reorganized doc directory
authorJosh Coalson <jcoalson@users.sourceforce.net>
Tue, 16 Jul 2002 16:09:53 +0000 (16:09 +0000)
committerJosh Coalson <jcoalson@users.sourceforce.net>
Tue, 16 Jul 2002 16:09:53 +0000 (16:09 +0000)
30 files changed:
doc/Doxyfile [new file with mode: 0644]
doc/Makefile.am [new file with mode: 0644]
doc/html/Makefile.am [new file with mode: 0644]
doc/html/comparison.html [new file with mode: 0644]
doc/html/developers.html [new file with mode: 0644]
doc/html/documentation.html [new file with mode: 0644]
doc/html/download.html [new file with mode: 0644]
doc/html/features.html [new file with mode: 0644]
doc/html/format.html [new file with mode: 0644]
doc/html/goals.html [new file with mode: 0644]
doc/html/id.html [new file with mode: 0644]
doc/html/images/1x1.gif [new file with mode: 0644]
doc/html/images/Makefile.am [new file with mode: 0644]
doc/html/images/cafebug.gif [new file with mode: 0644]
doc/html/images/logo.jpg [new file with mode: 0644]
doc/html/index.html [new file with mode: 0644]
doc/html/news.html [new file with mode: 0644]
doc/html/ru/Makefile.am [new file with mode: 0644]
doc/html/ru/authors.html [new file with mode: 0644]
doc/html/ru/comparison.html [new file with mode: 0644]
doc/html/ru/developers.html [new file with mode: 0644]
doc/html/ru/documentation.html [new file with mode: 0644]
doc/html/ru/download.html [new file with mode: 0644]
doc/html/ru/features.html [new file with mode: 0644]
doc/html/ru/format.html [new file with mode: 0644]
doc/html/ru/goals.html [new file with mode: 0644]
doc/html/ru/id.html [new file with mode: 0644]
doc/html/ru/index.html [new file with mode: 0644]
doc/html/ru/links.html [new file with mode: 0644]
doc/html/ru/news.html [new file with mode: 0644]

diff --git a/doc/Doxyfile b/doc/Doxyfile
new file mode 100644 (file)
index 0000000..7f4fcc9
--- /dev/null
@@ -0,0 +1,946 @@
+# Doxyfile 1.2.16
+
+# This file describes the settings to be used by the documentation system
+# doxygen (www.doxygen.org) for a project
+#
+# All text after a hash (#) is considered a comment and will be ignored
+# The format is:
+#       TAG = value [value, ...]
+# For lists items can also be appended using:
+#       TAG += value [value, ...]
+# Values that contain spaces should be placed between quotes (" ")
+
+#---------------------------------------------------------------------------
+# General configuration options
+#---------------------------------------------------------------------------
+
+# The PROJECT_NAME tag is a single word (or a sequence of words surrounded 
+# by quotes) that should identify the project.
+
+PROJECT_NAME           = FLAC
+
+# The PROJECT_NUMBER tag can be used to enter a project or revision number. 
+# This could be handy for archiving the generated documentation or 
+# if some version control system is used.
+
+PROJECT_NUMBER         = 1.0.3
+
+# The OUTPUT_DIRECTORY tag is used to specify the (relative or absolute) 
+# base path where the generated documentation will be put. 
+# If a relative path is entered, it will be relative to the location 
+# where doxygen was started. If left blank the current directory will be used.
+
+OUTPUT_DIRECTORY       = doxytmp
+
+# The OUTPUT_LANGUAGE tag is used to specify the language in which all 
+# documentation generated by doxygen is written. Doxygen will use this 
+# information to generate all constant output in the proper language. 
+# The default language is English, other supported languages are: 
+# Brazilian, Chinese, Chinese-Traditional, Croatian, Czech, Danish, Dutch, 
+# Finnish, French, German, Greek, Hungarian, Italian, Japanese, Korean, 
+# Norwegian, Polish, Portuguese, Romanian, Russian, Slovak, Slovene, 
+# Spanish, Swedish and Ukrainian.
+
+OUTPUT_LANGUAGE        = English
+
+# If the EXTRACT_ALL tag is set to YES doxygen will assume all entities in 
+# documentation are documented, even if no documentation was available. 
+# Private class members and static file members will be hidden unless 
+# the EXTRACT_PRIVATE and EXTRACT_STATIC tags are set to YES
+
+EXTRACT_ALL            = NO
+
+# If the EXTRACT_PRIVATE tag is set to YES all private members of a class 
+# will be included in the documentation.
+
+EXTRACT_PRIVATE        = NO
+
+# If the EXTRACT_STATIC tag is set to YES all static members of a file 
+# will be included in the documentation.
+
+EXTRACT_STATIC         = NO
+
+# If the EXTRACT_LOCAL_CLASSES tag is set to YES classes (and structs) 
+# defined locally in source files will be included in the documentation. 
+# If set to NO only classes defined in header files are included.
+
+EXTRACT_LOCAL_CLASSES  = YES
+
+# If the HIDE_UNDOC_MEMBERS tag is set to YES, Doxygen will hide all 
+# undocumented members of documented classes, files or namespaces. 
+# If set to NO (the default) these members will be included in the 
+# various overviews, but no documentation section is generated. 
+# This option has no effect if EXTRACT_ALL is enabled.
+
+HIDE_UNDOC_MEMBERS     = NO
+
+# If the HIDE_UNDOC_CLASSES tag is set to YES, Doxygen will hide all 
+# undocumented classes that are normally visible in the class hierarchy. 
+# If set to NO (the default) these class will be included in the various 
+# overviews. This option has no effect if EXTRACT_ALL is enabled.
+
+HIDE_UNDOC_CLASSES     = NO
+
+# If the BRIEF_MEMBER_DESC tag is set to YES (the default) Doxygen will 
+# include brief member descriptions after the members that are listed in 
+# the file and class documentation (similar to JavaDoc). 
+# Set to NO to disable this.
+
+BRIEF_MEMBER_DESC      = YES
+
+# If the REPEAT_BRIEF tag is set to YES (the default) Doxygen will prepend 
+# the brief description of a member or function before the detailed description. 
+# Note: if both HIDE_UNDOC_MEMBERS and BRIEF_MEMBER_DESC are set to NO, the 
+# brief descriptions will be completely suppressed.
+
+REPEAT_BRIEF           = YES
+
+# If the ALWAYS_DETAILED_SEC and REPEAT_BRIEF tags are both set to YES then 
+# Doxygen will generate a detailed section even if there is only a brief 
+# description.
+
+ALWAYS_DETAILED_SEC    = NO
+
+# If the INLINE_INHERITED_MEMB tag is set to YES, doxygen will show all inherited 
+# members of a class in the documentation of that class as if those members were 
+# ordinary class members. Constructors, destructors and assignment operators of 
+# the base classes will not be shown.
+
+INLINE_INHERITED_MEMB  = NO
+
+# If the FULL_PATH_NAMES tag is set to YES then Doxygen will prepend the full 
+# path before files name in the file list and in the header files. If set 
+# to NO the shortest path that makes the file name unique will be used.
+
+FULL_PATH_NAMES        = YES
+
+# If the FULL_PATH_NAMES tag is set to YES then the STRIP_FROM_PATH tag 
+# can be used to strip a user defined part of the path. Stripping is 
+# only done if one of the specified strings matches the left-hand part of 
+# the path. It is allowed to use relative paths in the argument list.
+
+STRIP_FROM_PATH        = 
+
+# The INTERNAL_DOCS tag determines if documentation 
+# that is typed after a \internal command is included. If the tag is set 
+# to NO (the default) then the documentation will be excluded. 
+# Set it to YES to include the internal documentation.
+
+INTERNAL_DOCS          = NO
+
+# Setting the STRIP_CODE_COMMENTS tag to YES (the default) will instruct 
+# doxygen to hide any special comment blocks from generated source code 
+# fragments. Normal C and C++ comments will always remain visible.
+
+STRIP_CODE_COMMENTS    = YES
+
+# If the CASE_SENSE_NAMES tag is set to NO then Doxygen will only generate 
+# file names in lower case letters. If set to YES upper case letters are also 
+# allowed. This is useful if you have classes or files whose names only differ 
+# in case and if your file system supports case sensitive file names. Windows 
+# users are adviced to set this option to NO.
+
+CASE_SENSE_NAMES       = YES
+
+# If the SHORT_NAMES tag is set to YES, doxygen will generate much shorter 
+# (but less readable) file names. This can be useful is your file systems 
+# doesn't support long names like on DOS, Mac, or CD-ROM.
+
+SHORT_NAMES            = NO
+
+# If the HIDE_SCOPE_NAMES tag is set to NO (the default) then Doxygen 
+# will show members with their full class and namespace scopes in the 
+# documentation. If set to YES the scope will be hidden.
+
+HIDE_SCOPE_NAMES       = NO
+
+# If the VERBATIM_HEADERS tag is set to YES (the default) then Doxygen 
+# will generate a verbatim copy of the header file for each class for 
+# which an include is specified. Set to NO to disable this.
+
+VERBATIM_HEADERS       = YES
+
+# If the SHOW_INCLUDE_FILES tag is set to YES (the default) then Doxygen 
+# will put list of the files that are included by a file in the documentation 
+# of that file.
+
+SHOW_INCLUDE_FILES     = YES
+
+# If the JAVADOC_AUTOBRIEF tag is set to YES then Doxygen 
+# will interpret the first line (until the first dot) of a JavaDoc-style 
+# comment as the brief description. If set to NO, the JavaDoc 
+# comments  will behave just like the Qt-style comments (thus requiring an 
+# explict @brief command for a brief description.
+
+JAVADOC_AUTOBRIEF      = NO
+
+# If the DETAILS_AT_TOP tag is set to YES then Doxygen 
+# will output the detailed description near the top, like JavaDoc.
+# If set to NO, the detailed description appears after the member 
+# documentation.
+
+DETAILS_AT_TOP         = YES
+
+# If the INHERIT_DOCS tag is set to YES (the default) then an undocumented 
+# member inherits the documentation from any documented member that it 
+# reimplements.
+
+INHERIT_DOCS           = YES
+
+# If the INLINE_INFO tag is set to YES (the default) then a tag [inline] 
+# is inserted in the documentation for inline members.
+
+INLINE_INFO            = YES
+
+# If the SORT_MEMBER_DOCS tag is set to YES (the default) then doxygen 
+# will sort the (detailed) documentation of file and class members 
+# alphabetically by member name. If set to NO the members will appear in 
+# declaration order.
+
+SORT_MEMBER_DOCS       = NO
+
+# If member grouping is used in the documentation and the DISTRIBUTE_GROUP_DOC 
+# tag is set to YES, then doxygen will reuse the documentation of the first 
+# member in the group (if any) for the other members of the group. By default 
+# all members of a group must be documented explicitly.
+
+DISTRIBUTE_GROUP_DOC   = NO
+
+# The TAB_SIZE tag can be used to set the number of spaces in a tab. 
+# Doxygen uses this value to replace tabs by spaces in code fragments.
+
+TAB_SIZE               = 4
+
+# The GENERATE_TODOLIST tag can be used to enable (YES) or 
+# disable (NO) the todo list. This list is created by putting \todo 
+# commands in the documentation.
+
+GENERATE_TODOLIST      = YES
+
+# The GENERATE_TESTLIST tag can be used to enable (YES) or 
+# disable (NO) the test list. This list is created by putting \test 
+# commands in the documentation.
+
+GENERATE_TESTLIST      = YES
+
+# The GENERATE_BUGLIST tag can be used to enable (YES) or 
+# disable (NO) the bug list. This list is created by putting \bug 
+# commands in the documentation.
+
+GENERATE_BUGLIST       = YES
+
+# This tag can be used to specify a number of aliases that acts 
+# as commands in the documentation. An alias has the form "name=value". 
+# For example adding "sideeffect=\par Side Effects:\n" will allow you to 
+# put the command \sideeffect (or @sideeffect) in the documentation, which 
+# will result in a user defined paragraph with heading "Side Effects:". 
+# You can put \n's in the value part of an alias to insert newlines.
+
+ALIASES                = "assert=\par Assertions:\n"
+
+# The ENABLED_SECTIONS tag can be used to enable conditional 
+# documentation sections, marked by \if sectionname ... \endif.
+
+ENABLED_SECTIONS       = 
+
+# The MAX_INITIALIZER_LINES tag determines the maximum number of lines 
+# the initial value of a variable or define consist of for it to appear in 
+# the documentation. If the initializer consists of more lines than specified 
+# here it will be hidden. Use a value of 0 to hide initializers completely. 
+# The appearance of the initializer of individual variables and defines in the 
+# documentation can be controlled using \showinitializer or \hideinitializer 
+# command in the documentation regardless of this setting.
+
+MAX_INITIALIZER_LINES  = 30
+
+# Set the OPTIMIZE_OUTPUT_FOR_C tag to YES if your project consists of C sources 
+# only. Doxygen will then generate output that is more tailored for C. 
+# For instance some of the names that are used will be different. The list 
+# of all members will be omitted, etc.
+
+OPTIMIZE_OUTPUT_FOR_C  = NO
+
+# Set the OPTIMIZE_OUTPUT_JAVA tag to YES if your project consists of Java sources 
+# only. Doxygen will then generate output that is more tailored for Java. 
+# For instance namespaces will be presented as packages, qualified scopes 
+# will look different, etc.
+
+OPTIMIZE_OUTPUT_JAVA   = NO
+
+# Set the SHOW_USED_FILES tag to NO to disable the list of files generated 
+# at the bottom of the documentation of classes and structs. If set to YES the 
+# list will mention the files that were used to generate the documentation.
+
+SHOW_USED_FILES        = YES
+
+#---------------------------------------------------------------------------
+# configuration options related to warning and progress messages
+#---------------------------------------------------------------------------
+
+# The QUIET tag can be used to turn on/off the messages that are generated 
+# by doxygen. Possible values are YES and NO. If left blank NO is used.
+
+QUIET                  = NO
+
+# The WARNINGS tag can be used to turn on/off the warning messages that are 
+# generated by doxygen. Possible values are YES and NO. If left blank 
+# NO is used.
+
+WARNINGS               = YES
+
+# If WARN_IF_UNDOCUMENTED is set to YES, then doxygen will generate warnings 
+# for undocumented members. If EXTRACT_ALL is set to YES then this flag will 
+# automatically be disabled.
+
+WARN_IF_UNDOCUMENTED   = YES
+
+# The WARN_FORMAT tag determines the format of the warning messages that 
+# doxygen can produce. The string should contain the $file, $line, and $text 
+# tags, which will be replaced by the file and line number from which the 
+# warning originated and the warning text.
+
+WARN_FORMAT            = "$file:$line: $text"
+
+# The WARN_LOGFILE tag can be used to specify a file to which warning 
+# and error messages should be written. If left blank the output is written 
+# to stderr.
+
+WARN_LOGFILE           = 
+
+#---------------------------------------------------------------------------
+# configuration options related to the input files
+#---------------------------------------------------------------------------
+
+# The INPUT tag can be used to specify the files and/or directories that contain 
+# documented source files. You may enter file names like "myfile.cpp" or 
+# directories like "/usr/src/myproject". Separate the files or directories 
+# with spaces.
+
+INPUT                  = ../include/FLAC ../include/FLAC++
+
+# If the value of the INPUT tag contains directories, you can use the 
+# FILE_PATTERNS tag to specify one or more wildcard pattern (like *.cpp 
+# and *.h) to filter out the source-files in the directories. If left 
+# blank the following patterns are tested: 
+# *.c *.cc *.cxx *.cpp *.c++ *.java *.ii *.ixx *.ipp *.i++ *.inl *.h *.hh *.hxx *.hpp 
+# *.h++ *.idl *.odl
+
+FILE_PATTERNS          = 
+
+# The RECURSIVE tag can be used to turn specify whether or not subdirectories 
+# should be searched for input files as well. Possible values are YES and NO. 
+# If left blank NO is used.
+
+RECURSIVE              = NO
+
+# The EXCLUDE tag can be used to specify files and/or directories that should 
+# excluded from the INPUT source files. This way you can easily exclude a 
+# subdirectory from a directory tree whose root is specified with the INPUT tag.
+
+EXCLUDE                = 
+
+# The EXCLUDE_SYMLINKS tag can be used select whether or not files or directories 
+# that are symbolic links (a Unix filesystem feature) are excluded from the input.
+
+EXCLUDE_SYMLINKS       = NO
+
+# If the value of the INPUT tag contains directories, you can use the 
+# EXCLUDE_PATTERNS tag to specify one or more wildcard patterns to exclude 
+# certain files from those directories.
+
+EXCLUDE_PATTERNS       = 
+
+# The EXAMPLE_PATH tag can be used to specify one or more files or 
+# directories that contain example code fragments that are included (see 
+# the \include command).
+
+EXAMPLE_PATH           = 
+
+# If the value of the EXAMPLE_PATH tag contains directories, you can use the 
+# EXAMPLE_PATTERNS tag to specify one or more wildcard pattern (like *.cpp 
+# and *.h) to filter out the source-files in the directories. If left 
+# blank all files are included.
+
+EXAMPLE_PATTERNS       = 
+
+# If the EXAMPLE_RECURSIVE tag is set to YES then subdirectories will be 
+# searched for input files to be used with the \include or \dontinclude 
+# commands irrespective of the value of the RECURSIVE tag. 
+# Possible values are YES and NO. If left blank NO is used.
+
+EXAMPLE_RECURSIVE      = NO
+
+# The IMAGE_PATH tag can be used to specify one or more files or 
+# directories that contain image that are included in the documentation (see 
+# the \image command).
+
+IMAGE_PATH             = 
+
+# The INPUT_FILTER tag can be used to specify a program that doxygen should 
+# invoke to filter for each input file. Doxygen will invoke the filter program 
+# by executing (via popen()) the command <filter> <input-file>, where <filter> 
+# is the value of the INPUT_FILTER tag, and <input-file> is the name of an 
+# input file. Doxygen will then use the output that the filter program writes 
+# to standard output.
+
+INPUT_FILTER           = 
+
+# If the FILTER_SOURCE_FILES tag is set to YES, the input filter (if set using 
+# INPUT_FILTER) will be used to filter the input files when producing source 
+# files to browse.
+
+FILTER_SOURCE_FILES    = NO
+
+#---------------------------------------------------------------------------
+# configuration options related to source browsing
+#---------------------------------------------------------------------------
+
+# If the SOURCE_BROWSER tag is set to YES then a list of source files will 
+# be generated. Documented entities will be cross-referenced with these sources.
+
+SOURCE_BROWSER         = NO
+
+# Setting the INLINE_SOURCES tag to YES will include the body 
+# of functions and classes directly in the documentation.
+
+INLINE_SOURCES         = NO
+
+# If the REFERENCED_BY_RELATION tag is set to YES (the default) 
+# then for each documented function all documented 
+# functions referencing it will be listed.
+
+REFERENCED_BY_RELATION = YES
+
+# If the REFERENCES_RELATION tag is set to YES (the default) 
+# then for each documented function all documented entities 
+# called/used by that function will be listed.
+
+REFERENCES_RELATION    = YES
+
+#---------------------------------------------------------------------------
+# configuration options related to the alphabetical class index
+#---------------------------------------------------------------------------
+
+# If the ALPHABETICAL_INDEX tag is set to YES, an alphabetical index 
+# of all compounds will be generated. Enable this if the project 
+# contains a lot of classes, structs, unions or interfaces.
+
+ALPHABETICAL_INDEX     = YES
+
+# If the alphabetical index is enabled (see ALPHABETICAL_INDEX) then 
+# the COLS_IN_ALPHA_INDEX tag can be used to specify the number of columns 
+# in which this list will be split (can be a number in the range [1..20])
+
+COLS_IN_ALPHA_INDEX    = 5
+
+# In case all classes in a project start with a common prefix, all 
+# classes will be put under the same header in the alphabetical index. 
+# The IGNORE_PREFIX tag can be used to specify one or more prefixes that 
+# should be ignored while generating the index headers.
+
+IGNORE_PREFIX          = 
+
+#---------------------------------------------------------------------------
+# configuration options related to the HTML output
+#---------------------------------------------------------------------------
+
+# If the GENERATE_HTML tag is set to YES (the default) Doxygen will 
+# generate HTML output.
+
+GENERATE_HTML          = YES
+
+# The HTML_OUTPUT tag is used to specify where the HTML docs will be put. 
+# If a relative path is entered the value of OUTPUT_DIRECTORY will be 
+# put in front of it. If left blank `html' will be used as the default path.
+
+HTML_OUTPUT            = html
+
+# The HTML_FILE_EXTENSION tag can be used to specify the file extension for 
+# each generated HTML page (for example: .htm,.php,.asp). If it is left blank 
+# doxygen will generate files with .html extension.
+
+HTML_FILE_EXTENSION    = .html
+
+# The HTML_HEADER tag can be used to specify a personal HTML header for 
+# each generated HTML page. If it is left blank doxygen will generate a 
+# standard header.
+
+HTML_HEADER            = 
+
+# The HTML_FOOTER tag can be used to specify a personal HTML footer for 
+# each generated HTML page. If it is left blank doxygen will generate a 
+# standard footer.
+
+HTML_FOOTER            = 
+
+# The HTML_STYLESHEET tag can be used to specify a user defined cascading 
+# style sheet that is used by each HTML page. It can be used to 
+# fine-tune the look of the HTML output. If the tag is left blank doxygen 
+# will generate a default style sheet
+
+HTML_STYLESHEET        = 
+
+# If the HTML_ALIGN_MEMBERS tag is set to YES, the members of classes, 
+# files or namespaces will be aligned in HTML using tables. If set to 
+# NO a bullet list will be used.
+
+HTML_ALIGN_MEMBERS     = YES
+
+# If the GENERATE_HTMLHELP tag is set to YES, additional index files 
+# will be generated that can be used as input for tools like the 
+# Microsoft HTML help workshop to generate a compressed HTML help file (.chm) 
+# of the generated HTML documentation.
+
+GENERATE_HTMLHELP      = NO
+
+# If the GENERATE_HTMLHELP tag is set to YES, the GENERATE_CHI flag 
+# controls if a separate .chi index file is generated (YES) or that 
+# it should be included in the master .chm file (NO).
+
+GENERATE_CHI           = NO
+
+# If the GENERATE_HTMLHELP tag is set to YES, the BINARY_TOC flag 
+# controls whether a binary table of contents is generated (YES) or a 
+# normal table of contents (NO) in the .chm file.
+
+BINARY_TOC             = NO
+
+# The TOC_EXPAND flag can be set to YES to add extra items for group members 
+# to the contents of the Html help documentation and to the tree view.
+
+TOC_EXPAND             = NO
+
+# The DISABLE_INDEX tag can be used to turn on/off the condensed index at 
+# top of each HTML page. The value NO (the default) enables the index and 
+# the value YES disables it.
+
+DISABLE_INDEX          = NO
+
+# This tag can be used to set the number of enum values (range [1..20]) 
+# that doxygen will group on one line in the generated HTML documentation.
+
+ENUM_VALUES_PER_LINE   = 4
+
+# If the GENERATE_TREEVIEW tag is set to YES, a side panel will be
+# generated containing a tree-like index structure (just like the one that 
+# is generated for HTML Help). For this to work a browser that supports 
+# JavaScript and frames is required (for instance Mozilla, Netscape 4.0+, 
+# or Internet explorer 4.0+). Note that for large projects the tree generation 
+# can take a very long time. In such cases it is better to disable this feature. 
+# Windows users are probably better off using the HTML help feature.
+
+GENERATE_TREEVIEW      = NO
+
+# If the treeview is enabled (see GENERATE_TREEVIEW) then this tag can be 
+# used to set the initial width (in pixels) of the frame in which the tree 
+# is shown.
+
+TREEVIEW_WIDTH         = 250
+
+#---------------------------------------------------------------------------
+# configuration options related to the LaTeX output
+#---------------------------------------------------------------------------
+
+# If the GENERATE_LATEX tag is set to YES (the default) Doxygen will 
+# generate Latex output.
+
+GENERATE_LATEX         = YES
+
+# The LATEX_OUTPUT tag is used to specify where the LaTeX docs will be put. 
+# If a relative path is entered the value of OUTPUT_DIRECTORY will be 
+# put in front of it. If left blank `latex' will be used as the default path.
+
+LATEX_OUTPUT           = latex
+
+# The LATEX_CMD_NAME tag can be used to specify the LaTeX command name to be invoked. If left blank `latex' will be used as the default command name.
+
+LATEX_CMD_NAME         = latex
+
+# The MAKEINDEX_CMD_NAME tag can be used to specify the command name to 
+# generate index for LaTeX. If left blank `makeindex' will be used as the 
+# default command name.
+
+MAKEINDEX_CMD_NAME     = makeindex
+
+# If the COMPACT_LATEX tag is set to YES Doxygen generates more compact 
+# LaTeX documents. This may be useful for small projects and may help to 
+# save some trees in general.
+
+COMPACT_LATEX          = NO
+
+# The PAPER_TYPE tag can be used to set the paper type that is used 
+# by the printer. Possible values are: a4, a4wide, letter, legal and 
+# executive. If left blank a4wide will be used.
+
+PAPER_TYPE             = a4wide
+
+# The EXTRA_PACKAGES tag can be to specify one or more names of LaTeX 
+# packages that should be included in the LaTeX output.
+
+EXTRA_PACKAGES         = 
+
+# The LATEX_HEADER tag can be used to specify a personal LaTeX header for 
+# the generated latex document. The header should contain everything until 
+# the first chapter. If it is left blank doxygen will generate a 
+# standard header. Notice: only use this tag if you know what you are doing!
+
+LATEX_HEADER           = 
+
+# If the PDF_HYPERLINKS tag is set to YES, the LaTeX that is generated 
+# is prepared for conversion to pdf (using ps2pdf). The pdf file will 
+# contain links (just like the HTML output) instead of page references 
+# This makes the output suitable for online browsing using a pdf viewer.
+
+PDF_HYPERLINKS         = NO
+
+# If the USE_PDFLATEX tag is set to YES, pdflatex will be used instead of 
+# plain latex in the generated Makefile. Set this option to YES to get a 
+# higher quality PDF documentation.
+
+USE_PDFLATEX           = NO
+
+# If the LATEX_BATCHMODE tag is set to YES, doxygen will add the \\batchmode. 
+# command to the generated LaTeX files. This will instruct LaTeX to keep 
+# running if errors occur, instead of asking the user for help. 
+# This option is also used when generating formulas in HTML.
+
+LATEX_BATCHMODE        = NO
+
+#---------------------------------------------------------------------------
+# configuration options related to the RTF output
+#---------------------------------------------------------------------------
+
+# If the GENERATE_RTF tag is set to YES Doxygen will generate RTF output 
+# The RTF output is optimised for Word 97 and may not look very pretty with 
+# other RTF readers or editors.
+
+GENERATE_RTF           = NO
+
+# The RTF_OUTPUT tag is used to specify where the RTF docs will be put. 
+# If a relative path is entered the value of OUTPUT_DIRECTORY will be 
+# put in front of it. If left blank `rtf' will be used as the default path.
+
+RTF_OUTPUT             = rtf
+
+# If the COMPACT_RTF tag is set to YES Doxygen generates more compact 
+# RTF documents. This may be useful for small projects and may help to 
+# save some trees in general.
+
+COMPACT_RTF            = NO
+
+# If the RTF_HYPERLINKS tag is set to YES, the RTF that is generated 
+# will contain hyperlink fields. The RTF file will 
+# contain links (just like the HTML output) instead of page references. 
+# This makes the output suitable for online browsing using WORD or other 
+# programs which support those fields. 
+# Note: wordpad (write) and others do not support links.
+
+RTF_HYPERLINKS         = NO
+
+# Load stylesheet definitions from file. Syntax is similar to doxygen's 
+# config file, i.e. a series of assigments. You only have to provide 
+# replacements, missing definitions are set to their default value.
+
+RTF_STYLESHEET_FILE    = 
+
+# Set optional variables used in the generation of an rtf document. 
+# Syntax is similar to doxygen's config file.
+
+RTF_EXTENSIONS_FILE    = 
+
+#---------------------------------------------------------------------------
+# configuration options related to the man page output
+#---------------------------------------------------------------------------
+
+# If the GENERATE_MAN tag is set to YES (the default) Doxygen will 
+# generate man pages
+
+GENERATE_MAN           = YES
+
+# The MAN_OUTPUT tag is used to specify where the man pages will be put. 
+# If a relative path is entered the value of OUTPUT_DIRECTORY will be 
+# put in front of it. If left blank `man' will be used as the default path.
+
+MAN_OUTPUT             = man
+
+# The MAN_EXTENSION tag determines the extension that is added to 
+# the generated man pages (default is the subroutine's section .3)
+
+MAN_EXTENSION          = .3
+
+# If the MAN_LINKS tag is set to YES and Doxygen generates man output, 
+# then it will generate one additional man file for each entity 
+# documented in the real man page(s). These additional files 
+# only source the real man page, but without them the man command 
+# would be unable to find the correct page. The default is NO.
+
+MAN_LINKS              = NO
+
+#---------------------------------------------------------------------------
+# configuration options related to the XML output
+#---------------------------------------------------------------------------
+
+# If the GENERATE_XML tag is set to YES Doxygen will 
+# generate an XML file that captures the structure of 
+# the code including all documentation. Note that this 
+# feature is still experimental and incomplete at the 
+# moment.
+
+GENERATE_XML           = NO
+
+#---------------------------------------------------------------------------
+# configuration options for the AutoGen Definitions output
+#---------------------------------------------------------------------------
+
+# If the GENERATE_AUTOGEN_DEF tag is set to YES Doxygen will 
+# generate an AutoGen Definitions (see autogen.sf.net) file 
+# that captures the structure of the code including all 
+# documentation. Note that this feature is still experimental 
+# and incomplete at the moment.
+
+GENERATE_AUTOGEN_DEF   = NO
+
+#---------------------------------------------------------------------------
+# Configuration options related to the preprocessor   
+#---------------------------------------------------------------------------
+
+# If the ENABLE_PREPROCESSING tag is set to YES (the default) Doxygen will 
+# evaluate all C-preprocessor directives found in the sources and include 
+# files.
+
+ENABLE_PREPROCESSING   = YES
+
+# If the MACRO_EXPANSION tag is set to YES Doxygen will expand all macro 
+# names in the source code. If set to NO (the default) only conditional 
+# compilation will be performed. Macro expansion can be done in a controlled 
+# way by setting EXPAND_ONLY_PREDEF to YES.
+
+MACRO_EXPANSION        = NO
+
+# If the EXPAND_ONLY_PREDEF and MACRO_EXPANSION tags are both set to YES 
+# then the macro expansion is limited to the macros specified with the 
+# PREDEFINED and EXPAND_AS_PREDEFINED tags.
+
+EXPAND_ONLY_PREDEF     = NO
+
+# If the SEARCH_INCLUDES tag is set to YES (the default) the includes files 
+# in the INCLUDE_PATH (see below) will be search if a #include is found.
+
+SEARCH_INCLUDES        = YES
+
+# The INCLUDE_PATH tag can be used to specify one or more directories that 
+# contain include files that are not input files but should be processed by 
+# the preprocessor.
+
+INCLUDE_PATH           = 
+
+# You can use the INCLUDE_FILE_PATTERNS tag to specify one or more wildcard 
+# patterns (like *.h and *.hpp) to filter out the header-files in the 
+# directories. If left blank, the patterns specified with FILE_PATTERNS will 
+# be used.
+
+INCLUDE_FILE_PATTERNS  = 
+
+# The PREDEFINED tag can be used to specify one or more macro names that 
+# are defined before the preprocessor is started (similar to the -D option of 
+# gcc). The argument of the tag is a list of macros of the form: name 
+# or name=definition (no spaces). If the definition and the = are 
+# omitted =1 is assumed.
+
+PREDEFINED             = 
+
+# If the MACRO_EXPANSION and EXPAND_PREDEF_ONLY tags are set to YES then 
+# this tag can be used to specify a list of macro names that should be expanded. 
+# The macro definition that is found in the sources will be used. 
+# Use the PREDEFINED tag if you want to use a different macro definition.
+
+EXPAND_AS_DEFINED      = 
+
+# If the SKIP_FUNCTION_MACROS tag is set to YES (the default) then 
+# doxygen's preprocessor will remove all function-like macros that are alone 
+# on a line and do not end with a semicolon. Such function macros are typically 
+# used for boiler-plate code, and will confuse the parser if not removed.
+
+SKIP_FUNCTION_MACROS   = YES
+
+#---------------------------------------------------------------------------
+# Configuration::addtions related to external references   
+#---------------------------------------------------------------------------
+
+# The TAGFILES tag can be used to specify one or more tagfiles.
+
+TAGFILES               = 
+
+# When a file name is specified after GENERATE_TAGFILE, doxygen will create 
+# a tag file that is based on the input files it reads.
+
+GENERATE_TAGFILE       = FLAC.tag
+
+# If the ALLEXTERNALS tag is set to YES all external classes will be listed 
+# in the class index. If set to NO only the inherited external classes 
+# will be listed.
+
+ALLEXTERNALS           = NO
+
+# If the EXTERNAL_GROUPS tag is set to YES all external groups will be listed 
+# in the modules index. If set to NO, only the current project's groups will 
+# be listed.
+
+EXTERNAL_GROUPS        = YES
+
+# The PERL_PATH should be the absolute path and name of the perl script 
+# interpreter (i.e. the result of `which perl').
+
+PERL_PATH              = /usr/bin/perl
+
+#---------------------------------------------------------------------------
+# Configuration options related to the dot tool   
+#---------------------------------------------------------------------------
+
+# If the CLASS_DIAGRAMS tag is set to YES (the default) Doxygen will 
+# generate a inheritance diagram (in Html, RTF and LaTeX) for classes with base or 
+# super classes. Setting the tag to NO turns the diagrams off. Note that this 
+# option is superceded by the HAVE_DOT option below. This is only a fallback. It is 
+# recommended to install and use dot, since it yield more powerful graphs.
+
+CLASS_DIAGRAMS         = YES
+
+# If set to YES, the inheritance and collaboration graphs will hide 
+# inheritance and usage relations if the target is undocumented 
+# or is not a class.
+
+HIDE_UNDOC_RELATIONS   = YES
+
+# If you set the HAVE_DOT tag to YES then doxygen will assume the dot tool is 
+# available from the path. This tool is part of Graphviz, a graph visualization 
+# toolkit from AT&T and Lucent Bell Labs. The other options in this section 
+# have no effect if this option is set to NO (the default)
+
+HAVE_DOT               = NO
+
+# If the CLASS_GRAPH and HAVE_DOT tags are set to YES then doxygen 
+# will generate a graph for each documented class showing the direct and 
+# indirect inheritance relations. Setting this tag to YES will force the 
+# the CLASS_DIAGRAMS tag to NO.
+
+CLASS_GRAPH            = YES
+
+# If the COLLABORATION_GRAPH and HAVE_DOT tags are set to YES then doxygen 
+# will generate a graph for each documented class showing the direct and 
+# indirect implementation dependencies (inheritance, containment, and 
+# class references variables) of the class with other documented classes.
+
+COLLABORATION_GRAPH    = YES
+
+# If set to YES, the inheritance and collaboration graphs will show the 
+# relations between templates and their instances.
+
+TEMPLATE_RELATIONS     = YES
+
+# If the ENABLE_PREPROCESSING, SEARCH_INCLUDES, INCLUDE_GRAPH, and HAVE_DOT 
+# tags are set to YES then doxygen will generate a graph for each documented 
+# file showing the direct and indirect include dependencies of the file with 
+# other documented files.
+
+INCLUDE_GRAPH          = YES
+
+# If the ENABLE_PREPROCESSING, SEARCH_INCLUDES, INCLUDED_BY_GRAPH, and 
+# HAVE_DOT tags are set to YES then doxygen will generate a graph for each 
+# documented header file showing the documented files that directly or 
+# indirectly include this file.
+
+INCLUDED_BY_GRAPH      = YES
+
+# If the GRAPHICAL_HIERARCHY and HAVE_DOT tags are set to YES then doxygen 
+# will graphical hierarchy of all classes instead of a textual one.
+
+GRAPHICAL_HIERARCHY    = YES
+
+# The DOT_IMAGE_FORMAT tag can be used to set the image format of the images 
+# generated by dot. Possible values are png, jpg, or gif
+# If left blank png will be used.
+
+DOT_IMAGE_FORMAT       = png
+
+# The tag DOT_PATH can be used to specify the path where the dot tool can be 
+# found. If left blank, it is assumed the dot tool can be found on the path.
+
+DOT_PATH               = 
+
+# The DOTFILE_DIRS tag can be used to specify one or more directories that 
+# contain dot files that are included in the documentation (see the 
+# \dotfile command).
+
+DOTFILE_DIRS           = 
+
+# The MAX_DOT_GRAPH_WIDTH tag can be used to set the maximum allowed width 
+# (in pixels) of the graphs generated by dot. If a graph becomes larger than 
+# this value, doxygen will try to truncate the graph, so that it fits within 
+# the specified constraint. Beware that most browsers cannot cope with very 
+# large images.
+
+MAX_DOT_GRAPH_WIDTH    = 1024
+
+# The MAX_DOT_GRAPH_HEIGHT tag can be used to set the maximum allows height 
+# (in pixels) of the graphs generated by dot. If a graph becomes larger than 
+# this value, doxygen will try to truncate the graph, so that it fits within 
+# the specified constraint. Beware that most browsers cannot cope with very 
+# large images.
+
+MAX_DOT_GRAPH_HEIGHT   = 1024
+
+# If the GENERATE_LEGEND tag is set to YES (the default) Doxygen will 
+# generate a legend page explaining the meaning of the various boxes and 
+# arrows in the dot generated graphs.
+
+GENERATE_LEGEND        = YES
+
+# If the DOT_CLEANUP tag is set to YES (the default) Doxygen will 
+# remove the intermedate dot files that are used to generate 
+# the various graphs.
+
+DOT_CLEANUP            = YES
+
+#---------------------------------------------------------------------------
+# Configuration::addtions related to the search engine   
+#---------------------------------------------------------------------------
+
+# The SEARCHENGINE tag specifies whether or not a search engine should be 
+# used. If set to NO the values of all tags below this one will be ignored.
+
+SEARCHENGINE           = NO
+
+# The CGI_NAME tag should be the name of the CGI script that 
+# starts the search engine (doxysearch) with the correct parameters. 
+# A script with this name will be generated by doxygen.
+
+CGI_NAME               = search.cgi
+
+# The CGI_URL tag should be the absolute URL to the directory where the 
+# cgi binaries are located. See the documentation of your http daemon for 
+# details.
+
+CGI_URL                = 
+
+# The DOC_URL tag should be the absolute URL to the directory where the 
+# documentation is located. If left blank the absolute path to the 
+# documentation, with file:// prepended to it, will be used.
+
+DOC_URL                = 
+
+# The DOC_ABSPATH tag should be the absolute path to the directory where the 
+# documentation is located. If left blank the directory on the local machine 
+# will be used.
+
+DOC_ABSPATH            = 
+
+# The BIN_ABSPATH tag must point to the directory where the doxysearch binary 
+# is installed.
+
+BIN_ABSPATH            = /usr/local/bin/
+
+# The EXT_DOC_PATHS tag can be used to specify one or more paths to 
+# documentation generated for other projects. This allows doxysearch to search 
+# the documentation for these projects as well.
+
+EXT_DOC_PATHS          = 
diff --git a/doc/Makefile.am b/doc/Makefile.am
new file mode 100644 (file)
index 0000000..315e9e0
--- /dev/null
@@ -0,0 +1,33 @@
+#  flac - Command-line FLAC encoder/decoder
+#  Copyright (C) 2002  Josh Coalson
+#
+#  This program is free software; you can redistribute it and/or
+#  modify it under the terms of the GNU General Public License
+#  as published by the Free Software Foundation; either version 2
+#  of the License, or (at your option) any later version.
+#
+#  This program is distributed in the hope that it will be useful,
+#  but WITHOUT ANY WARRANTY; without even the implied warranty of
+#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+#  GNU General Public License for more details.
+#
+#  You should have received a copy of the GNU General Public License
+#  along with this program; if not, write to the Free Software
+#  Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
+
+AUTOMAKE_OPTIONS = foreign
+
+SUBDIRS = . html
+
+FLAC.tag: Doxyfile
+       doxygen Doxyfile
+       rm -rf html/api
+       mv doxytmp/html html/api
+       rm -rf doxytmp
+
+docdir = $(datadir)/doc/$(PACKAGE)-$(VERSION)
+
+doc_DATA =  \
+       FLAC.tag
+
+EXTRA_DIST = Doxyfile $(doc_DATA)
diff --git a/doc/html/Makefile.am b/doc/html/Makefile.am
new file mode 100644 (file)
index 0000000..34b18eb
--- /dev/null
@@ -0,0 +1,37 @@
+#  FLAC - Free Lossless Audio Codec
+#  Copyright (C) 2001,2002  Josh Coalson
+#
+#  This program is part of FLAC; you can redistribute it and/or
+#  modify it under the terms of the GNU General Public License
+#  as published by the Free Software Foundation; either version 2
+#  of the License, or (at your option) any later version.
+#
+#  This program is distributed in the hope that it will be useful,
+#  but WITHOUT ANY WARRANTY; without even the implied warranty of
+#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+#  GNU General Public License for more details.
+#
+#  You should have received a copy of the GNU General Public License
+#  along with this program; if not, write to the Free Software
+#  Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
+
+AUTOMAKE_OPTIONS = foreign
+
+SUBDIRS = ru images
+
+docdir = $(datadir)/doc/$(PACKAGE)-$(VERSION)/html
+
+doc_DATA =  \
+       comparison.html \
+       developers.html \
+       documentation.html \
+       download.html \
+       features.html \
+       format.html \
+       goals.html \
+       id.html \
+       index.html \
+       news.html \
+       api
+
+EXTRA_DIST = $(doc_DATA)
diff --git a/doc/html/comparison.html b/doc/html/comparison.html
new file mode 100644 (file)
index 0000000..41c6761
--- /dev/null
@@ -0,0 +1,1067 @@
+<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
+<!-- Copyright (c)  2000,2001,2002  Josh Coalson -->
+<!-- Permission is granted to copy, distribute and/or modify this document -->
+<!-- under the terms of the GNU Free Documentation License, Version 1.1 -->
+<!-- or any later version published by the Free Software Foundation; -->
+<!-- with no invariant sections. -->
+<!-- A copy of the license can be found at http://www.gnu.org/copyleft/fdl.html -->
+<HTML>
+<HEAD>
+       <TITLE>FLAC - comparison</TITLE>
+</HEAD>
+
+<BODY MARGINWIDTH="0" MARGINHEIGHT="0" LEFTMARGIN="0" RIGHTMARGIN="0" TOPMARGIN="0" BGCOLOR="#99CC99" TEXT="#000000" LINK="#336699" VLINK="#336699" ALINK="#336699">
+
+<TABLE BORDER=0 WIDTH="100%" CELLPADDING=1 CELLSPACING=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#000000"><A HREF="http://flac.sourceforge.net/"><IMG SRC="images/logo.jpg" ALIGN=CENTER ALT="FLAC Logo" BORDER=0 HSPACE=0></a></TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="25" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#D3D4C5">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="index.html">home</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="news.html">news</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="download.html">download</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="features.html">features</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="goals.html">goals</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="format.html">format</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="id.html">id</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;comparison&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="documentation.html">documentation</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="developers.html">developers</A>&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#EEEED4">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;english&nbsp;&nbsp;</TD><TD BGCOLOR="#EEEED4" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="ru/comparison.html">russian</A>&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<CENTER>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="15" ALT=""></TD></TR></TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <B><FONT SIZE="+2">comparison</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               The purpose of the comparison page is not only to show how compression ratios and encoding/decoding times using the flac reference encoder compare to other lossless encoders, but also to compare features (for example, some coders archive only and files must be uncompressed completely before playback can start).  Keep a few things in mind:
+       </P>
+       <P>
+       <UL>
+               <LI>
+                       As far as I know, only three of the lossless encoders out there (Bonk, flac and Kexis) are truly free (source code for Shorten and Monkey's Audio is available but the licenses are more restrictive).  Most others give out free binaries, but without access to the source, you are leaving your data to the whim of the maintainer for eternity; you have no way to port the program to another OS or fix it if it breaks.  This can be a serious drawback unless the format has world-class clout (like MP3).
+               </LI>
+               <LI>
+                       The compression ratios and times for flac are representative only of the reference encoder.  They are not indicative of the limits of all FLAC encoders or the FLAC format itself since the format is open and extensible, and anyone is free to write a better FLAC encoder.  And it is almost certain that the reference encoder itself will improve.
+               </LI>
+               <LI>
+                       Since FLAC supports streaming, it is at a slight disadvantage to the formats that don't because they don't have the extra overhead of all those frame headers.
+               </LI>
+       </UL>
+       </P>
+       <P>
+               I make an effort to keep this information as accurate as possible, but if any of the data is wrong, <A HREF="mailto:jcoalson@users.sourceforge.net">let me know</A> and I'll correct it.
+       </P>
+       <P>
+               Reviewed encoders:
+       </P>
+       <P>
+       <UL>
+               <LI>
+                       <A HREF="index.html">flac</A> of course.
+               </LI>
+               <LI>
+                       <A HREF="http://yoyo.cc.monash.edu.au/~phf/bonk">Bonk</A> - An open-source source codec.  No player or library support yet.
+               </LI>
+               <LI>
+                       <A HREF="http://sourceforge.net/projects/kexis">Kexis</A> - An open-source source codec.  Still in the alpha stage.  No player or library support yet.
+               </LI>
+               <LI>
+                       <A HREF="http://www-ft.ee.tu-berlin.de/~liebchen/lpac.html">LPAC</A> - A closed source codec.  At least it's available for more than just Windows, but there's only a Winamp plugin.
+               </LI>
+               <LI>
+                       <A HREF="http://www.monkeysaudio.com/">Monkey's Audio</A> - A symmetric adaptive codec with good compression.  Source is available under a non-OSI license.  There are two versions available now, one by the original author Matt Ashland and one by Frank Klemm.  The one tested here is from the original author.
+               </LI>
+               <LI>
+                       Ogg Squish - An open source source codec that is no longer maintained.  The version I tested, 0.98, was the latest I could find.  I don't have Windows timing results but it is among the 'fast' coders, based on UNIX tests.
+               </LI>
+               <LI>
+                       <A HREF="http://ghido.shelter.ro/">optimFROG</A> - A closed source, Windows/Linux codec, with Winamp and XMMS plugins.  Slow but best compression ratios.
+               </LI>
+               <LI>
+                       <A HREF="http://www.jpg.com/products/sound.html">Pegasus-SPS</A> - A closed source, Windows-only codec.
+               </LI>
+               <LI>
+                       <A HREF="http://rksoft.virtualave.net/">RKAU</A> - A closed source, Windows-only codec.
+               </LI>
+               <LI>
+                       <A HREF="http://www.softsound.com/Shorten.html">Shorten</A> - A.J. Robinson's well-known codec; source is available <A HREF="http://rpmfind.net/linux/rpm2html/search.php?query=shorten">here</A>.
+               </LI>
+               <LI>
+                       WaveZIP - A closed source, Windows-only archiver.  Uses the <A HREF="http://members.aol.com/_ht_a/sndspace/index.html">MUSICompress</A>[tm] engine which supposedly has a patent.  I used to have a link to the company that makes WaveZIP (GadgetLabs) but apparently they have gone out of business (maybe for trying to sell something that should cost nothing).
+               </LI>
+               <LI>
+                       <A HREF="http://www.wavpack.com/">WavPack</A> - A closed source, Windows-only archiver.
+               </LI>
+       </UL>
+       </P>
+       <P>
+               Encoders I couldn't get a copy of:
+       </P>
+       <P>
+       <UL>
+               <LI>
+                       AudioPak
+               </LI>
+               <LI>
+                       WavARC
+               </LI>
+       </UL>
+       </P>
+       <P>
+               If you take maximum compression ratio and speed out of the picture (as you will see later, most coders exhibit similar performance), here is a subjective sort based on overall "usefulness".  As far as features go, having source code gives you the most freedom since you can add anything you need that is missing; besides, open source projects tend to get better faster than closed source ones.  A close second (depending on the user) would be OS support or plugin support.
+       </P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD ALIGN="RIGHT">
+                               <FONT SIZE="+1"><B>Codec</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Source Available?</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Plugins Available?</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Hardware Support?</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Streamable?</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Seekable?</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Cost</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>OS support</B></FONT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               flac v1.0.3
+                       </TD>
+                       <TD>
+                               YES (<A HREF="http://www.opensource.org/licenses/index.html">OSI</A> approved license)
+                       </TD>
+                       <TD>
+                               YES (<A HREF="http://www.winamp.com/">Winamp</A>, <A HREF="http://www.xmms.org/">XMMS</A>, <A HREF="http://www.hut.fi/~hylinen/apollo/Plug-ins.html">Apollo</A>, <A HREF="http://www.dbpoweramp.com/">dBpowerAMP</A>, <A HREF="http://alsaplayer.org/">AlsaPlayer</A>)
+                       </TD>
+                       <TD>
+                               YES (<A HREF="http://www.phatnoise.com/forum/showthread.php?threadid=74">PhatBox</A>, Kenwood MusicKeg)
+                       </TD>
+                       <TD>
+                               YES
+                       </TD>
+                       <TD>
+                               YES
+                       </TD>
+                       <TD>
+                               FREE
+                       </TD>
+                       <TD>
+                               ANY (source)
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               Shorten v3.2
+                       </TD>
+                       <TD>
+                               YES (non-<A HREF="http://www.opensource.org/licenses/index.html">OSI</A> license)
+                       </TD>
+                       <TD>
+                               YES (Winamp, XMMS)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               YES (v3 only)
+                       </TD>
+                       <TD>
+                               FREE
+                       </TD>
+                       <TD>
+                               ANY (source)
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               Monkey's Audio v3.96
+                       </TD>
+                       <TD>
+                               YES (non-<A HREF="http://www.opensource.org/licenses/index.html">OSI</A> license)
+                       </TD>
+                       <TD>
+                               YES (Winamp, MediaJukebox, dBpowerAMP)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               YES
+                       </TD>
+                       <TD>
+                               FREE
+                       </TD>
+                       <TD>
+                               Windows/Linux console source
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               Ogg Squish 0.98
+                       </TD>
+                       <TD>
+                               YES (<A HREF="http://www.opensource.org/licenses/index.html">OSI</A> approved license)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no (?)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               YES
+                       </TD>
+                       <TD>
+                               YES
+                       </TD>
+                       <TD>
+                               FREE
+                       </TD>
+                       <TD>
+                               ANY (source)
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               Bonk 0.5
+                       </TD>
+                       <TD>
+                               YES (<A HREF="http://www.opensource.org/licenses/index.html">OSI</A> approved license)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               FREE
+                       </TD>
+                       <TD>
+                               ANY (source)
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               Kexis 0.2.2
+                       </TD>
+                       <TD>
+                               YES (<A HREF="http://www.opensource.org/licenses/index.html">OSI</A> approved license)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               FREE
+                       </TD>
+                       <TD>
+                               ANY (source)
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               LPAC v1.31 (codec 3.0)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               YES (Winamp only)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no?
+                       </TD>
+                       <TD>
+                               YES
+                       </TD>
+                       <TD>
+                               FREE
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               Windows/Linux/Solaris console
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               WavPack v3.91
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               YES (Winamp only)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               YES
+                       </TD>
+                       <TD>
+                               FREE
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               Windows only
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               optimFROG 4.21
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               YES (Winamp, XMMS)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               YES
+                       </TD>
+                       <TD>
+                               FREE
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               Windows/Linux
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               RKAU v1.07
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               YES (Winamp only)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               YES
+                       </TD>
+                       <TD>
+                               FREE
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               Windows only
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               WaveZIP v2
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD>
+                               FREE (24-bit costs $)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               Windows only
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" BGCOLOR="#F4F4CC">
+                               Pegasus-SPS
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               no
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               $39 (free trial)
+                       </TD>
+                       <TD BGCOLOR="#D3D4C5">
+                               Windows only
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+       <P>
+               The machine I used for encoding the test files is a PII-333 with 256 megs of RAM, running Windows NT 4.0 SP5.  Unfortunately, Windows is the lowest-common-denominator platform for all the encoders.
+       </P>
+       <P>
+               The input corpus currently consists entirely of CD music tracks.  In the future it may include more kinds of input (like speech, other sample rates/resolutions, etc).  There are 14 tracks whose genres range from rock to pop to death metal to classical to chant.
+       </P>
+       <P>
+               The first table is a summary of results on all input tracks.  The remaining tables show the results of the encoders on each track.  The summary table has more modes, whereas the individual tables have just the interesting ones.
+       </P>
+       <P>
+               In the summary table, entries are sorted by average compression ratio, which is the average of the ratios for each track; this keeps long tracks from having more influence than short ones.  In the individual tables, this is the same as the straight compression ratio, which is compressed size / uncompressed size.
+       </P>
+       <P>
+               Some interesting things to note:
+               <UL>
+                       <LI>flac -5 is right in the middle with respect to compression, relatively fast on the encoding range, and one of the fastest decoding.  This is about what you would expect; FLAC is designed to put most of the processing on the encoding side, which is only done once, whereas the adaptive codecs take as long to decode as encode.  FLAC is more suited in this way for playback on low-power devices and is one of the reasons it is the only lossless codec with any kind of hardware support.</LI>
+                       <LI>LPAC quality settings are not too stable with -r (which allows seeking during playback) turned on.</LI>
+                       <LI>RKAU also has a tendency to get bigger in the 'high' mode.</LI>
+                       <LI>Another ironic fact is that the encoders that are patented or cost money turn out to be the worst by most measures.  SPS is so archane and crippled that I gave up trying to put together results for it after one track.</LI>
+               </UL>
+       </P>
+       <P>
+               This is a summary table with just the most 'economic' modes (the ones that give the most compression for the least amount of encode/decode time) for each codec.
+       </P>
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Codec</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Encode time</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Decode time</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Compressed<BR>size</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Overall<BR>compression<BR>ratio</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Average<BR>compression<BR>ratio</B></FONT>
+                       </TD>
+               </TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>26:52.07</TD><TD>28:44.55</TD><TD>386.96 MB</TD><TD>0.4958</TD><TD>0.5119</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>24:19.58</TD><TD>25:37.44</TD><TD>389.04 MB</TD><TD>0.4984</TD><TD>0.5151</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (high)</TD><TD>13:59.07</TD><TD>15:30.69</TD><TD>391.76 MB</TD><TD>0.5019</TD><TD>0.5179</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>16:34.96</TD><TD>17:57.28</TD><TD>394.69 MB</TD><TD>0.5056</TD><TD>0.5223</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (normal)</TD><TD>11:42.34</TD><TD>13:11.29</TD><TD>395.04 MB</TD><TD>0.5061</TD><TD>0.5223</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>RKAU 1.07 (normal)</TD><TD>53:46.74</TD><TD>23:31.10</TD><TD>395.71 MB</TD><TD>0.5070</TD><TD>0.5229</TD></TR>
+               <TR><TD>RKAU 1.07 (fast)</TD><TD>26:35.34</TD><TD>20:13.22</TD><TD>399.25 MB</TD><TD>0.5115</TD><TD>0.5262</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>LPAC 1.40 (-r, medium)</TD><TD>18:52.79</TD><TD>10:43.32</TD><TD>403.52 MB</TD><TD>0.5170</TD><TD>0.5319</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (fast)</TD><TD>9:05.59</TD><TD>10:51.09</TD><TD>401.63 MB</TD><TD>0.5145</TD><TD>0.5327</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>13:26.74</TD><TD>7:19.04</TD><TD>413.43 MB</TD><TD>0.5297</TD><TD>0.5458</TD></TR>
+               <TR><TD>flac 1.0.3 (-3)</TD><TD>10:07.34</TD><TD>7:14.68</TD><TD>419.26 MB</TD><TD>0.5371</TD><TD>0.5543</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Bonk 0.5</TD><TD>36:56.36</TD><TD>27:09.35</TD><TD>418.65 MB</TD><TD>0.5364</TD><TD>0.5543</TD></TR>
+               <TR><TD>WavPack 3.91 (high)</TD><TD>7:15.37</TD><TD>?</TD><TD>418.09 MB</TD><TD>0.5356</TD><TD>0.5556</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>431.08 MB</TD><TD>0.5522</TD><TD>0.5714</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>9:44.48</TD><TD>6:31.74</TD><TD>433.56 MB</TD><TD>0.5555</TD><TD>0.5729</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Kexis 0.2.2</TD><TD>17:49.06</TD><TD>14:53.90</TD><TD>434.33 MB</TD><TD>0.5564</TD><TD>0.5750</TD></TR>
+               <TR><TD>WaveZIP</TD><TD>8:41.72</TD><TD>?</TD><TD>452.95 MB</TD><TD>0.5802</TD><TD>0.5986</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>RIFF WAVE</TD><TD>&nbsp;</TD><TD>&nbsp;</TD><TD>780.56 MB</TD><TD>1.0000</TD><TD>1.0000</TD></TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+       <P>
+               Here are the summary results for all codecs and all modes:
+       </P>
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Codec</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Encode time</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Decode time</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Compressed<BR>size</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Overall<BR>compression<BR>ratio</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Average<BR>compression<BR>ratio</B></FONT>
+                       </TD>
+               </TR>
+               <!-- BGCOLOR="#D3D4C5" -->
+               <TR><TD>optimFROG 4.21 (mode 4 @ 2x)</TD><TD>183:05.29</TD><TD>184:13.42</TD><TD>386.13 MB</TD><TD>0.4947</TD><TD>0.5105</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>338:34.96</TD><TD>339:23.24</TD><TD>386.22 MB</TD><TD>0.4948</TD><TD>0.5105</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 4 @ 4x)</TD><TD>105:15.85</TD><TD>106:36.23</TD><TD>386.21 MB</TD><TD>0.4948</TD><TD>0.5107</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 3 @ 2x)</TD><TD>92:48.79</TD><TD>93:49.75</TD><TD>386.52 MB</TD><TD>0.4952</TD><TD>0.5110</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 3 @ 1x)</TD><TD>161:51.00</TD><TD>162:10.62</TD><TD>386.55 MB</TD><TD>0.4952</TD><TD>0.5110</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 3 @ 4x)</TD><TD>58:18.40</TD><TD>59:30.51</TD><TD>386.71 MB</TD><TD>0.4954</TD><TD>0.5114</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>26:52.07</TD><TD>28:44.55</TD><TD>386.96 MB</TD><TD>0.4958</TD><TD>0.5119</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 2 @ 1x)</TD><TD>68:22.58</TD><TD>69:29.50</TD><TD>387.71 MB</TD><TD>0.4967</TD><TD>0.5128</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 2 @ 2x)</TD><TD>44:17.55</TD><TD>45:31.33</TD><TD>387.72 MB</TD><TD>0.4967</TD><TD>0.5129</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 2 @ 4x)</TD><TD>32:16.85</TD><TD>33:30.92</TD><TD>387.93 MB</TD><TD>0.4970</TD><TD>0.5133</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 1 @ 1x)</TD><TD>43:00.91</TD><TD>44:13.07</TD><TD>388.71 MB</TD><TD>0.4980</TD><TD>0.5146</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 1 @ 2x)</TD><TD>30:35.00</TD><TD>31:50.50</TD><TD>388.81 MB</TD><TD>0.4981</TD><TD>0.5147</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>24:19.58</TD><TD>25:37.44</TD><TD>389.04 MB</TD><TD>0.4984</TD><TD>0.5151</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (high)</TD><TD>13:59.07</TD><TD>15:30.69</TD><TD>391.76 MB</TD><TD>0.5019</TD><TD>0.5179</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 0 @ 1x)</TD><TD>20:51.21</TD><TD>22:08.44</TD><TD>394.35 MB</TD><TD>0.5052</TD><TD>0.5218</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 0 @ 2x)</TD><TD>17:59.86</TD><TD>19:20.53</TD><TD>394.48 MB</TD><TD>0.5054</TD><TD>0.5220</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>16:34.96</TD><TD>17:57.28</TD><TD>394.69 MB</TD><TD>0.5056</TD><TD>0.5223</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (normal)</TD><TD>11:42.34</TD><TD>13:11.29</TD><TD>395.04 MB</TD><TD>0.5061</TD><TD>0.5223</TD></TR>
+               <TR><TD>RKAU 1.07 (normal)</TD><TD>53:46.74</TD><TD>23:31.10</TD><TD>395.71 MB</TD><TD>0.5070</TD><TD>0.5229</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>RKAU 1.07 (high)</TD><TD>136:56.62</TD><TD>27:55.98</TD><TD>395.89 MB</TD><TD>0.5072</TD><TD>0.5235</TD></TR>
+               <TR><TD>RKAU 1.07 (fast)</TD><TD>26:35.34</TD><TD>20:13.22</TD><TD>399.25 MB</TD><TD>0.5115</TD><TD>0.5262</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>LPAC 1.40 (-r, medium)</TD><TD>18:52.79</TD><TD>10:43.32</TD><TD>403.52 MB</TD><TD>0.5170</TD><TD>0.5319</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, extra high)</TD><TD>30:30.93</TD><TD>12:20.26</TD><TD>404.08 MB</TD><TD>0.5177</TD><TD>0.5322</TR>
+               <TR BGCOLOR="#D3D4C5"><TD>LPAC 1.40 (-r, high)</TD><TD>24:56.56</TD><TD>11:51.64</TD><TD>404.03 MB</TD><TD>0.5176</TD><TD>0.5323</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (fast)</TD><TD>9:05.59</TD><TD>10:51.09</TD><TD>401.63 MB</TD><TD>0.5145</TD><TD>0.5327</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-8)</TD><TD>55:49.51</TD><TD>7:25.36</TD><TD>411.85 MB</TD><TD>0.5276</TD><TD>0.5436</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>13:26.74</TD><TD>7:19.04</TD><TD>413.43 MB</TD><TD>0.5297</TD><TD>0.5458</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-3)</TD><TD>10:07.34</TD><TD>7:14.68</TD><TD>419.26 MB</TD><TD>0.5371</TD><TD>0.5543</TD></TR>
+               <TR><TD>Bonk 0.5</TD><TD>36:56.36</TD><TD>27:09.35</TD><TD>418.65 MB</TD><TD>0.5364</TD><TD>0.5543</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WavPack 3.91 (high)</TD><TD>7:15.37</TD><TD>?</TD><TD>418.09 MB</TD><TD>0.5356</TD><TD>0.5556</TD></TR>
+               <TR><TD>flac 1.0.3 (-1)</TD><TD>8:55.94</TD><TD>7:22.51</TD><TD>432.29 MB</TD><TD>0.5538</TD><TD>0.5704</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>431.08 MB</TD><TD>0.5522</TD><TD>0.5714</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>9:44.48</TD><TD>6:31.74</TD><TD>433.56 MB</TD><TD>0.5555</TD><TD>0.5729</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Kexis 0.2.2</TD><TD>17:49.06</TD><TD>14:53.90</TD><TD>434.33 MB</TD><TD>0.5564</TD><TD>0.5750</TD></TR>
+               <TR><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>12:00.04</TD><TD>7:25.12</TD><TD>438.86 MB</TD><TD>0.5622</TD><TD>0.5810</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>8:41.72</TD><TD>?</TD><TD>452.95 MB</TD><TD>0.5802</TD><TD>0.5986</TD></TR>
+               <TR><TD>RIFF WAVE</TD><TD>&nbsp;</TD><TD>&nbsp;</TD><TD>780.56 MB</TD><TD>1.0000</TD><TD>1.0000</TD></TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+       <P>
+               Here are the results for each individual track.
+       </P>
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD ALIGN="RIGHT">
+                               <FONT SIZE="+1"><B>Track</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Codec</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Encode<BR>time</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Decode<BR>time</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Compressed<BR>size</B></FONT>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1"><B>Compression<BR>ratio</B></FONT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="22" BGCOLOR="#F4F4CC">
+                               Dream Theater<BR><I>6:00</I><BR>58.47 MB
+                       </TD>
+               </TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>2:05.36</TD><TD>2:13.44</TD><TD>43.24 MB</TD><TD>0.7395</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>1:53.28</TD><TD>2:00.45</TD><TD>43.26 MB</TD><TD>0.7398</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>25:32.41</TD><TD>25:38.15</TD><TD>43.26 MB</TD><TD>0.7398</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (high)</TD><TD>1:07.92</TD><TD>1:13.20</TD><TD>43.39 MB</TD><TD>0.7421</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>1:17.95</TD><TD>1:25.67</TD><TD>43.42 MB</TD><TD>0.7426</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:57.04</TD><TD>1:03.37</TD><TD>43.48 MB</TD><TD>0.7436</TD></TR>
+               <TR><TD>RKAU 1.07 (normal)</TD><TD>1:57.68</TD><TD>1:33.38</TD><TD>43.81 MB</TD><TD>0.7493</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:44.33</TD><TD>0:51.22</TD><TD>43.97 MB</TD><TD>0.7520</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>1:27.61</TD><TD>0:56.18</TD><TD>44.12 MB</TD><TD>0.7545</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-8)</TD><TD>4:20.41</TD><TD>0:37.86</TD><TD>44.33 MB</TD><TD>0.7581</TD></TR>
+               <TR><TD>Bonk 0.5</TD><TD>2:56.03</TD><TD>2:11.58</TD><TD>44.35 MB</TD><TD>0.7585</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>1:03.87</TD><TD>0:37.55</TD><TD>44.40 MB</TD><TD>0.7594</TD></TR>
+               <TR><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>0:58.81</TD><TD>0:37.63</TD><TD>44.75 MB</TD><TD>0.7654</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-3)</TD><TD>0:48.52</TD><TD>0:37.22</TD><TD>44.78 MB</TD><TD>0.7658</TD></TR>
+               <TR><TD>WavPack 3.91 (high)</TD><TD>0:35.86</TD><TD>?</TD><TD>45.14 MB</TD><TD>0.7720</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>45.17 MB</TD><TD>0.7725</TD></TR>
+               <TR><TD>Pegasus-SPS</TD><TD>4:45.00</TD><TD>?</TD><TD>45.40 MB</TD><TD>0.7765</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Kexis 0.2.2</TD><TD>1:24.83</TD><TD>1:10.93</TD><TD>46.52 MB</TD><TD>0.7956</TD></TR>
+               <TR><TD>flac 1.0.3 (-1)</TD><TD>0:44.07</TD><TD>0:37.77</TD><TD>46.64 MB</TD><TD>0.7977</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:47.75</TD><TD>0:32.56</TD><TD>46.68 MB</TD><TD>0.7984</TD></TR>
+               <TR><TD>WaveZIP</TD><TD>0:38.99</TD><TD>?</TD><TD>47.22 MB</TD><TD>0.8077</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               Eddie Warner<BR><I>Titus</I><BR>27.87 MB
+                       </TD>
+               </TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>0:40.76</TD><TD>0:21.21</TD><TD>14.77 MB</TD><TD>0.5298</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-8)</TD><TD>1:58.90</TD><TD>0:17.33</TD><TD>15.01 MB</TD><TD>0.5384</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>0:53.39</TD><TD>0:55.52</TD><TD>15.01 MB</TD><TD>0.5385</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>12:02.54</TD><TD>12:03.76</TD><TD>15.02 MB</TD><TD>0.5390</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>0:29.56</TD><TD>0:15.57</TD><TD>15.11 MB</TD><TD>0.5423</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>0:36.81</TD><TD>0:39.19</TD><TD>15.13 MB</TD><TD>0.5429</TD></TR>
+               <TR><TD>RKAU 1.07 (normal)</TD><TD>0:54.82</TD><TD>0:42.71</TD><TD>15.15 MB</TD><TD>0.5435</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (extra high)</TD><TD>0:58.52</TD><TD>1:01.81</TD><TD>15.25 MB</TD><TD>0.5471</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (high)</TD><TD>0:30.88</TD><TD>0:33.55</TD><TD>15.34 MB</TD><TD>0.5505</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:25.45</TD><TD>0:28.37</TD><TD>15.35 MB</TD><TD>0.5509</TD></TR>
+               <TR><TD>flac 1.0.3 (-3)</TD><TD>0:22.29</TD><TD>0:15.29</TD><TD>15.43 MB</TD><TD>0.5537</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:19.85</TD><TD>0:22.90</TD><TD>15.58 MB</TD><TD>0.5592</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:21.16</TD><TD>0:13.55</TD><TD>15.78 MB</TD><TD>0.5662</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>0:26.82</TD><TD>0:16.75</TD><TD>16.21 MB</TD><TD>0.5818</TD></TR>
+               <TR><TD>flac 1.0.3 (-1)</TD><TD>0:19.92</TD><TD>0:15.67</TD><TD>16.38 MB</TD><TD>0.5879</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Bonk 0.5</TD><TD>1:22.01</TD><TD>1:00.12</TD><TD>16.73 MB</TD><TD>0.6003</TD></TR>
+               <TR><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>17.03 MB</TD><TD>0.6112</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WavPack 3.91 (high)</TD><TD>0:10.55</TD><TD>?</TD><TD>17.13 MB</TD><TD>0.6148</TD></TR>
+               <TR><TD>Kexis 0.2.2</TD><TD>0:38.72</TD><TD>0:32.25</TD><TD>17.40 MB</TD><TD>0.6242</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>0:17.55</TD><TD>?</TD><TD>17.89 MB</TD><TD>0.6420</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               Tool<BR><I>Forty-six & 2</I><BR>64.25 MB
+                       </TD>
+               </TR>
+               <TR><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>27:58.28</TD><TD>28:01.87</TD><TD>37.96 MB</TD><TD>0.5907</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>2:03.43</TD><TD>2:09.27</TD><TD>38.15 MB</TD><TD>0.5937</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>2:14.70</TD><TD>2:24.30</TD><TD>38.23 MB</TD><TD>0.5950</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (high)</TD><TD>1:09.82</TD><TD>1:18.09</TD><TD>38.42 MB</TD><TD>0.5979</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:58.69</TD><TD>1:07.02</TD><TD>38.59 MB</TD><TD>0.6005</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>1:24.44</TD><TD>1:30.97</TD><TD>38.68 MB</TD><TD>0.6020</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:46.50</TD><TD>0:55.41</TD><TD>39.18 MB</TD><TD>0.6098</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>RKAU 1.07 (normal)</TD><TD>2:16.00</TD><TD>1:41.84</TD><TD>39.42 MB</TD><TD>0.6135</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>1:38.01</TD><TD>0:57.56</TD><TD>40.25 MB</TD><TD>0.6263</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-8)</TD><TD>4:39.65</TD><TD>0:38.25</TD><TD>40.88 MB</TD><TD>0.6363</TD></TR>
+               <TR><TD>Bonk 0.5</TD><TD>3:07.20</TD><TD>2:21.28</TD><TD>40.98 MB</TD><TD>0.6378</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>1:08.27</TD><TD>0:38.81</TD><TD>41.04 MB</TD><TD>0.6387</TD></TR>
+               <TR><TD>WavPack 3.91 (high)</TD><TD>0:37.51</TD><TD>?</TD><TD>41.51 MB</TD><TD>0.6461</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-3)</TD><TD>0:51.78</TD><TD>0:38.54</TD><TD>41.74 MB</TD><TD>0.6496</TD></TR>
+               <TR><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>42.27 MB</TD><TD>0.6578</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-1)</TD><TD>0:46.19</TD><TD>0:41.18</TD><TD>42.70 MB</TD><TD>0.6646</TD></TR>
+               <TR><TD>Kexis 0.2.2</TD><TD>1:30.09</TD><TD>1:16.29</TD><TD>42.75 MB</TD><TD>0.6652</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>1:02.42</TD><TD>0:37.84</TD><TD>43.06 MB</TD><TD>0.6701</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:51.29</TD><TD>0:34.59</TD><TD>43.18 MB</TD><TD>0.6721</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>0:42.84</TD><TD>?</TD><TD>44.52 MB</TD><TD>0.6930</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               Cannibal Corpse<BR><I>Mummified In Barbed Wire</I><BR>33.37 MB
+                       </TD>
+               </TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>1:10.94</TD><TD>1:15.92</TD><TD>22.95 MB</TD><TD>0.6876</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>14:34.28</TD><TD>14:37.69</TD><TD>22.95 MB</TD><TD>0.6877</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (high)</TD><TD>0:37.63</TD><TD>0:41.34</TD><TD>23.19 MB</TD><TD>0.6948</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:31.71</TD><TD>0:34.87</TD><TD>23.26 MB</TD><TD>0.6968</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>1:03.96</TD><TD>1:08.85</TD><TD>23.31 MB</TD><TD>0.6984</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>RKAU 1.07 (normal)</TD><TD>1:09.71</TD><TD>0:56.66</TD><TD>23.34 MB</TD><TD>0.6993</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>1:05.38</TD><TD>0:36.20</TD><TD>23.53 MB</TD><TD>0.7050</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>0:44.14</TD><TD>0:48.71</TD><TD>23.95 MB</TD><TD>0.7176</TD></TR>
+               <TR><TD>flac 1.0.3 (-8)</TD><TD>2:27.82</TD><TD>0:22.15</TD><TD>24.18 MB</TD><TD>0.7244</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:25.05</TD><TD>0:28.99</TD><TD>24.20 MB</TD><TD>0.7250</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>0:36.19</TD><TD>0:21.00</TD><TD>24.30 MB</TD><TD>0.7281</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Bonk 0.5</TD><TD>1:40.38</TD><TD>1:14.58</TD><TD>24.36 MB</TD><TD>0.7297</TD></TR>
+               <TR><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>0:33.74</TD><TD>0:22.47</TD><TD>25.12 MB</TD><TD>0.7526</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-3)</TD><TD>0:28.29</TD><TD>0:20.82</TD><TD>25.16 MB</TD><TD>0.7539</TD></TR>
+               <TR><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>25.23 MB</TD><TD>0.7558</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WavPack 3.91 (high)</TD><TD>0:20.50</TD><TD>?</TD><TD>25.33 MB</TD><TD>0.7589</TD></TR>
+               <TR><TD>Kexis 0.2.2</TD><TD>0:47.13</TD><TD>0:40.67</TD><TD>26.03 MB</TD><TD>0.7799</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-1)</TD><TD>0:24.95</TD><TD>0:22.70</TD><TD>26.09 MB</TD><TD>0.7818</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:28.20</TD><TD>0:20.46</TD><TD>26.61 MB</TD><TD>0.7972</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>0:22.25</TD><TD>?</TD><TD>26.89 MB</TD><TD>0.8058</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               Alanis Morisette<BR><I>Hand In My Pocket</I><BR>39.09 MB
+                       </TD>
+               </TR>
+               <TR><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>16:51.82</TD><TD>16:54.34</TD><TD>21.24 MB</TD><TD>0.5433</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>1:14.29</TD><TD>1:18.06</TD><TD>21.36 MB</TD><TD>0.5464</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>1:21.38</TD><TD>1:27.28</TD><TD>21.54 MB</TD><TD>0.5509</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (high)</TD><TD>0:42.54</TD><TD>0:47.41</TD><TD>21.75 MB</TD><TD>0.5563</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:35.45</TD><TD>0:39.65</TD><TD>21.84 MB</TD><TD>0.5586</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>0:51.39</TD><TD>0:54.97</TD><TD>21.89 MB</TD><TD>0.5598</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:28.23</TD><TD>0:33.21</TD><TD>22.16 MB</TD><TD>0.5668</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>RKAU 1.07 (normal)</TD><TD>1:21.18</TD><TD>1:01.60</TD><TD>22.80 MB</TD><TD>0.5833</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>1:01.11</TD><TD>0:33.79</TD><TD>23.25 MB</TD><TD>0.5948</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Bonk 0.5</TD><TD>1:53.41</TD><TD>1:23.52</TD><TD>23.35 MB</TD><TD>0.5972</TD></TR>
+               <TR><TD>flac 1.0.3 (-8)</TD><TD>2:48.45</TD><TD>0:23.47</TD><TD>23.45 MB</TD><TD>0.5997</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>0:41.00</TD><TD>0:23.59</TD><TD>23.55 MB</TD><TD>0.6025</TD></TR>
+               <TR><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>24.11 MB</TD><TD>0.6167</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WavPack 3.91 (high)</TD><TD>0:22.50</TD><TD>?</TD><TD>24.22 MB</TD><TD>0.6196</TD></TR>
+               <TR><TD>flac 1.0.3 (-3)</TD><TD>0:31.65</TD><TD>0:22.27</TD><TD>24.32 MB</TD><TD>0.6220</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>0:37.49</TD><TD>0:22.93</TD><TD>24.72 MB</TD><TD>0.6323</TD></TR>
+               <TR><TD>Kexis 0.2.2</TD><TD>0:54.26</TD><TD>0:45.64</TD><TD>24.80 MB</TD><TD>0.6345</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-1)</TD><TD>0:27.51</TD><TD>0:22.72</TD><TD>24.81 MB</TD><TD>0.6347</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:29.71</TD><TD>0:18.92</TD><TD>25.34 MB</TD><TD>0.6481</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>0:28.05</TD><TD>?</TD><TD>25.95 MB</TD><TD>0.6638</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               Gloria Estefan<BR><I>Conga</I><BR>45.15 MB
+                       </TD>
+               </TR>
+               <TR><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>19:40.53</TD><TD>19:44.47</TD><TD>29.43 MB</TD><TD>0.6517</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>1:26.64</TD><TD>1:32.23</TD><TD>29.58 MB</TD><TD>0.6550</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>1:35.65</TD><TD>1:42.11</TD><TD>29.65 MB</TD><TD>0.6567</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>0:59.59</TD><TD>1:05.29</TD><TD>29.78 MB</TD><TD>0.6595</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (high)</TD><TD>0:50.17</TD><TD>0:56.40</TD><TD>29.85 MB</TD><TD>0.6610</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:42.27</TD><TD>0:47.74</TD><TD>29.97 MB</TD><TD>0.6637</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:33.46</TD><TD>0:39.22</TD><TD>30.30 MB</TD><TD>0.6710</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>RKAU 1.07 (normal)</TD><TD>1:37.85</TD><TD>1:12.15</TD><TD>30.34 MB</TD><TD>0.6719</TD></TR>
+               <TR><TD>Bonk 0.5</TD><TD>2:13.34</TD><TD>1:39.44</TD><TD>30.64 MB</TD><TD>0.6785</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-8)</TD><TD>3:18.77</TD><TD>0:28.57</TD><TD>30.75 MB</TD><TD>0.6811</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>1:14.08</TD><TD>0:44.64</TD><TD>30.81 MB</TD><TD>0.6823</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>0:48.11</TD><TD>0:27.84</TD><TD>30.85 MB</TD><TD>0.6833</TD></TR>
+               <TR><TD>WavPack 3.91 (high)</TD><TD>0:26.73</TD><TD>?</TD><TD>30.91 MB</TD><TD>0.6845</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>31.06 MB</TD><TD>0.6879</TD></TR>
+               <TR><TD>flac 1.0.3 (-3)</TD><TD>0:37.46</TD><TD>0:27.80</TD><TD>31.63 MB</TD><TD>0.7005</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-1)</TD><TD>0:32.55</TD><TD>0:27.42</TD><TD>31.99 MB</TD><TD>0.7085</TD></TR>
+               <TR><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>0:44.76</TD><TD>0:27.48</TD><TD>31.76 MB</TD><TD>0.7034</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Kexis 0.2.2</TD><TD>1:03.91</TD><TD>0:53.54</TD><TD>31.86 MB</TD><TD>0.7056</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:35.74</TD><TD>0:23.64</TD><TD>32.47 MB</TD><TD>0.7191</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>0:29.42</TD><TD>?</TD><TD>33.02 MB</TD><TD>0.7313</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               Cream<BR><I>White Room</I><BR>53.01 MB
+                       </TD>
+               </TR>
+               <TR><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>22:59.05</TD><TD>23:02.98</TD><TD>33.93 MB</TD><TD>0.6399</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>1:42.28</TD><TD>1:47.55</TD><TD>33.96 MB</TD><TD>0.6405</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>1:51.77</TD><TD>2:00.37</TD><TD>34.14 MB</TD><TD>0.6441</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (high)</TD><TD>0:58.45</TD><TD>1:04.59</TD><TD>34.29 MB</TD><TD>0.6468</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>1:10.30</TD><TD>1:15.99</TD><TD>34.29 MB</TD><TD>0.6468</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:49.32</TD><TD>0:56.89</TD><TD>34.42 MB</TD><TD>0.6493</TD></TR>
+               <TR><TD>RKAU 1.07 (normal)</TD><TD>1:50.80</TD><TD>1:24.98</TD><TD>34.60 MB</TD><TD>0.6527</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>LPAC 1.40 (-r, normal)</TD><TD>1:25.16</TD><TD>0:48.67</TD><TD>34.84 MB</TD><TD>0.6572</TD></TR>
+               <TR><TD>Bonk 0.5</TD><TD>2:35.36</TD><TD>1:56.20</TD><TD>34.96 MB</TD><TD>0.6595</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:38.75</TD><TD>0:46.80</TD><TD>34.99 MB</TD><TD>0.6601</TD></TR>
+               <TR><TD>flac 1.0.3 (-8)</TD><TD>3:53.21</TD><TD>0:32.08</TD><TD>35.00 MB</TD><TD>0.6602</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>0:57.28</TD><TD>0:32.27</TD><TD>35.17 MB</TD><TD>0.6634</TD></TR>
+               <TR><TD>flac 1.0.3 (-3)</TD><TD>0:42.71</TD><TD>0:32.56</TD><TD>35.37 MB</TD><TD>0.6672</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>0:51.44</TD><TD>0:33.02</TD><TD>35.40 MB</TD><TD>0.6677</TD></TR>
+               <TR><TD>WavPack 3.91 (high)</TD><TD>0:31.05</TD><TD>?</TD><TD>35.60 MB</TD><TD>0.6715</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>35.74 MB</TD><TD>0.6742</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:41.14</TD><TD>0:28.91</TD><TD>36.42 MB</TD><TD>0.6870</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-1)</TD><TD>0:38.47</TD><TD>0:30.66</TD><TD>36.56 MB</TD><TD>0.6896</TD></TR>
+               <TR><TD>Kexis 0.2.2</TD><TD>1:13.87</TD><TD>1:02.90</TD><TD>36.64 MB</TD><TD>0.6911</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>0:35.77</TD><TD>?</TD><TD>37.13 MB</TD><TD>0.7004</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               Maurice Ravel<BR><I>Fanfare from "L'eventail de Jeanne"</I><BR>20.82 MB
+                       </TD>
+               </TR>
+               <TR><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>8:22.42</TD><TD>8:23.32</TD><TD>6.82 MB</TD><TD>0.3274</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (extra high)</TD><TD>0:39.93</TD><TD>0:41.69</TD><TD>6.85 MB</TD><TD>0.3289</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>0:36.57</TD><TD>0:38.26</TD><TD>7.09 MB</TD><TD>0.3406</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (high)</TD><TD>0:21.22</TD><TD>0:23.04</TD><TD>7.16 MB</TD><TD>0.3437</TD></TR>
+               <TR><TD>RKAU 1.07 (normal)</TD><TD>0:40.67</TD><TD>0:28.52</TD><TD>7.18 MB</TD><TD>0.3451</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>0:24.98</TD><TD>0:26.37</TD><TD>7.21 MB</TD><TD>0.3462</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>0:29.01</TD><TD>0:15.11</TD><TD>7.33 MB</TD><TD>0.3520</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:18.19</TD><TD>0:19.54</TD><TD>7.44 MB</TD><TD>0.3575</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:13.70</TD><TD>0:15.72</TD><TD>7.64 MB</TD><TD>0.3671</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-8)</TD><TD>1:21.11</TD><TD>0:09.72</TD><TD>7.68 MB</TD><TD>0.3691</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>0:19.97</TD><TD>0:09.77</TD><TD>7.71 MB</TD><TD>0.3702</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-3)</TD><TD>0:15.09</TD><TD>0:10.09</TD><TD>7.77 MB</TD><TD>0.3733</TD></TR>
+               <TR><TD>Bonk 0.5</TD><TD>0:55.92</TD><TD>0:40.23</TD><TD>7.83 MB</TD><TD>0.3762</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WavPack 3.91 (high)</TD><TD>0:11.42</TD><TD>?</TD><TD>7.89 MB</TD><TD>0.3791</TD></TR>
+               <TR><TD>flac 1.0.3 (-1)</TD><TD>0:13.18</TD><TD>0:09.78</TD><TD>8.12 MB</TD><TD>0.3901</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>8.15 MB</TD><TD>0.3914</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:13.81</TD><TD>0:08.88</TD><TD>8.19 MB</TD><TD>0.3932</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>0:17.45</TD><TD>0:10.30</TD><TD>8.29 MB</TD><TD>0.3983</TD></TR>
+               <TR><TD>Kexis 0.2.2</TD><TD>0:26.78</TD><TD>0:21.90</TD><TD>8.52 MB</TD><TD>0.4091</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>0:13.11</TD><TD>?</TD><TD>8.72 MB</TD><TD>0.4193</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               Maurice Ravel<BR><I>String Quartet (4th movement)</I><BR>56.18 MB
+                       </TD>
+               </TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>1:54.09</TD><TD>2:01.72</TD><TD>20.47 MB</TD><TD>0.3642</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>24:26.99</TD><TD>24:29.36</TD><TD>20.62 MB</TD><TD>0.3671</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (high)</TD><TD>0:58.14</TD><TD>1:06.45</TD><TD>20.80 MB</TD><TD>0.3702</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>1:42.57</TD><TD>1:47.48</TD><TD>20.93 MB</TD><TD>0.3725</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:48.61</TD><TD>0:54.73</TD><TD>21.14 MB</TD><TD>0.3763</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>1:09.17</TD><TD>1:13.14</TD><TD>21.23 MB</TD><TD>0.3779</TD></TR>
+               <TR><TD>RKAU 1.07 (normal)</TD><TD>1:52.65</TD><TD>1:25.39</TD><TD>21.30 MB</TD><TD>0.3791</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:37.30</TD><TD>0:44.79</TD><TD>21.54 MB</TD><TD>0.3835</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>1:20.84</TD><TD>0:42.73</TD><TD>21.96 MB</TD><TD>0.3909</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WavPack 3.91 (high)</TD><TD>0:30.03</TD><TD>?</TD><TD>22.30 MB</TD><TD>0.3969</TD></TR>
+               <TR><TD>flac 1.0.3 (-8)</TD><TD>3:57.24</TD><TD>0:28.26</TD><TD>22.61 MB</TD><TD>0.4024</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>0:56.47</TD><TD>0:28.31</TD><TD>22.67 MB</TD><TD>0.4036</TD></TR>
+               <TR><TD>Bonk 0.5</TD><TD>2:33.53</TD><TD>1:51.94</TD><TD>23.18 MB</TD><TD>0.4125</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-3)</TD><TD>0:41.66</TD><TD>0:27.63</TD><TD>23.21 MB</TD><TD>0.4131</TD></TR>
+               <TR><TD>flac 1.0.3 (-1)</TD><TD>0:36.76</TD><TD>0:29.54</TD><TD>23.36 MB</TD><TD>0.4158</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Kexis 0.2.2</TD><TD>1:15.05</TD><TD>1:03.86</TD><TD>23.42 MB</TD><TD>0.4168</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:39.96</TD><TD>0:27.36</TD><TD>23.71 MB</TD><TD>0.4221</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>24.12 MB</TD><TD>0.4293</TD></TR>
+               <TR><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>0:49.06</TD><TD>0:29.94</TD><TD>25.59 MB</TD><TD>0.4554</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>0:36.60</TD><TD>?</TD><TD>25.84 MB</TD><TD>0.4600</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               Sergei Prokofiev<BR><I>Piano Concerto No.3 (3rd movement)</I><BR>100.68 MB
+                       </TD>
+               </TR>
+               <TR><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>43:21.88</TD><TD>43:26.18</TD><TD>33.58 MB</TD><TD>0.3335</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (extra high)</TD><TD>3:21.33</TD><TD>3:35.91</TD><TD>33.72 MB</TD><TD>0.3349</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>3:00.57</TD><TD>3:08.19</TD><TD>33.83 MB</TD><TD>0.3360</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>2:00.85</TD><TD>2:09.52</TD><TD>34.14 MB</TD><TD>0.3390</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (high)</TD><TD>1:43.17</TD><TD>1:55.31</TD><TD>34.23 MB</TD><TD>0.3400</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (normal)</TD><TD>1:26.19</TD><TD>1:35.90</TD><TD>34.66 MB</TD><TD>0.3442</TD></TR>
+               <TR><TD>RKAU 1.07 (normal)</TD><TD>3:08.70</TD><TD>2:26.17</TD><TD>35.21 MB</TD><TD>0.3496</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>LPAC 1.40 (-r, normal)</TD><TD>2:06.21</TD><TD>1:11.92</TD><TD>35.27 MB</TD><TD>0.3502</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (fast)</TD><TD>1:06.28</TD><TD>1:18.56</TD><TD>35.43 MB</TD><TD>0.3518</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WavPack 3.91 (high)</TD><TD>0:53.49</TD><TD>?</TD><TD>37.44 MB</TD><TD>0.3718</TD></TR>
+               <TR><TD>flac 1.0.3 (-8)</TD><TD>6:57.76</TD><TD>0:51.45</TD><TD>38.07 MB</TD><TD>0.3781</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>1:39.78</TD><TD>0:49.47</TD><TD>38.17 MB</TD><TD>0.3791</TD></TR>
+               <TR><TD>flac 1.0.3 (-3)</TD><TD>1:13.49</TD><TD>0:49.51</TD><TD>38.50 MB</TD><TD>0.3824</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-1)</TD><TD>1:04.89</TD><TD>0:53.25</TD><TD>39.30 MB</TD><TD>0.3903</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>1:10.57</TD><TD>0:50.00</TD><TD>39.49 MB</TD><TD>0.3921</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Kexis 0.2.2</TD><TD>2:12.39</TD><TD>1:49.00</TD><TD>39.89 MB</TD><TD>0.3962</TD></TR>
+               <TR><TD>Bonk 0.5</TD><TD>4:33.71</TD><TD>3:19.38</TD><TD>40.31 MB</TD><TD>0.4003</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>41.86 MB</TD><TD>0.4157</TD></TR>
+               <TR><TD>WaveZIP</TD><TD>1:05.60</TD><TD>?</TD><TD>43.67 MB</TD><TD>0.4337</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>1:26.84</TD><TD>0:53.19</TD><TD>45.34 MB</TD><TD>0.4502</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               Frederic Chopin<BR><I>Prelude No.24 in d minor</I><BR>27.46 MB
+                       </TD>
+               </TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>0:55.83</TD><TD>0:59.52</TD><TD>10.25 MB</TD><TD>0.3734</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>11:56.24</TD><TD>11:58.77</TD><TD>10.34 MB</TD><TD>0.3764</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>0:50.07</TD><TD>0:53.49</TD><TD>10.41 MB</TD><TD>0.3790</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (high)</TD><TD>0:28.52</TD><TD>0:31.32</TD><TD>10.47 MB</TD><TD>0.3812</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>0:34.06</TD><TD>0:38.70</TD><TD>10.53 MB</TD><TD>0.3833</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:23.70</TD><TD>0:27.61</TD><TD>10.59 MB</TD><TD>0.3854</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>0:34.91</TD><TD>0:20.01</TD><TD>10.74 MB</TD><TD>0.3911</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>RKAU 1.07 (normal)</TD><TD>0:54.46</TD><TD>0:41.54</TD><TD>10.88 MB</TD><TD>0.3963</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:18.53</TD><TD>0:21.80</TD><TD>10.94 MB</TD><TD>0.3982</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WavPack 3.91 (high)</TD><TD>0:14.46</TD><TD>?</TD><TD>11.38 MB</TD><TD>0.4145</TD></TR>
+               <TR><TD>flac 1.0.3 (-8)</TD><TD>1:54.97</TD><TD>0:13.68</TD><TD>11.69 MB</TD><TD>0.4255</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>0:27.35</TD><TD>0:14.06</TD><TD>11.71 MB</TD><TD>0.4264</TD></TR>
+               <TR><TD>flac 1.0.3 (-3)</TD><TD>0:20.43</TD><TD>0:14.59</TD><TD>11.73 MB</TD><TD>0.4273</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-1)</TD><TD>0:17.24</TD><TD>0:14.17</TD><TD>11.86 MB</TD><TD>0.4318</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:19.20</TD><TD>0:11.89</TD><TD>12.05 MB</TD><TD>0.4386</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Kexis 0.2.2</TD><TD>0:36.70</TD><TD>0:30.26</TD><TD>12.14 MB</TD><TD>0.4419</TD></TR>
+               <TR><TD>Bonk 0.5</TD><TD>1:15.45</TD><TD>0:55.07</TD><TD>12.86 MB</TD><TD>0.4684</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>0:18.75</TD><TD>?</TD><TD>13.08 MB</TD><TD>0.4765</TD></TR>
+               <TR><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>13.31 MB</TD><TD>0.4845</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>0:24.43</TD><TD>0:14.32</TD><TD>14.40 MB</TD><TD>0.5242</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               Domenico Scarlatti<BR><I>Sonata K.42 (arr.Yepes for guitar)</I><BR>16.39 MB
+                       </TD>
+               </TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>0:33.99</TD><TD>0:35.53</TD><TD>6.80 MB</TD><TD>0.4145</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>7:08.77</TD><TD>7:09.52</TD><TD>6.87 MB</TD><TD>0.4190</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (high)</TD><TD>0:17.47</TD><TD>0:19.11</TD><TD>6.94 MB</TD><TD>0.4230</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>0:30.47</TD><TD>0:31.95</TD><TD>6.98 MB</TD><TD>0.4255</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:14.84</TD><TD>0:16.17</TD><TD>7.02 MB</TD><TD>0.4284</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>RKAU 1.07 (normal)</TD><TD>0:37.09</TD><TD>0:26.56</TD><TD>7.05 MB</TD><TD>0.4297</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>0:20.68</TD><TD>0:22.06</TD><TD>7.07 MB</TD><TD>0.4310</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:11.39</TD><TD>0:12.95</TD><TD>7.19 MB</TD><TD>0.4384</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>0:29.99</TD><TD>0:14.33</TD><TD>7.21 MB</TD><TD>0.4397</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WavPack 3.91 (high)</TD><TD>0:09.23</TD><TD>?</TD><TD>7.35 MB</TD><TD>0.4481</TD></TR>
+               <TR><TD>flac 1.0.3 (-8)</TD><TD>1:09.87</TD><TD>0:08.22</TD><TD>7.37 MB</TD><TD>0.4497</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>0:16.58</TD><TD>0:08.13</TD><TD>7.40 MB</TD><TD>0.4512</TD></TR>
+               <TR><TD>flac 1.0.3 (-3)</TD><TD>0:12.15</TD><TD>0:08.47</TD><TD>7.43 MB</TD><TD>0.4529</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Bonk 0.5</TD><TD>0:45.59</TD><TD>0:32.63</TD><TD>7.46 MB</TD><TD>0.4548</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:11.67</TD><TD>0:07.17</TD><TD>7.48 MB</TD><TD>0.4564</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Kexis 0.2.2</TD><TD>0:22.06</TD><TD>0:18.62</TD><TD>7.50 MB</TD><TD>0.4572</TD></TR>
+               <TR><TD>flac 1.0.3 (-1)</TD><TD>0:10.74</TD><TD>0:08.20</TD><TD>7.52 MB</TD><TD>0.4590</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>7.74 MB</TD><TD>0.4723</TD></TR>
+               <TR><TD>WaveZIP</TD><TD>0:10.56</TD><TD>?</TD><TD>7.83 MB</TD><TD>0.4781</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>0:14.29</TD><TD>0:09.07</TD><TD>8.20 MB</TD><TD>0.5004</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               The Benedictine Monks of<BR>Santo Domingo de Silos<BR><I>Laetatus sum</I><BR>24.26 MB
+                       </TD>
+               </TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>0:50.72</TD><TD>0:54.20</TD><TD>12.15 MB</TD><TD>0.5006</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>10:34.75</TD><TD>10:36.83</TD><TD>12.17 MB</TD><TD>0.5015</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (high)</TD><TD>0:26.36</TD><TD>0:28.52</TD><TD>12.25 MB</TD><TD>0.5048</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>RKAU 1.07 (normal)</TD><TD>0:57.64</TD><TD>0:41.39</TD><TD>12.25 MB</TD><TD>0.5049</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>0:45.20</TD><TD>0:48.45</TD><TD>12.43 MB</TD><TD>0.5121</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (normal)</TD><TD>0:21.81</TD><TD>0:24.33</TD><TD>12.47 MB</TD><TD>0.5139</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>0:45.66</TD><TD>0:23.29</TD><TD>12.62 MB</TD><TD>0.5200</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>0:31.03</TD><TD>0:33.69</TD><TD>12.63 MB</TD><TD>0.5207</TD></TR>
+               <TR><TD>Bonk 0.5</TD><TD>1:08.18</TD><TD>0:49.70</TD><TD>12.71 MB</TD><TD>0.5237</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (fast)</TD><TD>0:16.91</TD><TD>0:20.60</TD><TD>12.80 MB</TD><TD>0.5277</TD></TR>
+               <TR><TD>flac 1.0.3 (-8)</TD><TD>1:45.27</TD><TD>0:14.22</TD><TD>12.82 MB</TD><TD>0.5285</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>0:24.98</TD><TD>0:13.94</TD><TD>12.92 MB</TD><TD>0.5324</TD></TR>
+               <TR><TD>flac 1.0.3 (-3)</TD><TD>0:19.00</TD><TD>0:13.16</TD><TD>12.98 MB</TD><TD>0.5348</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WavPack 3.91 (high)</TD><TD>0:13.72</TD><TD>?</TD><TD>13.06 MB</TD><TD>0.5384</TD></TR>
+               <TR><TD>Kexis 0.2.2</TD><TD>0:33.43</TD><TD>0:28.11</TD><TD>13.30 MB</TD><TD>0.5481</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>0:17.80</TD><TD>0:11.46</TD><TD>13.32 MB</TD><TD>0.5489</TD></TR>
+               <TR><TD>flac 1.0.3 (-1)</TD><TD>0:17.29</TD><TD>0:14.53</TD><TD>13.34 MB</TD><TD>0.5499</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>13.41 MB</TD><TD>0.5528</TD></TR>
+               <TR><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>0:22.30</TD><TD>0:13.87</TD><TD>13.42 MB</TD><TD>0.5531</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>0:16.37</TD><TD>?</TD><TD>13.72 MB</TD><TD>0.5655</TD></TR>
+               <TR>
+                       <TD COLSPAN="6" BGCOLOR="#F4F4CC">&nbsp;</TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" ROWSPAN="21" BGCOLOR="#F4F4CC">
+                               L. Subramaniam<BR><I>Raga Sivapriya</I><BR>213.56 MB
+                       </TD>
+               </TR>
+               <TR><TD>Monkey's Audio 3.96 (extra high)</TD><TD>7:17.86</TD><TD>7:50.75</TD><TD>91.73 MB</TD><TD>0.4295</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>optimFROG 4.21 (mode 4 @ 1x)</TD><TD>93:05.00</TD><TD>93:16.00</TD><TD>92.05 MB</TD><TD>0.4310</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 1 @ 4x)</TD><TD>6:36.86</TD><TD>6:57.69</TD><TD>92.76 MB</TD><TD>0.4343</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (high)</TD><TD>3:46.78</TD><TD>4:12.36</TD><TD>93.69 MB</TD><TD>0.4387</TD></TR>
+               <TR><TD>optimFROG 4.21 (mode 0 @ 4x)</TD><TD>4:29.57</TD><TD>4:53.01</TD><TD>94.74 MB</TD><TD>0.4436</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Monkey's Audio 3.96 (normal)</TD><TD>3:09.07</TD><TD>3:35.10</TD><TD>94.81 MB</TD><TD>0.4439</TD></TR>
+               <TR><TD>Monkey's Audio 3.96 (fast)</TD><TD>2:25.31</TD><TD>2:58.92</TD><TD>95.70 MB</TD><TD>0.4481</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>RKAU 1.07 (normal)</TD><TD>7:16.09</TD><TD>5:30.33</TD><TD>95.91 MB</TD><TD>0.4490</TD></TR>
+               <TR><TD>LPAC 1.40 (-r, normal)</TD><TD>4:34.06</TD><TD>2:37.68</TD><TD>96.84 MB</TD><TD>0.4534</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-8)</TD><TD>15:16.08</TD><TD>2:00.10</TD><TD>98.01 MB</TD><TD>0.4589</TD></TR>
+               <TR BGCOLOR="#FFFFFF"><TD>flac 1.0.3 (-5, default)</TD><TD>3:37.33</TD><TD>1:58.73</TD><TD>98.41 MB</TD><TD>0.4608</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WavPack 3.91 (high)</TD><TD>1:58.32</TD><TD>?</TD><TD>98.82 MB</TD><TD>0.4627</TD></TR>
+               <TR><TD>Bonk 0.5</TD><TD>9:56.25</TD><TD>7:13.68</TD><TD>98.94 MB</TD><TD>0.4633</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>flac 1.0.3 (-3)</TD><TD>2:42.82</TD><TD>1:56.73</TD><TD>99.21 MB</TD><TD>0.4645</TD></TR>
+               <TR><TD>Ogg Squish 0.98</TD><TD>?</TD><TD>?</TD><TD>101.88 MB</TD><TD>0.4770</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Shorten 3.2a (-p8 -b2048)</TD><TD>3:10.19</TD><TD>1:56.31</TD><TD>102.60 MB</TD><TD>0.4804</TD></TR>
+               <TR><TD>Shorten 3.2a (-p0 -b256, default)</TD><TD>2:36.48</TD><TD>1:42.35</TD><TD>102.84 MB</TD><TD>0.4815</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>Kexis 0.2.2</TD><TD>4:49.84</TD><TD>3:59.93</TD><TD>103.57 MB</TD><TD>0.4849</TD></TR>
+               <TR><TD>flac 1.0.3 (-1)</TD><TD>2:22.18</TD><TD>1:54.92</TD><TD>103.59 MB</TD><TD>0.4850</TD></TR>
+               <TR BGCOLOR="#D3D4C5"><TD>WaveZIP</TD><TD>2:25.86</TD><TD>?</TD><TD>107.47 MB</TD><TD>0.5032</TD></TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+</CENTER>
+
+<P>&nbsp;Copyright (c) 2000,2001,2002 Josh Coalson</P>
+
+</BODY>
+</HTML>
diff --git a/doc/html/developers.html b/doc/html/developers.html
new file mode 100644 (file)
index 0000000..09fd049
--- /dev/null
@@ -0,0 +1,170 @@
+<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
+<!-- Copyright (c)  2000,2001,2002  Josh Coalson -->
+<!-- Permission is granted to copy, distribute and/or modify this document -->
+<!-- under the terms of the GNU Free Documentation License, Version 1.1 -->
+<!-- or any later version published by the Free Software Foundation; -->
+<!-- with no invariant sections. -->
+<!-- A copy of the license can be found at http://www.gnu.org/copyleft/fdl.html -->
+<HTML>
+<HEAD>
+       <TITLE>FLAC - developers</TITLE>
+</HEAD>
+
+<BODY MARGINWIDTH="0" MARGINHEIGHT="0" LEFTMARGIN="0" RIGHTMARGIN="0" TOPMARGIN="0" BGCOLOR="#99CC99" TEXT="#000000" LINK="#336699" VLINK="#336699" ALINK="#336699">
+
+<TABLE BORDER=0 WIDTH="100%" CELLPADDING=1 CELLSPACING=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#000000"><A HREF="http://flac.sourceforge.net/"><IMG SRC="images/logo.jpg" ALIGN=CENTER ALT="FLAC Logo" BORDER=0 HSPACE=0></a></TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="25" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#D3D4C5">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="index.html">home</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="news.html">news</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="download.html">download</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="features.html">features</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="goals.html">goals</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="format.html">format</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="id.html">id</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="comparison.html">comparison</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="documentation.html">documentation</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;developers&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#EEEED4">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;english&nbsp;&nbsp;</TD><TD BGCOLOR="#EEEED4" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="ru/developers.html">russian</A>&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<CENTER>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="15" ALT=""></TD></TR></TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <B><FONT SIZE="+2">FLAC developers</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               FLAC is an open source project and we are happy to enlist the help of anyone who wants to contribute.  You can do this to a limited extent through the <A HREF="http://sourceforge.net/mail/?group_id=13478">mailing list</A> but if you have major changes to make to the code it's best to <A HREF="http://sourceforge.net/project/memberlist.php?group_id=13478">sign up as a developer</A>.  In either case, make sure to check out the <A HREF="goals.html">FLAC goals</A> first; there are some thing the we <B>don't</B> want added to FLAC, like copy protection and lossy compression.
+       </P>
+       <P>
+               High priority items are:
+       </P>
+       <P>
+       <UL>
+               <LI>
+                       More input plugins.  Currently there are plugins for XMMS and Winamp; Freeamp is in the works.  More is better!
+               </LI>
+               <LI>
+                       Improving the compression methods.
+               </LI>
+       </UL>
+       </P>
+       <P>
+               Some other "nice-to-haves":
+       </P>
+       <P>
+       <UL>
+               <LI>
+                       Fix the MSVC makefiles to make libFLAC.dll (instead of just the .lib).
+               </LI>
+               <LI>
+                       Configurable ID3V1 support and ID3V2 support in the plugins.
+               </LI>
+               <LI>
+                       Support more input types than just WAVE and raw in <TT><B>flac</B></TT>.
+               </LI>
+               <LI>
+                       A better logo!  Gimp jedi I'm not...
+               </LI>
+       </UL>
+       </P>
+       <P>
+               Things that are in the works (check the flac-dev mailing list):
+       </P>
+       <P>
+       <UL>
+               <LI>
+                       Windows ACM codec.
+               </LI>
+               <LI>
+                       FLAC module for <A HREF="http://oss.sgi.com/projects/audiofile/">Audiofile</A>.
+               </LI>
+       </UL>
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <B><FONT SIZE="+2">third-party developers</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               FLAC is open to third-party developers who want to add support for FLAC into their programs.  All the necessary functionality is contained the library libFLAC which is licensed under the <A HREF="http://www.gnu.org/copyleft/lesser.html">LGPL</A>.  The relevant documentation here is:
+               <UL>
+                       <LI>The <A HREF="documentation.html#libflac">libFLAC API documentation</A> and <A HREF="documentation.html#libflacpp">libFLAC++ API documentation</A></LI>
+                       <LI>The <A HREF="format.html">formal description</A> of the FLAC format.</LI>
+                       <LI>The <A HREF="id.html">ID registration page</A> for registering an ID if you need to write custom metadata.</LI>
+               </UL>
+       </P>
+       <P>
+               There also are several examples in the FLAC code base of the use of libFLAC and libFLAC++ that may also be helpful.  Visit the <A HREF="download.html">download page</A> for instructions on how to get the source.
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+</CENTER>
+
+<P>&nbsp;Copyright (c) 2000,2001,2002 Josh Coalson</P>
+
+</BODY>
+</HTML>
diff --git a/doc/html/documentation.html b/doc/html/documentation.html
new file mode 100644 (file)
index 0000000..776d125
--- /dev/null
@@ -0,0 +1,1095 @@
+<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
+<!-- Copyright (c)  2000,2001,2002  Josh Coalson -->
+<!-- Permission is granted to copy, distribute and/or modify this document -->
+<!-- under the terms of the GNU Free Documentation License, Version 1.1 -->
+<!-- or any later version published by the Free Software Foundation; -->
+<!-- with no invariant sections. -->
+<!-- A copy of the license can be found at http://www.gnu.org/copyleft/fdl.html -->
+<HTML>
+<HEAD>
+       <TITLE>FLAC - documentation</TITLE>
+</HEAD>
+
+<BODY MARGINWIDTH="0" MARGINHEIGHT="0" LEFTMARGIN="0" RIGHTMARGIN="0" TOPMARGIN="0" BGCOLOR="#99CC99" TEXT="#000000" LINK="#336699" VLINK="#336699" ALINK="#336699">
+
+<TABLE BORDER=0 WIDTH="100%" CELLPADDING=1 CELLSPACING=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#000000"><A HREF="http://flac.sourceforge.net/"><IMG SRC="images/logo.jpg" ALIGN=CENTER ALT="FLAC Logo" BORDER=0 HSPACE=0></a></TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="25" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#D3D4C5">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="index.html">home</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="news.html">news</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="download.html">download</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="features.html">features</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="goals.html">goals</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="format.html">format</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="id.html">id</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="comparison.html">comparison</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;documentation&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="developers.html">developers</A>&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#EEEED4">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;english&nbsp;&nbsp;</TD><TD BGCOLOR="#EEEED4" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="ru/documentation.html">russian</A>&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<CENTER>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="15" ALT=""></TD></TR></TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <B><FONT SIZE="+2">documentation</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               This page is broken up into the following sections:
+               <UL>
+                       <LI><A HREF="#format">format</A> - the user-level view of the FLAC format (for a more detailed explanation see the <A HREF="format.html">format page</A>).</LI>
+                       <LI><A HREF="#flac">flac</A> - the usage of the command-line file encoder/decoder <B><TT>flac</TT></B>.</LI>
+                       <LI><A HREF="#metaflac">metaflac</A> - the usage of the command-line FLAC metadata editor <B><TT>metaflac</TT></B>.</LI>
+                       <LI><A HREF="#plugins">plugins</A> - documentation for the various input plugins.</LI>
+                       <LI><A HREF="#libflac">libFLAC API</A> - for developers who want to add FLAC support to their programs.</LI>
+                       <LI><A HREF="#libflacpp">libFLAC++ API</A> - the documentation for object layer around libFLAC.</LI>
+                       <LI><A HREF="#bugs">bugs</A> - known bugs.</LI>
+                       <LI><A HREF="#monkey">How to add FLAC support to the Monkey's Audio GUI</A></LI>
+               </UL>
+       </P>
+       <P>
+               Keep in mind that the online version of this document will always apply to the latest release.  For older releases, check the documentation included with the release package.
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <A NAME="format"><B><FONT SIZE="+2">format</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               <B><TT>flac</TT></B> has been tuned so that the default options yield a good speed vs. compression tradeoff for many kinds of input.  However, if you are looking to maximize the compression rate or speed, or want to use the full power of FLAC's metadata system, this section is for you.  If not, just skip to the <A HREF="#flac">next section</A>.
+       </P>
+       <P>
+               The basic structure of a FLAC stream is:
+               <UL>
+                       <LI>The four byte string "fLaC"</LI>
+                       <LI>The <A HREF="format.html#def_STREAMINFO">STREAMINFO</A> metadata block</LI>
+                       <LI>Zero or more other metadata blocks</LI>
+                       <LI>One or more audio frames</LI>
+               </UL>
+       </P>
+       <P>
+               The first four bytes are to identify the FLAC stream.  The metadata that follows contains all the information about the stream except for the audio data itself.  After the metadata comes the encoded audio data.
+       </P>
+       <P>
+               <B>METADATA</B>
+       </P>
+       <P>
+               FLAC defines several types of metadata blocks (see the <A HREF="format.html">format</A> page for the complete list.  Metadata blocks can be any length and new ones can be defined.  A decoder is allowed to skip any metadata types it does not understand.  Only one is mandatory: the STREAMINFO block.  This block has information like the sample rate, number of channels, etc., and data that can help the decoder manage its buffers, like the minimum and maximum data rate and minimum and maximum block size.  Also included in the STREAMINFO block is the MD5 signature of the <I>unencoded</I> audio data.  This is useful for checking an entire stream for transmission errors.
+       </P>
+       <P>
+               Other blocks allow for padding, seek tables, and application-specific data.  You can see <B><TT>flac</TT></B> options below for adding PADDING blocks or specifying seek points.  FLAC does not require seek points for seeking but they can speed up seeks, or be used for cueing in editing applications.
+       <P>
+       </P>
+               Also, if you have a need of a custom metadata block, you can define your own and request an ID <A HREF="id.html">here</A>.  Then you can reserve a PADDING block of the correct size when encoding, and overwrite the padding block with your APPLICATION block after encoding.  The resulting stream will be FLAC compatible; decoders that are aware of your metadata can use it and the rest will safely ignore it.
+       </P>
+       <P>
+               <B>AUDIO DATA</B>
+       </P>
+       <P>
+               After the metadata comes the encoded audio data.  Audio data and metadata are not interleaved.  Like most audio codecs, FLAC splits the unencoded audio data into blocks, and encodes each block separately.  The encoded block is packed into a frame and appended to the stream.  The reference encoder uses a single block size for the whole stream but the FLAC format does not require it.
+       </P>
+       <P>
+               <B>BLOCKING</B>
+       </P>
+       <P>
+               The block size is an important parameter to encoding.  If it is too small, the frame overhead will lower the compression.  If it is too large, the modeling stage of the compressor will not be able to generate an efficient model.  Understanding FLAC's modeling will help you to improve compression for some kinds of input by varying the block size.  In the most general case, using linear prediction on 44.1kHz audio, the optimal block size will be between 2-6 ksamples.  <B><TT>flac</TT></B> defaults to a block size of 4608 in this case.  Using the fast fixed predictors, a smaller block size is usually preferable because of the smaller frame header.
+       </P>
+       <P>
+               <B>INTER-CHANNEL DECORRELATION</B>
+       </P>
+       <P>
+               In the case of stereo input, once the data is blocked it is optionally passed through an inter-channel decorrelation stage.  The left and right channels are converted to center and side channels through the following transformation: mid = (left + right) / 2, side = left - right.  This is a lossless process, unlike joint stereo.  For normal CD audio this can result in significant extra compression.  <B><TT>flac</TT></B> has two options for this: <TT>-m</TT> always compresses both the left-right and mid-side versions of the block and takes the smallest frame, and <TT>-M</TT>, which adaptively switches between left-right and mid-side.
+       </P>
+       <P>
+               <B>MODELING</B>
+       </P>
+       <P>
+               In the next stage, the encoder tries to approximate the signal with a function in such a way that when the approximation is subracted, the result (called the <I>residual</I>, <I>residue</I>, or <I>error</I>) requires fewer bits-per-sample to encode.  The function's parameters also have to be transmitted so they should not be so complex as to eat up the savings.  FLAC has two methods of forming approximations: 1) fitting a simple polynomial to the signal; and 2) general linear predictive coding (LPC).  I will not go into the details here, only some generalities that involve the encoding options.
+       </P>
+       <P>
+               First, fixed polynomial prediction (specified with <TT>-l 0</TT>) is much faster, but less accurate than LPC.  The higher the maximum LPC order, the slower, but more accurate, the model will be.  However, there are diminishing returns with increasing orders.  Also, at some point (usually around order 9) the part of the encoder that guesses what is the best order to use will start to get it wrong and the compression will actually decrease slightly; at that point you will have to you will have to use the exhaustive search option <TT>-e</TT> to overcome this, which is significantly slower.
+       </P>
+       <P>
+               Second, the parameters for the fixed predictors can be transmitted in 3 bits whereas the parameters for the LPC model depend on the bits-per-sample and LPC order.  This means the frame header length varies depending on the method and order you choose and can affect the optimal block size.
+       </P>
+       <P>
+               <B>RESIDUAL CODING</B>
+       </P>
+       <P>
+               Once the model is generated, the encoder subracts the approximation from the original signal to get the residual (error) signal.  The error signal is then losslessly coded.  To do this, FLAC takes advantage of the fact that the error signal generally has a Laplacian (two-sided geometric) distribution, and that there are a set of special Huffman codes called Rice codes that can be used to efficiently encode these kind of signals quickly and without needing a dictionary.
+       </P>
+       <P>
+               Rice coding involves finding a single parameter that matches a signal's distribution, then using that parameter to generate the codes.  As the distribution changes, the optimal parameter changes, so FLAC supports a method that allows the parameter to change as needed.  The residual can be broken into several <I>contexts</I> or <I>partitions</I>, each with it's own Rice parameter.  <B><TT>flac</TT></B> allows you to specify how the partitioning is done with the <TT>-r</TT> option.  The residual can be broken into 2^<I>n</I> partitions, by using the option <TT>-r n,n</TT>.  The parameter <I>n</I> is called the <I>partition order</I>.  Furthermore, the encoder can be made to search through <I>m</I> to <I>n</I> partition orders, taking the best one, by specifying <TT>-r m,n</TT>.  Generally, the choice of n does not affect encoding speed but m,n does.  The larger the difference between m and n, the more time it will take the encoder to search for the best order.  The block size will also affect the optimal order.
+       </P>
+       <P>
+               <B>FRAMING</B>
+       </P>
+       <P>
+               An audio frame is preceded by a frame header and trailed by a frame footer.  The header starts with a sync code, and contains the minimum information necessary for a decoder to play the stream, like sample rate, bits per sample, etc.  It also contains the block or sample number and an 8-bit CRC of the frame header.  The sync code, frame header CRC, and block/sample number allow resynchronization and seeking even in the absence of seek points.  The frame footer contains a 16-bit CRC of the entire encoded frame for error detection.  If the reference decoder detects a CRC error it will generate a silent block.
+       </P>
+       <P>
+               <B>MISCELLANEOUS</B>
+       </P>
+       <P>
+               In order to support come common types of metadata, the reference decoder knows how to skip ID3V1 and ID3V2 tags so it is safe to tag FLAC files in this way.  ID3V2 tags must come at the beginning of the file (before the "fLaC" marker) and ID3V1 tags must come at the end of the file.
+       </P>
+       <P>
+               <B><TT>flac</TT></B> has a verify option <TT>-V</TT> that verifies the output while encoding.  With this option, a decoder is run in parallel to the encoder and its output is compared against the original input.  If a difference is found <B><TT>flac</TT></B> will stop with an error.
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <A NAME="flac"><B><FONT SIZE="+2">flac</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               <B><TT>flac</TT></B> is the command-line file encoder/decoder.  The input to the encoder and the output to the decoder must either be RIFF WAVE format, or raw interleaved sample data.  <B><TT>flac</TT></B> only supports linear PCM samples (in other words, no A-LAW, uLAW, etc.).  Another restriction (hopefully short-term) is that the input must be 8, 16, or 24 bits per sample.  This is not a limitation of the FLAC format, just the reference encoder/decoder.
+       </P>
+       <P>
+               <B><TT>flac</TT></B> assumes that files ending in ".wav" or that have the RIFF WAVE header present are WAVE files; this may be overridden with a command-line option; it also assumes that files ending in ".ogg" are Ogg-FLAC files.  Other than this, <B><TT>flac</TT></B> makes no assumptions about file extensions, though the convention is that FLAC files have the extension ".flac" (or ".fla" on ancient file systems like FAT-16).
+       </P>
+       <P>
+               Before going into the full command-line description, a few other things help to sort it out: 1) <B><TT>flac</TT></B> encodes by default, so you must use <B>-d</B> to decode; 2) the options <B><TT>-0</TT></B> .. <B><TT>-8</TT></B> (or <B><TT>--fast</B></TT> and <B><TT>--best</B></TT>) that control the compression level actually are just synonyms for different groups of specific encoding options (described later) and you can get the same effect by using the same options; 3) <B><TT>flac</TT></B> behaves similarly to gzip in the way it handles input and output files.
+       </P>
+       <P>
+               <B><TT>flac</TT></B> will be invoked one of four ways, depending on whether you are encoding, decoding, testing, or analyzing:
+               <UL>
+               <LI>
+                       Encoding: flac [-s] [--skip #] [-V] [<I><A HREF="#format_options">&lt;format-options&gt;</A></I>] [<I><A HREF="#encoding_options">&lt;encoding options&gt;</A></I>] [inputfile [...]]
+               </LI>
+               <LI>
+                       Decoding: flac -d [-s] [--skip #] [-F] [<I><A HREF="#format_options">&lt;format-options&gt;</A></I>] [inputfile [...]]
+               </LI>
+               <LI>
+                       Testing: flac -t [-s] [inputfile [...]]
+               </LI>
+               <LI>
+                       Analyzing: flac -a [-s] [--skip #] [<I><A HREF="#analysis_options">&lt;analysis-options&gt;</A></I>] [inputfile [...]]
+               </LI>
+               </UL>
+       </P>
+       <P>
+               In any case, if no <TT>inputfile</TT> is specified, stdin is assumed.  If only one inputfile is specified, it may be "-" for stdin.  When stdin is used as input, <B><TT>flac</TT></B> will write to stdout.  Otherwise <B><TT>flac</TT></B> will perform the desired operation on each input file to similarly named output files (meaning for encoding, the extension will be replaced with ".flac", or appended with ".flac" if the input file has no extension, and for decoding, the extension will be ".wav" for WAVE output and ".raw" for raw output).  The original file is not deleted unless --delete-input-file is specified.
+       </P>
+       <P>
+               If you are encoding/decoding from stdin to a file, you should use the -o option like so:
+               <UL>
+               <LI>
+                       flac [options] -o outputfile
+               </LI>
+               <LI>
+                       flac -d [options] -o outputfile
+               </LI>
+               </UL>
+               which are better than:
+               <UL>
+               <LI>
+                       flac [options] &gt; outputfile
+               </LI>
+               <LI>
+                       flac -d [options] &gt; outputfile
+               </LI>
+               </UL>
+               since the former allows flac to seek backwards to write the STREAMINFO or RIFF WAVE header contents when necessary.
+       </P>
+       <P>
+               Also, you can force output data to go to stdout using <TT>-c</TT>.
+       </P>
+       <P>The encoding options affect the compression ratio and encoding speed.  The format options are used to tell <B><TT>flac</TT></B> the arrangement of samples if the input file (or output file when decoding) is a raw file.  If it is a RIFF WAVE file the format options are not needed since they are read from the WAVE header.
+       </P>
+       <P>
+               In test mode, <B><TT>flac</TT></B> acts just like in decode mode, except no output file is written.  Both decode and test modes detect errors in the stream, but they also detect when the MD5 signature of the decoded audio does not match the stored MD5 signature, even when the bitstream is valid.
+       </P>
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="general_options"><FONT SIZE="+1"><B>General Options</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-H</TT>
+                       </TD>
+                       <TD>
+                               Show the long usage screen.  Running <B><TT>flac</TT></B> without arguments shows the short help screen by default.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-d</TT>
+                       </TD>
+                       <TD>
+                               Decode (<B><TT>flac</TT></B> encodes by default).  <B><TT>flac</TT></B> will exit with an exit code of <TT>1</TT> (and print a message, even in silent mode) if there were any errors during decoding, including when the MD5 checksum does not match the decoded output.  Otherwise the exit code will be <TT>0</TT>.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-t</TT>
+                       </TD>
+                       <TD>
+                               Test (same as <B><TT>-d</TT></B> except no decoded file is written).  The exit codes are the same as in decode mode.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-a</TT>
+                       </TD>
+                       <TD>
+                               Analyze (same as <B><TT>-d</TT></B> except an analysis file is written).  The exit codes are the same as in decode mode.  This option is mainly for developers; the output will be a text file that has data about each frame and subframe.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-c</TT>
+                       </TD>
+                       <TD>
+                               Write output to stdout
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-s</TT>
+                       </TD>
+                       <TD>
+                               Silent: do not show encoding/decoding statistics.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-o filename</TT>
+                       </TD>
+                       <TD>
+                               Force the output file name (usually <TT><B>flac</B></TT> just changes the extension).  May only be used when encoding a single file.  May not be used in conjunction with <TT>--output-prefix</T>.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>--output-prefix string</TT>
+                       </TD>
+                       <TD>
+                               Prefix each output file name with the given string.  This can be useful for encoding/decoding files to a different directory.  Make sure if your string is a path name that it ends with a trailing '<TT>/</TT>' slash.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>--delete-input-file</TT>
+                       </TD>
+                       <TD>
+                               Automatically delete the input file after a successful encode or decode.  If there was an error (including a verify error) the input file is left intact.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>--skip #</TT>
+                       </TD>
+                       <TD>
+                               Skip over the first # of samples of the input.  This works for both encoding and decoding, but not testing.
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="analysis_options"><FONT SIZE="+1"><B>Analysis Options</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>--a-rtext</TT>
+                       </TD>
+                       <TD>
+                               Includes the residual signal in the analysis file.  This will make the file <B>very</B> big, much larger than even the decoded file.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>--a-rgp</TT>
+                       </TD>
+                       <TD>
+                               Generates a gnuplot file for every subframe; each file will contain the residual distribution of the subframe.  This will create a <B>lot</B> of files.
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="decoding_options"><FONT SIZE="+1"><B>Decoding Options</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-F</TT>
+                       </TD>
+                       <TD>
+                               By default <B><TT>flac</TT></B> stops decoding with an error and removes the partially decoded file if it encounters a bitstream error.  With <TT>-F</TT>, errors are still printed but <B><TT>flac</TT></B> will continue decoding to completion.  Note that errors may cause the decoded audio to be missing some samples or have silent sections.
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="encoding_options"><FONT SIZE="+1"><B>Encoding Options</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>--ogg</TT>
+                       </TD>
+                       <TD>
+                               When encoding, generate Ogg-FLAC output instead of native-FLAC.  Ogg-FLAC streams are FLAC streams wrapped in an Ogg transport layer.  The resulting file should have an '.ogg' extension and will still be decodable by <TT><B>flac</B></TT>.<P>
+                               When decoding, force the input to be treated as Ogg-FLAC.  This is useful when piping input from stdin or when the filename does not end in '.ogg'.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>--lax</TT>
+                       </TD>
+                       <TD>
+                               Allow encoder to generate non-Subset files.  The resulting FLAC file may not be streamable, so you should only use this option in combination with custom encoding options meant for archival.  File decoders will still be able play (and seek in) such files.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>--sector-align</TT>
+                       </TD>
+                       <TD>
+                               Align encoding of multiple CD format WAVE files on sector boundaries.  This option is only allowed when encoding WAVE files, all of which have a 44.1kHz sample rate and 2 channels.  With <TT>--sector-align</TT>, the encoder will align the resulting .flac streams so that their lengths are even multiples of a CD sector (1/75th of a second, or 588 samples).  It does this by carrying over any partial sector at the end of each WAVE file to the next stream.  The last stream will be padded to alignment with zeroes.<P>
+                               This option will have no effect if the files are already aligned (as is the normally the case with WAVE files ripped from a CD).  <TT><B>flac</B></TT> can only align a set of files given in one invocation of <TT><B>flac</B></TT>.<P>
+                               <B>WARNING:</B> The ordering of files is important!  If you give a command like '<TT>flac --sector-align *.wav</TT>' the shell may not expand the wildcard to the order you expect.  To be safe you should '<TT>echo *.wav</TT>' first to confirm the order, or be explicit like '<TT>flac --sector-align 8.wav 9.wav 10.wav</TT>'.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-S { # | X | #x }</TT>
+                       </TD>
+                       <TD>
+                               Include a point or points in a SEEKTABLE:<BR>
+                               <UL>
+                               <LI>
+                                       <TT>#&nbsp;</TT> : a specific sample number for a seek point
+                               </LI>
+                               <LI>
+                                       <TT>X&nbsp;</TT> : a placeholder point (always goes at the end of the SEEKTABLE)
+                               </LI>
+                               <LI>
+                                       <TT>#x</TT> : # evenly spaced seekpoints, the first being at sample 0
+                               </LI>
+                               </UL>
+                               You may use many -S options; the resulting SEEKTABLE will be the unique-ified union of all such values.<BR>
+                               With no -S options, flac defaults to '-S 100x'.  Use -S- for no SEEKTABLE.<BR>
+                               <B>NOTE:</B> -S #x will not work if the encoder can't determine the input size before starting.<BR>
+                               <B>NOTE:</B> if you use -S # and # is >= samples in the input, there will be either no seek point entered (if the input size is determinable before encoding starts) or a placeholder point (if input size is not determinable).<BR>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-P #</TT>
+                       </TD>
+                       <TD>
+                               Tell the encoder to write a <TT>PADDING</TT> metadata block of the given length (in bytes) after the <TT>STREAMINFO</TT> block.  This is useful if you plan to tag the file later with an <TT>APPLICATION</TT> block; instead of having to rewrite the entire file later just to insert your block, you can write directly over the <TT>PADDING</TT> block.  Note that the total length of the <TT>PADDING</TT> block will be 4 bytes longer than the length given because of the 4 metadata block header bytes.  You can force no <TT>PADDING</TT> block at all to be written with <TT>-P-</TT>, which is the default.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-b #</TT>
+                       </TD>
+                       <TD>
+                               Specify the block size in samples.  The default is 1152 for -l 0, otherwise 4608.  Subset streams must use one of 192/576/1152/2304/4608/256/512/1024/2048/4096/8192/16384/32768.  The reference encoder uses the same block size for the entire stream.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-m</TT>
+                       </TD>
+                       <TD>
+                               Enable mid-side coding (only for stereo streams).  Tends to increase compression by a few percent on average.  For each block both the stereo pair and mid-side versions of the block will be encoded, and smallest resulting frame will be stored.  Currently mid-side encoding is only available when bits-per-sample <= 16.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-M</TT>
+                       </TD>
+                       <TD>
+                               Enable loose mid-side coding (only for stereo streams).  Like <TT>-m</TT> but the encoder adaptively switches between independent and mid-side coding, which is faster but yields less compression than <TT>-m</TT> (which does an exhaustive search).
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-0 .. -8</TT>
+                       </TD>
+                       <TD>
+                               Fastest compression .. highest compression.  The default is <TT>-5</TT>.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-0</TT>
+                       </TD>
+                       <TD>
+                               Synonymous with <TT>-l 0 -b 1152 -r 2,2</TT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-1</TT>
+                       </TD>
+                       <TD>
+                               Synonymous with <TT>-l 0 -b 1152 -M -r 2,2</TT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-2</TT>
+                       </TD>
+                       <TD>
+                               Synonymous with <TT>-l 0 -b 1152 -m -r 3</TT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-3</TT>
+                       </TD>
+                       <TD>
+                               Synonymous with <TT>-l 6 -b 4608 -r 3,3</TT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-4</TT>
+                       </TD>
+                       <TD>
+                               Synonymous with <TT>-l 8 -b 4608 -M -r 3,3</TT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-5</TT>
+                       </TD>
+                       <TD>
+                               Synonymous with <TT>-l 8 -b 4608 -m -r 3,3</TT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-6</TT>
+                       </TD>
+                       <TD>
+                               Synonymous with <TT>-l 8 -b 4608 -m -r 4</TT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-7</TT>
+                       </TD>
+                       <TD>
+                               Synonymous with <TT>-l 8 -b 4608 -m -e -r 6</TT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-8</TT>
+                       </TD>
+                       <TD>
+                               Synonymous with <TT>-l 12 -b 4608 -m -e -r 6</TT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>--fast</TT>
+                       </TD>
+                       <TD>
+                               Fastest compression.  Currently synonymous with <TT>-0</TT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>--best</TT>
+                       </TD>
+                       <TD>
+                               Highest compression.  Currently synonymous with <TT>-8</TT>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-e</TT>
+                       </TD>
+                       <TD>
+                               Exhaustive model search (expensive!).  Normally the encoder estimates the best model to use and encodes once based on the estimate.  With an exhaustive model search, the encoder will generate subframes for every order and use the smallest.  If the max LPC order is high this can significantly increase the encode time but can shave off another 0.5%.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-E</TT>
+                       </TD>
+                       <TD>
+                               Do escape coding in the entropy coder.  This causes the encoder to use an unencoded representation of the residual in a partition if it is smaller.  It increases the runtime and usually results in an improvement of less than 1%.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-l #</TT>
+                       </TD>
+                       <TD>
+                               Specifies the maximum LPC order.  This number must be <= 32.  If 0, the encoder will not attempt generic linear prediction, and use only fixed predictors.  Using fixed predictors is faster but usually results in files being 5-10% larger.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-q #</TT>
+                       </TD>
+                       <TD>
+                               Specifies the precision of the quantized LP coefficients, in bits.  The default is <B><TT>-q 0</TT></B>, which means let the encoder decide based on the signal.  Unless you really know your input file it's best to leave this up to the encoder.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-p</TT>
+                       </TD>
+                       <TD>
+                               Do exhaustive LP coefficient quantization optimization.  This option overrides any <B><TT>-q</TT></B> option.  It is expensive and typically will only improve the compression a tiny fraction of a percent.  <B><TT>-q</TT></B> has no effect when <B><TT>-l 0</TT></B> is used.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-r [#,]#</TT>
+                       </TD>
+                       <TD>
+                               Set the [min,]max residual partition order.  The min value defaults to 0 if unspecified.<P>
+                               By default the encoder uses a single Rice parameter for the subframe's entire residual.  With this option, the residual is iteratively partitioned into 2^min# .. 2^max# pieces, each with its own Rice parameter.  Higher values of max# yield diminishing returns.  The most bang for the buck is usually with <B><TT>-r 2,2</TT></B> (more for higher block sizes).  This usually shaves off about 1.5%.  The technique tends to peak out about when blocksize/(2^n)=128.  Use <B><TT>-r 0,16</TT></B> to force the highest degree of optimization.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-V</TT>
+                       </TD>
+                       <TD>
+                               Verify the encoding process.  With this option, <B><TT>flac</TT></B> will create a parallel decoder that decodes the output of the encoder and compares the result against the original.  It will abort immediately with an error if a mismatch occurs.  <B><TT>-V</TT></B> increases the total encoding time but is guaranteed to catch any unforseen bug in the encoding process.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                       </TD>
+                       <TD>
+                               <TT>-F-</TT>, <TT>-S-</TT>, <TT>-P-</TT>, <TT>-m-</TT>, <TT>-e-</TT>, <TT>-E-</TT>, <TT>-p-</TT>, <TT>-V-</TT>, <TT>--delete-input-file-</TT>, <TT>--lax-</TT>, <TT>--sector-align-</TT> can all be used to turn off a particular option.
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="format_options"><FONT SIZE="+1"><B>Format Options</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-fb | -fl</TT>
+                       </TD>
+                       <TD>
+                               Specify big-endian | little-endian byte order in the raw file.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-fc #</TT>
+                       </TD>
+                       <TD>
+                               Specify the number of channels in the raw file.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-fp #</TT>
+                       </TD>
+                       <TD>
+                               Specify the number of bits per sample in the raw file.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-fs #</TT>
+                       </TD>
+                       <TD>
+                               Specify the sample rate of the raw file.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-fu</TT>
+                       </TD>
+                       <TD>
+                               Specify that the samples in the raw file are unsigned (the default is signed).
+                       </TD>
+               </TR>
+               <TR>
+                       <TD NOWRAP ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <TT>-fr</TT>
+                       </TD>
+                       <TD>
+                               Treat the input file (or output file if decoding) as a raw file, regardless of the extension.
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <A NAME="metaflac"><B><FONT SIZE="+2">metaflac</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               <B><TT>metaflac</TT></B> is the command-line <TT>.flac</TT> file metadata editor.  You can use it to list the contents of blocks, delete or insert blocks, and manage padding.
+       </P>
+       <P>
+               The documentation for <B><TT>metaflac</TT></B> is currently being rewritten, but the usage screen should explain it pretty well.  Do <TT>metaflac --help</TT> to see the full usage.
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<A NAME="plugins"></A>
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <A NAME="xmms_plugin"><B><FONT SIZE="+2">xmms plugin</FONT></B></A>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               All that is necessary is to copy <B><TT>libxmms-flac.so</TT></B> to the directory where XMMS looks for input plugins (usually <B><TT>/usr/lib/xmms/Input</TT></B>).  There is nothing else to configure.  Make sure to restart XMMS before trying to play any <TT>.flac</TT> files.
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <A NAME="winamp2_plugin"><B><FONT SIZE="+2">winamp2 plugin</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               There are two Winamp plugins; one for Winamp versions 2.x and one for Winamp versions 3.x.  If you are using Winamp 2.x, all that is necessary is to copy <B><TT>in_flac.dll</TT></B> to the <B><TT>Plugins/</TT></B> directory of your Winamp installation.  There is nothing else to configure.  Make sure to restart Winamp before trying to play any <TT>.flac</TT> files.
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <A NAME="winamp3_plugin"><B><FONT SIZE="+2">winamp3 plugin</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               There are two Winamp plugins; one for Winamp versions 2.x and one for Winamp versions 3.x.  If you are using Winamp 3.x, all that is necessary is to copy <B><TT>cnv_flacpcm.wac</TT></B> to the <B><TT>Wacs/</TT></B> directory of your Winamp installation.  There is nothing else to configure.  Make sure to restart Winamp before trying to play any <TT>.flac</TT> files.
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <A NAME="libflac"><B><FONT SIZE="+2">libFLAC</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               The FLAC library <B><TT>libFLAC</TT></B> is a C implementation of reference encoders and decoders, and a metadata interface.  By linking against <B><TT>libFLAC</TT></B> and writing a little code, it is relatively easy to add FLAC support to another program.  The library is licensed under the <A HREF="http://www.gnu.org/copyleft/lesser.html">LGPL</A>.  Complete source code of <B><TT>libFLAC</TT></B> as well as the command-line encoder and plugins is available and is a useful source of examples.
+       </P>
+       <P>
+               There is also a C++ object wrapper around <B><TT>libFLAC</TT></B> called <B><TT>libFLAC++</TT></B>; see the documentation <A HREF="#libflacpp">below</A>.
+       </P>
+       <P>
+               <B><TT>libFLAC</TT></B> usually only requires the standard C library and C math library.  In particular, threading is not used so there is no dependency on a thread library.  However, <B><TT>libFLAC</TT></B> does not use global variables and should be thread-safe.
+       </P>
+       <P>
+               The <B><TT>libFLAC</TT></B> interface is described in the public header files in the <TT>include/FLAC/</TT> directory.  The public headers and the compiled library are all that is needed to compile and link against the library.  Note that none of the code in <TT>src/libFLAC/</TT>, including the private header files in <TT>src/libFLAC/include/</TT> is required.
+       </P>
+       <P>
+               Aside from encoders and decoders, <B><TT>libFLAC</TT></B> provides a powerful metadata interface for manipulating metadata in FLAC files.  It allows the user to add, delete, and modify FLAC metadata blocks and it can automatically take advantage of PADDING blocks to avoid rewriting the entire FLAC file when changing the size of the metadata.  The documentation for the metadata interface is currently being rewritten but there are extensive usage comments in the header file <TT>include/FLAC/metadata.h</TT>.
+       </P>
+       <P>
+               The basic usage of a <B><TT>libFLAC</TT></B> encoder or decoder is as follows:
+               <OL>
+                       <LI>The program creates an instance of a decoder or encoder using <TT>*_new()</TT>.</LI>
+                       <LI>The program sets the parameters of the instance and callbacks for reading, writing, error reporting, and metadata reporting using <TT>*_set_*()</TT> functions.</LI>
+                       <LI>The program initializes the instance to validate the parameters and prepare for decoding/encoding using <TT>*_init()</TT>.</LI>
+                       <LI>The program calls <TT>*_process_*()</TT> functions to encode or decode data, which subsequently calls the callbacks.</LI>
+                       <LI>The program finishes the instance with <TT>*_finish()</TT>, which flushes the input and output and resets the encoder/decoder to the unitialized state.</LI>
+                       <LI>The instance may be used again or deleted with <TT>*_delete()</TT>.</LI>
+               </OL>
+       </P>
+       <P>
+               For decoding, <B><TT>libFLAC</TT></B> provides three layers of access.  The lowest layer is non-seekable stream-level decoding, the next is seekable stream-level decoding, and the highest layer is file-level decoding.  The interfaces are described in <TT>stream_decoder.h</TT>, <TT>seekable_stream_decoder.h</TT>, and <TT>file_decoder.h</TT> respectively.  Typically you will choose the highest layer that your input source will support.
+       </P>
+       <P>
+               The stream decoder relies on callbacks for all input and output and has no provisions for seeking.  The seekable stream decoder wraps the stream decoder and exposes functions for seeking.  However, you must provide extra callbacks for seek-related operations on your stream, like seek and tell.  The file decoder wraps the seekable stream decoder and supplies most of the callbacks internally, simplifying the processing of standard files.
+       </P>
+       <P>
+               Currently there is only one level of encoder implementation which is at the stream level (<TT>stream_encoder.h</TT>).  There is currently no file encoder because seeking within a file while encoding seemed like too obscure a feature.
+       </P>
+       <P>
+               Structures and constants related to the format are defined in <TT>format.h</TT>.
+       </P>
+       <P>
+               <B>STREAM DECODER</B>
+       </P>
+       <P>
+               First we discuss the stream decoder.  The instance type is <TT>FLAC__StreamDecoder</TT>.  Typically the program will create a new instance by calling <TT>FLAC__stream_decoder_new()</TT>, then call <TT>FLAC__stream_decoder_set_*()</TT> functions to set the callbacks and client data, and call <TT>FLAC__stream_decoder_init()</TT>.  The required callbacks are:
+               <UL>
+                       <LI>Read callback - This function will be called when the decoder needs more input data.  The address of the buffer to be filled is supplied, along with the number of bytes the buffer can hold.  The callback may choose to supply less data and modify the byte count but must be careful not to overflow the buffer.  The callback then returns a status code chosen from FLAC__StreamDecoderReadStatus.</LI>
+                       <LI>Write callback - This function will be called when the decoder has decoded a single frame of data.  The decoder will pass the frame metadata as well as an array of pointers (one for each channel) pointing to the decoded audio.</LI>
+                       <LI>Metadata callback - This function will be called when the decoder has decoded a metadata block.  There will always be one STREAMINFO block per stream, followed by zero or more other metadata blocks.  These will be supplied by the decoder in the same order as they appear in the stream and always before the first audio frame (i.e. write callback).  The metadata block that is passed in must not be modified, and it doesn't live beyond the callback, so you should make a copy of it with <TT>FLAC__metadata_object_clone()</TT> if you will need it elsewhere.  Since metadata blocks can potentially be large, by default the decoder only calls the metadata callback for the STREAMINFO block; you can instruct the decoder to pass or filter other blocks with <TT>FLAC__stream_decoder_set_metadata_*()</TT> calls.</LI>
+                       <LI>Error callback - This function will be called whenever an error occurs during decoding.</LI>
+               </UL>
+       </P>
+       <P>
+               Once the decoder is initialized, your program will call one of several functions to start the decoding process:
+               <UL>
+                       <LI><B><TT>FLAC__stream_decoder_process_whole_stream()</TT></B> - Tells the decoder to start and continue processing the stream until the read callback says <TT>FLAC__STREAM_DECODER_READ_END_OF_STREAM</TT> or <TT>FLAC__STREAM_DECODER_READ_ABORT</TT>.</LI>
+                       <LI><B><TT>FLAC__stream_decoder_process_metadata()</TT></B> - Tells the decoder to start processing the stream and stop upon reaching the first audio frame.</LI>
+                       <LI><B><TT>FLAC__stream_decoder_process_one_frame()</TT></B> - Tells the decoder to process one audio frame and return.  The decoder must have processed all metadata first before calling this function.</LI>
+                       <LI><B><TT>FLAC__stream_decoder_process_remaining_frames()</TT></B> - Tells the decoder to process all remaining frames.  The decoder must have processed all metadata first but may also have processed frames with <TT>FLAC__stream_decoder_process_one_frame()</TT>.</LI>
+               </UL>
+       </P>
+       <P>
+               When the decoder has finished decoding (normally or through an abort), the instance is finished by calling <TT>FLAC__stream_decoder_finish()</TT>, which ensures the decoder is in the correct state and frees memory.  Then the instance may be deleted with <TT>FLAC__stream_decoder_delete()</TT> or initialized again to decode another stream.
+       </P>
+       <P>
+               Note that the stream decoder has no real concept of stream position, it just converts data.  To seek within a stream the callbacks have only to flush the decoder using <TT>FLAC__stream_decoder_flush()</TT> and start feeding data from the new position through the read callback.  The seekable stream decoder does just this.
+       </P>
+       <P>
+               <B>SEEKABLE STREAM DECODER</B>
+       </P>
+       <P>
+               The seekable stream decoder is a wrapper around the stream decoder which also provides seeking capability.  The instance type is <TT>FLAC__SeekableStreamDecoder</TT>.  In addition to the Read/Write/Metadata/Error callbacks of the stream decoder, the user must also provide the following:
+               <UL>
+                       <LI>Seek callback - This function will be called when the decoder wants to seek to an absolute position in the stream.</LI>
+                       <LI>Tell callback - This function will be called when the decoder wants to know the current absolute position of the stream.</LI>
+                       <LI>Length callback - This function will be called when the decoder wants to know length of the stream.  The seeking algorithm currently requires that the overall stream length be known.</LI>
+                       <LI>EOF callback - This function will be called when the decoder wants to know if it is at the end of the stream.  This could be determined from the tell and length callbacks but it may be more expensive that way.</LI>
+               </UL>
+       </P>
+       <P>
+               Seeking is exposed through the <TT>FLAC__seekable_stream_decoder_seek_absolute()</TT> method.  At any point after the seekable stream decoder has been initialized, the user can call this function to seek to an exact sample within the stream.  Subsequently, the first time the write callback is called it will contain a (possibly partial) block starting at that sample.
+       </P>
+       <P>
+               The seekable stream decoder also provides MD5 signature checking.  If this is turned on before initialization, <TT>FLAC__seekable_stream_decoder_finish()</TT> will report when the decoded MD5 signature does not match the one stored in the STREAMINFO block.  MD5 checking is automatically turned off if there is no signature in the STREAMINFO block or when a seek is attempted.
+       </P>
+       <P>
+               <B>FILE DECODER</B>
+       </P>
+       <P>
+               The file decoder is a trivial wrapper around the seekable stream decoder meant to simplfy the process of decoding from a standard file.  The instance type is <TT>FLAC__FileDecoder</TT>.  The file decoder supplies all but the Write/Metadata/Error callbacks.  The user needs only to provide the path to the file and the file decoder handles the rest.
+       </P>
+       <P>
+               Like the seekable stream decoder, seeking is exposed through the <TT>FLAC__file_decoder_seek_absolute()</TT> method.  At any point after the file decoder has been initialized, the user can call this function to seek to an exact sample within the file.  Subsequently, the first time the write callback is called it will contain a (possibly partial) block starting at that sample.
+       </P>
+       <P>
+               The file decoder also inherits MD5 signature checking from the seekable stream decoder.  If this is turned on before initialization, <TT>FLAC__file_decoder_finish()</TT> will report when the decoded MD5 signature does not match the one stored in the STREAMINFO block.  MD5 checking is automatically turned off if there is no signature in the STREAMINFO block or when a seek is attempted.
+       </P>
+       <P>
+               <B>STREAM ENCODER</B>
+       </P>
+       <P>
+               The stream encoder functions similarly to the stream decoder, but has fewer callbacks and more options.  The instance type is <TT>FLAC__StreamEncoder</TT>.  Typically the user will create a new instance by calling <TT>FLAC__stream_encoder_new()</TT>, then set the necessary parameters with <TT>FLAC__stream_encoder_set_*()</TT>, and initialize it by calling <TT>FLAC__stream_encoder_init()</TT>.
+       </P>
+       <P>
+               Unlike the decoding process, FLAC encoding has many options that can affect the speed and compression ratio.  When the user calls <TT>FLAC__stream_encoder_init()</TT> the encoder will validate the values, so you should make sure to check the returned state to see that it is FLAC__STREAM_ENCODER_OK.  When setting these parameters you should have some basic knowledge of the format (see the <A HREF="#format">user-level documentation</A> or the <A HREF="format.html">formal description</A>) but the required parameters are summarized here:
+               <UL>
+                       <LI><B><TT>streamable_subset</TT></B> - true to force the encoder to generate a <A HREF="format.html#subset">Subset stream</A>, else false.</LI>
+                       <LI><B><TT>do_mid_side_stereo</TT></B> - true to try mid-side encoding on stereo input, else false.  <TT>channels</TT> must be 2.</LI>
+                       <LI><B><TT>loose_mid_side_stereo</TT></B> - true to do adaptive mid-side switching, else false.  <TT>do_mid_side_stereo</TT> must be true.</LI>
+                       <LI><B><TT>channels</TT></B> - must be &lt;= <TT>FLAC__MAX_CHANNELS</TT>.</LI>
+                       <LI><B><TT>bits_per_sample</TT></B> - do not give the encoder wider data than what you specify here or bad things will happen.</LI>
+                       <LI><B><TT>sample_rate</TT></B> - must be &lt;= <TT>FLAC__MAX_SAMPLE_RATE</TT>.</LI>
+                       <LI><B><TT>blocksize</TT></B> - must be between <TT>FLAC__MIN_BLOCKSIZE</TT> and <TT>FLAC__MAX_BLOCKSIZE</TT>.</LI>
+                       <LI><B><TT>max_lpc_order</TT></B> - 0 implies encoder will not try general LPC, only fixed predictors; must be &lt;= <TT>FLAC__MAX_LPC_ORDER</TT>.</LI>
+                       <LI><B><TT>qlp_coeff_precision</TT></B> - must be &gt;= <TT>FLAC__MIN_QLP_COEFF_PRECISION</TT>, or 0 to let encoder select based on blocksize.  In the current imlementation <TT>qlp_coeff_precision</TT>+<TT>bits_per_sample</TT> must be &lt; 32.</LI>
+                       <LI><B><TT>do_qlp_coeff_prec_search</TT></B> - false to use <TT>qlp_coeff_precision</TT>; true to search around <TT>qlp_coeff_precision</TT> and take best.</LI>
+                       <LI><B><TT>do_escape_coding</TT></B> - true => search for escape codes in the entropy coding stage for slightly better compression.</LI>
+                       <LI><B><TT>do_exhaustive_model_search</TT></B> - false to use estimated bits per residual for scoring; true to generate all and take shortest.</LI>
+                       <LI><B><TT>min_residual_partition_order</TT></B>, <B><TT>max_residual_partition_order</TT></B> - 0 to estimate Rice parameter based on residual variance; &gt; 0 to partition the residual and use parameter for each based on mean; <TT>min_residual_partition_order</TT> and <TT>max_residual_partition_order</TT> specify the min and max Rice partition order.</LI>
+                       <LI><B><TT>rice_parameter_search_dist</TT></B> - 0 to try only calculated parameter k; else try all [k-<TT>rice_parameter_search_dist</TT>..k+<TT>rice_parameter_search_dist</TT>] parameters and use the best.</LI>
+                       <LI><B><TT>total_samples_estimate</TT></B> - May be set to 0 if unknown.  Otherwise, set this to the number of samples to be encoded.  This will allow the STREAMINFO block to be more accurate during the first pass in the event that the encoder can't seek back to the beginning of the output file to write the updated STREAMINFO block.</LI>
+                       <LI><B><TT>metadata</TT></B> - Optional array of pointers to metadata blocks to be written; NULL implies no metadata.  The STREAMINFO block is always written automatically and must not be present in the array of pointers.</LI>
+               </UL>
+       </P>
+       <P>
+               The user provides addresses for the following callbacks:
+               <UL>
+                       <LI>Write callback - This function is called anytime there is raw encoded data to write.  It may include metadata mixed with encoded audio frames and the data is not guaranteed to be aligned on frame or metadata block boundaries.</LI>
+                       <LI>Metadata callback - This function is called once at the end of encoding with the populated STREAMINFO structure.  This is so file encoders can seek back to the beginning of the file and write the STREAMINFO block with the correct statistics after encoding (like minimum/maximum frame size).</LI>
+               </UL>
+               The call to <TT>FLAC__stream_encoder_init()</TT> currently will also immediately call the write callback with the "fLaC" signature and all the encoded metadata.
+       </P>
+       <P>
+               After initializing the instance, the user may feed audio data to the encoder in one of two ways:
+               <UL>
+                       <LI>Channel separate, through <B><TT>FLAC__stream_encoder_process()</TT></B> - The user will pass an array of pointers to buffers, one for each channel, to the encoder, each of the same length.  The samples need not be block-aligned.</LI>
+                       <LI>Channel interleaved, through <B><TT>FLAC__stream_encoder_process_interleaved()</TT></B> - The user will pass a single pointer to data that is channel-interleaved (i.e. <TT>channel0_sample0, channel1_sample0, ... , channelN_sample0, channel0_sample1, ...</TT>).  Again, the samples need not be block-aligned but they must be sample-aligned, i.e. the first value should be channel0_sampleX and the last value channelN_sampleY.</LI>
+               </UL>
+       </P>
+       <P>
+               When the user is finished encoding data, it calls <TT>FLAC__stream_encoder_finish()</TT>, which causes the encoder to encode any data still in its input pipe, and call the metadata callback with the final encoding statistics.  Then the instance may be deleted with <TT>FLAC__stream_encoder_delete()</TT> or initialized again to encode another stream.
+       </P>
+       <P>
+               <B>MISCELLANEOUS</B>
+       </P>
+       <P>
+               It should be noted that any time an array of pointers to audio data is passed, the channel order currently only has meaning for stereo streams.  Channel 0 corresponds to the left channel and channel 1 corresponds to the right channel.
+       </P>
+       <P>
+               <B>METADATA</B>
+       </P>
+       <P>
+               For programs that write their own metadata, but that do not know the actual metadata until after encoding, it is advantageous to instruct the encoder to write a PADDING block of the correct size, so that instead of rewriting the whole stream after encoding, the program can just overwrite the PADDING block.  If only the maximum size of the metadata is known, the program can write a slightly larger padding block, then split it after encoding.
+       </P>
+       <P>
+               Make sure you understand how lengths are calculated.  All FLAC metadata blocks have a 4 byte header which contains the type and length.  This length does not include the 4 bytes of the header.  See the <A HREF="format.html#metadata_block">format page</A> for the specification of metadata blocks and their lengths.
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <A NAME="libflacpp"><B><FONT SIZE="+2">libFLAC++</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               <B><TT>libFLAC++</TT></B> is a C++ object wrapper around <B><TT>libFLAC</TT></B>.  It provides classed for the encoders and decoders as well as the metadata interface.
+       </P>
+       <P>
+               The documentation for <B><TT>libFLAC++</TT></B> is currently being rewritten.  As a wrapper it is actually quite simple.  The method names and semantics generally follow those in the C layer and comments in the header files specify where there are differences.
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <A NAME="bugs"><B><FONT SIZE="+2">known bugs</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               Bug tracking is done on the Sourceforge project page <A HREF="http://sourceforge.net/bugs/?group_id=13478">here</A>.  If you submit a bug, make sure and provide an email contact or use the Monitor feature.
+       </P>
+       <P>
+               The following are major known bugs in the current (1.0.3) release:
+       </P>
+       <P>
+               <UL>
+                       <LI>
+                               The Winamp2 plugin has a bug where it shows a dialog "ERROR: invalid/missing FLAC metadata" for any .flac file.  This has been fixed; you can download the plugin <A HREF="http://sourceforge.net/project/shownotes.php?release_id=98266">here</A>.  The fixed source code is in CVS.
+                       </LI>
+               </UL>
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <A NAME="monkey"><B><FONT SIZE="+2">monkey</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               Monkey's Audio comes with a nice GUI that many people are familiar with.  It supports some external encoders, but not FLAC.  However, the FLAC Windows distribution comes with a utility that allows you to replace one the of the supported lossless external codecs with FLAC.  Here's how:
+               <UL>
+                       <LI>Copy <B><TT>flac.exe</TT></B> and <B><TT>flac_ren.exe</TT></B> to the <B><TT>External/</TT></B> directory of the Monkey's Audio installation.</LI>
+                       <LI>
+                               Choose a supported encoder to replace:
+                               <UL>
+                                       <LI>Shorten - copy <B><TT>flac_mac.exe</TT></B> on top of <B><TT>External/shortn32.exe</TT></B></LI>
+                                       <LI>WavPack - copy <B><TT>flac_mac.exe</TT></B> on top of both <B><TT>External/wavpack.exe</TT></B> and <B><TT>External/wvunpack.exe</TT></B></LI>
+                                       <LI>RKAU - copy <B><TT>flac_mac.exe</TT></B> on top of <B><TT>External/rkau.exe</TT></B></LI>
+                               </UL>
+                               If you choose WavPack you will also be able to use the WavPack Configuration menu to add flac options.
+                       </LI>
+                       <LI>Now you can encode FLAC files as if you were using the replaced encoder.  The renamed <B><TT>flac_mac.exe</TT></B> utility will call <B><TT>flac.exe</TT></B> and afterwards, <B><TT>flac_ren.exe</TT></B> will rename the resulting file to have the .flac extension.</LI>
+               </UL>
+       </P>
+       <P>
+               Other front-ends may be wedged in the same way; if you have one in mind, post it to the <A HREF="http://lists.sourceforge.net/lists/listinfo/flac-dev">flac-dev</A> mailing list.
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+</CENTER>
+
+<P>&nbsp;Copyright (c) 2000,2001,2002 Josh Coalson</P>
+
+</BODY>
+</HTML>
diff --git a/doc/html/download.html b/doc/html/download.html
new file mode 100644 (file)
index 0000000..e8c17e4
--- /dev/null
@@ -0,0 +1,141 @@
+<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
+<!-- Copyright (c)  2000,2001,2002  Josh Coalson -->
+<!-- Permission is granted to copy, distribute and/or modify this document -->
+<!-- under the terms of the GNU Free Documentation License, Version 1.1 -->
+<!-- or any later version published by the Free Software Foundation; -->
+<!-- with no invariant sections. -->
+<!-- A copy of the license can be found at http://www.gnu.org/copyleft/fdl.html -->
+<HTML>
+<HEAD>
+       <TITLE>FLAC - download</TITLE>
+</HEAD>
+
+<BODY MARGINWIDTH="0" MARGINHEIGHT="0" LEFTMARGIN="0" RIGHTMARGIN="0" TOPMARGIN="0" BGCOLOR="#99CC99" TEXT="#000000" LINK="#336699" VLINK="#336699" ALINK="#336699">
+
+<TABLE BORDER=0 WIDTH="100%" CELLPADDING=1 CELLSPACING=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#000000"><A HREF="http://flac.sourceforge.net/"><IMG SRC="images/logo.jpg" ALIGN=CENTER ALT="FLAC Logo" BORDER=0 HSPACE=0></a></TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="25" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#D3D4C5">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="index.html">home</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="news.html">news</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;download&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="features.html">features</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="goals.html">goals</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="format.html">format</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="id.html">id</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="comparison.html">comparison</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="documentation.html">documentation</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="developers.html">developers</A>&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#EEEED4">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;english&nbsp;&nbsp;</TD><TD BGCOLOR="#EEEED4" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="ru/download.html">russian</A>&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<CENTER>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="15" ALT=""></TD></TR></TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <B><FONT SIZE="+2">download</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               Currently all releases are made through SourceForge and can be found <A HREF="http://sourceforge.net/project/showfiles.php?group_id=13478">here</A>.  For each version there is a source release and binary releases for Linux, Windows, Solaris, and Darwin (includes OS X).
+       </P>
+       <P>
+               Debian packages can be found <A HREF="http://packages.debian.org/cgi-bin/search_packages.pl?keywords=flac+libflac+1.0&searchon=all&subword=1&version=all&release=all">here</A>.
+       </P>
+       <P>
+               You can also get the nightly CVS tarball <A HREF="http://cvs.sourceforge.net/cvstarballs/flac-cvsroot.tar.gz">here</A>.
+       </P>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <B><FONT SIZE="+2">extras</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               <B>GUI front-ends:</B>
+       <P>
+       <UL>
+               <LI>
+                       <B>Windows</B>
+                       <UL>
+                               <LI><A HREF="http://home.wanadoo.nl/~w.speek/flac.htm">FLAC frontend</A>, a Windows GUI.</LI>
+                               <LI><A HREF="http://www.inf.ufpr.br/~rja00/lossless.html">FLACdrop</A>, an Oggdrop-like frontend for Windows.</LI>
+                               <LI><A HREF="http://www.monkeysaudio.com/">Monkey's Audio</A> can be fooled into supporting FLAC as an external encoder; <A HREF="documentation.html#monkey">here's how</A>.</LI>
+                       </UL>
+               </LI>
+               <LI>
+                       <B>Unix</B>
+                       <UL>
+                               <LI><A HREF="http://www.nostatic.org/grip/">Grip</A> is a great ripping and encoding front end and can be easily configured to use flac.</LI>
+                       </UL>
+               </LI>
+       </UL>
+       </P>
+
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+</CENTER>
+
+<P>&nbsp;Copyright (c) 2000,2001,2002 Josh Coalson</P>
+
+</BODY>
+</HTML>
diff --git a/doc/html/features.html b/doc/html/features.html
new file mode 100644 (file)
index 0000000..ce451f1
--- /dev/null
@@ -0,0 +1,171 @@
+<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
+<!-- Copyright (c)  2000,2001,2002  Josh Coalson -->
+<!-- Permission is granted to copy, distribute and/or modify this document -->
+<!-- under the terms of the GNU Free Documentation License, Version 1.1 -->
+<!-- or any later version published by the Free Software Foundation; -->
+<!-- with no invariant sections. -->
+<!-- A copy of the license can be found at http://www.gnu.org/copyleft/fdl.html -->
+<HTML>
+<HEAD>
+       <TITLE>FLAC - features</TITLE>
+</HEAD>
+
+<BODY MARGINWIDTH="0" MARGINHEIGHT="0" LEFTMARGIN="0" RIGHTMARGIN="0" TOPMARGIN="0" BGCOLOR="#99CC99" TEXT="#000000" LINK="#336699" VLINK="#336699" ALINK="#336699">
+
+<TABLE BORDER=0 WIDTH="100%" CELLPADDING=1 CELLSPACING=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#000000"><A HREF="http://flac.sourceforge.net/"><IMG SRC="images/logo.jpg" ALIGN=CENTER ALT="FLAC Logo" BORDER=0 HSPACE=0></a></TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="25" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#D3D4C5">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="index.html">home</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="news.html">news</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="download.html">download</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;features&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="goals.html">goals</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="format.html">format</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="id.html">id</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="comparison.html">comparison</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="documentation.html">documentation</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="developers.html">developers</A>&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#EEEED4">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;english&nbsp;&nbsp;</TD><TD BGCOLOR="#EEEED4" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="ru/features.html">russian</A>&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<CENTER>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="15" ALT=""></TD></TR></TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <B><FONT SIZE="+2">features</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+       FLAC stands for Free Lossless Audio Codec.  The FLAC project consists of:
+       </P>
+       <P>
+       <UL>
+               <LI>the stream format</LI>
+               <LI><B><TT>libFLAC</TT></B>, a library of reference encoders and decoders, and a metadata interface</LI>
+               <LI><B><TT>libFLAC++</TT></B>, an object wrapper around libFLAC</LI>
+               <LI><B><TT>flac</TT></B>, a command-line wrapper around libFLAC to encode and decode .flac files</LI>
+               <LI><B><TT>metaflac</TT></B>, a command-line metadata editor for .flac files</LI>
+               <LI>input plugins for various music players (Winamp, XMMS, and more in the works)</LI>
+       </UL>
+       </P>
+       <P>
+               "Free" means that the specification of the stream format is in the public domain (the FLAC project reserves the right to set the FLAC specification and certify compliance), and that neither the FLAC format nor any of the implemented encoding/decoding methods are covered by any patent.  It also means that the sources for <B><TT>libFLAC</TT></B> and <B><TT>libFLAC++</TT></B> are available under the <A HREF="http://www.opensource.org/licenses/lgpl-license.html">LGPL</A> and the sources for <B><TT>flac</TT></B>, <B><TT>metaflac</TT></B>, and the plugins are available under the <A HREF="http://www.opensource.org/licenses/gpl-license.html">GPL</A>.
+       </P>
+       <P>
+               FLAC compiles on many platforms: most Unixes (Linux, *BSD, Solaris, OS X), Windows, BeOS, and OS/2.  There are build systems for autoconf/automake, MSVC, Watcom C, and Project Builder.
+       </P>
+       <P>
+               What FLAC is:
+       </P>
+       <UL>
+               <P><LI>
+                       FLAC is patent free.  The FLAC format or encoding/decoding methods are not covered by any patents.
+               </LI></P>
+               <P><LI>
+                       FLAC is lossless.  The encoding of PCM data incurs no loss of information, and the decoded audio is bit-for-bit identical to what went into the encoder.  Each frame contains a 16-bit CRC of the frame data for detecting transmission errors.  The integrity of the audio data is further insured by storing an <A HREF="http://userpages.umbc.edu/~mabzug1/cs/md5/md5.html">MD5 signature</A> of the original unencoded audio data in the file header, which can be compared against later during decoding or testing.
+               </LI></P>
+               <P><LI>
+                       FLAC is designed to compress audio data.  Technically, <B><TT>flac</TT></B> can "compress" other kinds of data losslessly (if you pass it in as a mono 8-bit raw file), but the output files tend to be bigger.
+               </LI></P>
+               <P><LI>
+                       The compression capabilities of FLAC are extendable, meaning that new methods can be added to future versions of the format without breaking older streams or decoders.
+               </LI></P>
+               <P><LI>
+                       The currently implemented compression methods in the reference encoder yield streams smaller than shorten.  The encoding time is variable, but is generally between that of <A HREF="http://www.softsound.com/Shorten.html">shorten</A>, and that of, say, <A HREF="http://www.mp3dev.org/mp3/">LAME</A>.  The most aggressive compression however can be quite slow.  For more info see the <A HREF="comparison.html">comparison page</A>.
+               </LI></P>
+               <P><LI>
+                       FLAC is asymmetric in favor of decode speed.  Decoding requires only integer arithmetic, and is much less compute-intensive than for most perceptual codecs.  Real-time decode performance is easily achievable on even modest hardare.
+               </LI></P>
+               <P><LI>
+                       FLAC is suitable for archiving, since there is no information loss.  You are not locked into the format since there is no generation loss if you decide to convert your data to another format in the future.  In addition to the frame CRCs and MD5 signature, <B><TT>flac</TT></B> has a verify option that decodes the encoded stream in parallel with the encoding process and compares the result to the original, aborting with an error if there is a mismatch.
+               </LI></P>
+               <P><LI>
+                       FLAC is suitable for streaming.  Each FLAC frame contains enough data to decode that frame.  FLAC does not even rely on previous or following frames.  FLAC uses sync codes and CRCs (similar to MPEG and other formats), which, along with framing, allow decoders to pick up in the middle of a stream with a minimum of delay.
+               </LI></P>
+               <P><LI>
+                       FLAC supports fast sample-accurate seeking.  Not only is this useful for playback, it makes FLAC files suitable for use in editing applications.
+               </LI></P>
+               <P><LI>
+                       FLAC has an extendable metadata system.  New metadata blocks can be defined and implemented in future versions of FLAC without breaking older streams or decoders.  Applications can write their own <A HREF="format.html#def_APPLICATION">APPLICATION</A> metadata once they <A HREF="id.html">register an ID</A>.  ID3 and ID3V2 tags may be attached to .flac files without disrupting the decoder.
+               </LI></P>
+       </UL>
+       <P>
+               Some things that follow from the features:
+       </P>
+       <UL>
+               <P><LI>
+                       FLAC streams can be played back consecutively with no audible gaps in between, unlike say, MP3s (this is one of the minor <A HREF="goals.html">goals</A>).  For example, you can encode a live album as individual tracks and still play them back seamlessly.
+               </LI></P>
+               <P><LI>
+                       The sample-accurate seeking allows versatile playback: a sophisticated player could do index points, complex looping, or other structured playback.  This could be useful in for say DJs, or practice sessions where you want to play along through specific passages.
+               </LI></P>
+               <P><LI>
+                       Basically, you get the versatility of a WAV file in a compressed streamable format.
+               </LI></P>
+       </UL>
+       <P>
+               What FLAC is <B>not</B>:
+       </P>
+       <UL>
+               <P><LI>
+                       Lossy.  FLAC is intended for lossless compression only, as there are many good lossy formats already, such as <A HREF="http://www.mp3-tech.org/">MP3</A> (see <A HREF="http://www.mp3dev.org/mp3/">LAME</A> for an excellent open-source implementation), and <A HREF="http://www.xiph.org/ogg/vorbis/index.html">Ogg Vorbis</A>.
+               </LI></P>
+               <P><LI>
+                       SDMI compliant, et cetera.  There is no intention to support any methods of copy protection, which are, for all practical purposes, a complete waste of bits.  (Another way to look at it is that since copy protection is futile, it really carries no information, so you might say FLAC already losslessly compresses all possible copy protection information down to zero bits!)  Of course, we can't stop what some misguided person does with proprietary meta-data blocks, but then again, non-proprietary decoders will skip them anyway.
+               </LI></P>
+       </UL>
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+</TD></TR>
+</TABLE>
+
+
+</CENTER>
+
+<P>&nbsp;Copyright (c) 2000,2001,2002 Josh Coalson</P>
+
+</BODY>
+</HTML>
diff --git a/doc/html/format.html b/doc/html/format.html
new file mode 100644 (file)
index 0000000..4d0b21f
--- /dev/null
@@ -0,0 +1,1331 @@
+<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
+<!-- Copyright (c)  2000,2001,2002  Josh Coalson -->
+<!-- Permission is granted to copy, distribute and/or modify this document -->
+<!-- under the terms of the GNU Free Documentation License, Version 1.1 -->
+<!-- or any later version published by the Free Software Foundation; -->
+<!-- with no invariant sections. -->
+<!-- A copy of the license can be found at http://www.gnu.org/copyleft/fdl.html -->
+<HTML>
+<HEAD>
+       <TITLE>FLAC - format</TITLE>
+</HEAD>
+
+<BODY MARGINWIDTH="0" MARGINHEIGHT="0" LEFTMARGIN="0" RIGHTMARGIN="0" TOPMARGIN="0" BGCOLOR="#99CC99" TEXT="#000000" LINK="#336699" VLINK="#336699" ALINK="#336699">
+
+<TABLE BORDER=0 WIDTH="100%" CELLPADDING=1 CELLSPACING=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#000000"><A HREF="http://flac.sourceforge.net/"><IMG SRC="images/logo.jpg" ALIGN=CENTER ALT="FLAC Logo" BORDER=0 HSPACE=0></a></TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="25" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#D3D4C5">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="index.html">home</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="news.html">news</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="download.html">download</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="features.html">features</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="goals.html">goals</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;format&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="id.html">id</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="comparison.html">comparison</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="documentation.html">documentation</A>&nbsp;&nbsp;</TD><TD BGCOLOR="#D3D4C5" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#D3D4C5" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="developers.html">developers</A>&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING=0 CELLSPACING=0 BORDER=0>
+       <TR>
+               <TD ALIGN="CENTER" BGCOLOR="#EEEED4">
+                       <TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0>
+                               <TR>
+                                       <TD HEIGHT=22 BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;english&nbsp;&nbsp;</TD><TD BGCOLOR="#EEEED4" ALIGN=CENTER>|</TD>
+                                       <TD           BGCOLOR="#EEEED4" ALIGN=CENTER NOWRAP>&nbsp;&nbsp;<A CLASS="topnav" HREF="ru/format.html">russian</A>&nbsp;&nbsp;</TD>
+                               </TR>
+                       </TABLE>
+               </TD>
+       </TR>
+</TABLE>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="2" ALT=""></TD></TR></TABLE>
+
+<CENTER>
+
+<TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#99CC99"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="15" ALT=""></TD></TR></TABLE>
+
+
+<TABLE WIDTH="100%" CELLPADDING="5" CELLSPACING="5" BORDER="0">
+<TR><TD>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#D3D4C5">
+               <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+               <B><FONT SIZE="+2">format</FONT></B>
+               </FONT></TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+       <TABLE CELLSPACING="0" CELLPADDING="3" WIDTH="100%" BORDER="0" BGCOLOR="#EEEED4">
+       <TR><TD><FONT FACE="Lucida,Verdana,Helvetica,Arial">
+       <P>
+               This is a detailed description of the FLAC format.
+       </P>
+       <P>
+               First, as the original developer I have to say that I am not a compression expert and I feel obligated to give credit where it is due.  FLAC owes a lot to the many people who have advanced the audio compression field so freely.  For instance:
+       </P>
+       <P>
+       <UL>
+               <LI>
+                       <A HREF="http://svr-www.eng.cam.ac.uk/~ajr/">A. J. Robinson</A> for his work on <A HREF="http://svr-www.eng.cam.ac.uk/~ajr/GroupPubs/Robinson94-tr156/index.html">Shorten</A>; his code and paper are a good starting point on some of the basic methods used by FLAC.  FLAC expands on the fixed predictors used in shorten.
+               </LI>
+               <LI>
+                       <A HREF="http://commsci.usc.edu/faculty/golomb.html">S. W. Golomb</A> and Robert F. Rice; their universal codes are used by FLAC's entropy coder.
+               </LI>
+               <LI>
+                       N. Levinson and J. Durbin; the reference encoder uses an algorithm developed and refined by them for determining the LPC coefficients from the autocorrelation coefficients.
+               </LI>
+               <LI>
+                       And of course, the main guy, <A HREF="http://www.digitalcentury.com/encyclo/update/shannon.html">Claude Shannon</A>
+               </LI>
+       </UL>
+       </P>
+       <P>
+               <A NAME="scope"><FONT SIZE="+1"><B><U>Scope</U></B></FONT></A>
+       </P>
+       <P>
+               It is a known fact that no algorithm can losslessly compress all possible input, so most compressors restrict themselves to a useful domain and try to work as well as possible within that domain.  FLAC's domain is audio data.  Though it can losslessly <B>code</B> any input, only certain kinds of input will get smaller.  FLAC exploits the fact that audio data typically has a high degree of sample-to-sample correlation.
+       </P>
+       <P>
+               Within the audio domain, there are many possible subdomains.  For example: low bitrate speech, high-bitrate multi-channel music, etc.  FLAC itself does not target a specific subdomain but many of the default parameters of the reference encoder are tuned to CD-quality music data (i.e. 44.1kHz, 2 channel, 16 bits per sample).  The effect of the encoding parameters on different kinds of audio data will be examined later.
+       </P>
+       <P>
+               <A NAME="architecture"><FONT SIZE="+1"><B><U>Architecture</U></B></FONT></A>
+       </P>
+       <P>
+               Similar to many audio coders, a FLAC encoder has the following stages:
+       </P>
+       <UL>
+               <P><LI>
+                       <A HREF="#blocking">Blocking</A>.  The input is broken up into many contiguous blocks.  With FLAC, the blocks may vary in size.  The optimal size of the block is usually affected by many factors, including the sample rate, spectral characteristics over time, etc.  Though FLAC allows the block size to vary within a stream, the reference encoder uses a fixed block size.
+               </LI></P>
+               <P><LI>
+                       <A HREF="#interchannel">Interchannel Decorrelation</A>.  In the case of stereo streams, the encoder will create mid and side signals based on the average and difference (respectively) of the left and right channels.  The encoder will then pass the best form of the signal to the next stage.
+               </LI></P>
+               <P><LI>
+                       <A HREF="#prediction">Prediction</A>.  The block is passed through a prediction stage where the encoder tries to find a mathematical description (usually an approximate one) of the signal.  This description is typically much smaller than the raw signal itself.  Since the methods of prediction are known to both the encoder and decoder, only the parameters of the predictor need be included in the compressed stream.  FLAC currently uses four different classes of predictors (described in the <A HREF="#prediction">prediction</A> section), but the format has reserved space for additional methods.  FLAC allows the class of predictor to change from block to block, or even within the channels of a block.
+               </LI></P>
+               <LI><P>
+                       <A HREF="#residualcoding">Residual coding</A>.  If the predictor does not describe the signal exactly, the difference between the original signal and the predicted signal (called the error or residual signal) must be coded losslessy.  If the predictor is effective, the residual signal will require fewer bits per sample than the original signal.  FLAC currently uses only one method for encoding the residual (see the <A HREF="#residualcoding">Residual coding</A> section), but the format has reserved space for additional methods.  FLAC allows the residual coding method to change from block to block, or even within the channels of a block.
+               </LI></P>
+       </UL>
+       <P>
+               In addition, FLAC specifies a metadata system, which allows arbitrary information about the stream to be included at the beginning of the stream.
+       </P>
+       <P>
+               <A NAME="definitions"><FONT SIZE="+1"><B><U>Definitions</U></B></FONT></A>
+       </P>
+       <P>
+               Many terms like "block" and "frame" are used to mean different things in differenct encoding schemes.  For example, a frame in MP3 corresponds to many samples across several channels, whereas an S/PDIF frame represents just one sample for each channel.  The definitions we use for FLAC follow.  Note that when we talk about blocks and subblocks we are refering to the raw unencoded audio data that is the input to the encoder, and when we talk about frames and subframes, we are refering to the FLAC-encoded data.
+       </P>
+       <UL>
+               <P><LI>
+                       <B>Block</B>: One or more audio samples that span several channels.
+               </LI></P>
+               <P><LI>
+                       <B>Subblock</B>: One or more audio samples within a channel.  So a block contains one subblock for each channel, and all subblocks contain the same number of samples.
+               </LI></P>
+               <P><LI>
+                       <B>Blocksize</B>: The number of samples in any of a block's subblocks.  For example, a one second block sampled at 44.1KHz has a blocksize of 44100, regardless of the number of channels.
+               </LI></P>
+               <P><LI>
+                       <B>Frame</B>: A frame header plus one or more subframes.
+               </LI></P>
+               <P><LI>
+                       <B>Subframe</B>: A subframe header plus one or more encoded samples from a given channel.  All subframes within a frame will contain the same number of samples.
+               </LI></P>
+       </UL>
+       <P>
+               <A NAME="blocking"><FONT SIZE="+1"><B><U>Blocking</U></B></FONT></A>
+       </P>
+       <P>
+               The size used for blocking the audio data has a direct effect on the compression ratio.  If the block size is too small, the resulting large number of frames mean that excess bits will be wasted on frame headers.  If the block size is too large, the characteristics of the signal may vary so much that the encoder will be unable to find a good predictor.  In order to simplify encoder/decoder design, FLAC imposes a minimum block size of 16 samples, and a maximum block size of 65535 samples.  This range covers the optimal size for all of the audio data FLAC supports.
+       </P>
+       <P>
+               Currently the reference encoder uses a fixed block size, optimized on the sample rate of the input.  Future versions may vary the block size depending on the characteristics of the signal.
+       </P>
+       <P>
+               Blocked data is passed to the predictor stage one subblock (channel) at a time.  Each subblock is independently coded into a subframe, and the subframes are concatenated into a frame.  Because each channel is coded separately, it means that one channel of a stereo frame may be encoded as a constant subframe, and the other an LPC subframe.
+       </P>
+       <P>
+               <A NAME="interchannel"><FONT SIZE="+1"><B><U>Interchannel Decorrelation</U></B></FONT></A>
+       </P>
+       <P>
+               In stereo streams, in many cases there is an exploitable amount of correlation between the left and right channels.  FLAC allows the frames of stereo streams to have different channel assignments, and an encoder may choose to use the best representation on a frame-by-frame basis.
+       </P>
+       <UL>
+               <P><LI>
+                       <B>Independent</B>.  The left and right channels are coded independently.
+               </LI></P>
+               <P><LI>
+                       <B>Mid-side</B>.  The left and right channels are transformed into mid and side channels.  The mid channel is the midpoint (average) of the left and right signals, and the side is the difference signal (left minus right).
+               </LI></P>
+               <P><LI>
+                       <B>Left-side</B>.  The left channel and side channel are coded.
+               </LI></P>
+               <P><LI>
+                       <B>Right-side</B>.  The right channel and side channel are coded
+               </LI></P>
+       </UL>
+       <P>
+               Surprisingly, the left-side and right-side forms can be the most efficient in many frames, even though the raw number of bits per sample needed for the original signal is slightly more than that needed for independent or mid-side coding.
+       </P>
+       <P>
+               <A NAME="prediction"><FONT SIZE="+1"><B><U>Prediction</U></B></FONT></A>
+       </P>
+       <P>
+               FLAC uses four methods for modeling the input signal:
+       </P>
+       <UL>
+               <P><LI>
+                       <B>Verbatim</B>.  This is essentially a zero-order predictor of the signal.  The predicted signal is zero, meaning the residual is the signal itself, and the compression is zero.  This is the baseline against which the other predictors are measured.  If you feed random data to the encoder, the verbatim predictor will probably be used for every subblock.  Since the raw signal is not actually passed through the residual coding stage (it is added to the stream 'verbatim'), the encoding results will not be the same as a zero-order linear predictor.
+               </LI></P>
+               <P><LI>
+                       <B>Constant</B>.  This predictor is used whenever the subblock is pure DC ("digital silence"), i.e. a constant value throughout.  The signal is run-length encoded and added to the stream.
+               </LI></P>
+               <P><LI>
+                       <B>Fixed linear predictor</B>.  FLAC uses a class of computationally-efficient fixed linear predictors (for a good description, see <A HREF="http://www.hpl.hp.com/techreports/1999/HPL-1999-144.pdf">audiopak</A> and <A HREF="http://svr-www.eng.cam.ac.uk/~ajr/GroupPubs/Robinson94-tr156/index.html">shorten</A>).  FLAC adds a fourth-order predictor to the zero-to-third-order predictors used by Shorten.  Since the predictors are fixed, the predictor order is the only parameter that needs to be stored in the compressed stream.  The error signal is then passed to the residual coder.
+               </LI></P>
+               <P><LI>
+                       <B>FIR Linear prediction</B>.  For more accurate modeling (at a cost of slower encoding), FLAC supports up to 32nd order FIR linear prediction (again, for info on linear prediction, see <A HREF="http://www.hpl.hp.com/techreports/1999/HPL-1999-144.pdf">audiopak</A> and <A HREF="http://svr-www.eng.cam.ac.uk/~ajr/GroupPubs/Robinson94-tr156/index.html">shorten</A>).  The reference encoder uses the Levinson-Durbin method for calculating the LPC coefficients from the autocorrelation coefficients, and the coefficients are quantized before computing the residual.  Whereas encoders such as Shorten used a fixed quantization for the entire input, FLAC allows the quantized coefficient precision to vary from subframe to subframe.  The FLAC reference encoder estimates the optimal precision to use based on the block size and dynamic range of the original signal.
+               </LI></P>
+       </UL>
+       <P>
+               <A NAME="residualcoding"><FONT SIZE="+1"><B><U>Residual Coding</U></B></FONT></A>
+       </P>
+       <P>
+               FLAC currently defines two similar methods for the coding of the error signal from the prediction stage.  The error signal is coded using Rice codes in one of two ways: 1) the encoder estimates a single rice parameter based on the variance of the residual and Rice codes the entire residual using this parameter; 2) the residual is partitioned into several equal-length regions of contiguous samples, and each region is coded with its own Rice parameter based on the region's mean.  (Note that the first method is a special case of the second method with one partition, except the Rice parameter is based on the residual variance instead of the mean.)
+       </P>
+       <P>
+               The FLAC format has reserved space for other coding methods.  Some possiblities for volunteers would be to explore better context-modeling of the Rice parameter, or Huffman coding.  See <A HREF="http://www.hpl.hp.com/techreports/98/HPL-98-193.html">LOCO-I</A> and <A HREF="http://www.cs.tut.fi/~albert/Dev/pucrunch/packing.html">pucrunch</A> for descriptions of several universal codes.
+       </P>
+       <P>
+               <FONT SIZE="+1"><B><U>Format</U></B></FONT>
+       </P>
+       <P>
+               This section specifies the FLAC bitstream format.  FLAC has no format version information, but it does contain reserved space in several places.  Future versions of the format may use this reserved space safely without breaking the format of older streams.  Older decoders may choose to abort decoding or skip data encoded with newer methods.  Apart from reserved patterns, in places the format specifies invalid patterns, meaning that the patterns may never appear in any valid bitstream, in any prior, present, or future versions of the format.  These invalid patterns are usually used to make the synchronization mechanism more robust.
+       </P>
+       <P>
+               All numbers used in a FLAC bitstream are integers; there are no floating-point representations.  All numbers are big-endian coded.  All numbers are unsigned unless otherwise specified.
+       </P>
+       <P>
+               <A NAME="overview">A FLAC bitstream may be appended with ID3V1 data or prepended with ID3V2 data.  FLAC has no knowledge of such data, but the reference decoder knows how to skip an ID3 tag.</A>
+       </P>
+       <P>
+               Before the formal description of the stream, an overview might be helpful.
+       </P>
+       <UL>
+               <P><LI>
+                       A FLAC bitstream consists of the "fLaC" marker at the beginning of the stream, followed by a mandatory metadata block (called the STREAMINFO block), any number of other metadata blocks, then the audio frames.
+               </LI></P>
+               <P><LI>
+                       FLAC supports up to 128 kinds of metadata blocks; currently the following are defined:
+                       <UL>
+                               <LI><A NAME="def_STREAMINFO"><B>STREAMINFO</B>: This block has information about the whole stream, like sample rate, number of channels, total number of samples, etc.  It must be present as the first metadata block in the stream.  Other metadata blocks may follow, and ones that the decoder doesn't understand, it will skip.</LI>
+                               <LI><A NAME="def_APPLICATION"><B>APPLICATION</B>: This block is for use by third-party applications.  The only mandatory field is a 32-bit identifier.  This ID is granted upon request to an application by the FLAC maintainers.  The remainder is of the block is defined by the registered application.  Visit the <A HREF="id.html">registration page</A> if you would like to register an ID for your application with FLAC.</LI>
+                               <LI><A NAME="def_PADDING"><B>PADDING</B>: This block allows for an arbitrary amount of padding.  The contents of a PADDING block have no meaning.  This block is useful when it is known that an APPLICATION block will be added after encoding; the user can instruct the encoder to reserve a PADDING block of the proper size so that the application may directly write over it later (which is relatively quick) instead of having to insert the APPLICATION block (which would normally require rewriting the entire file).</LI>
+                               <LI><A NAME="def_SEEKTABLE"><B>SEEKTABLE</B>: This is an optional block for storing seek points.  It is possible to seek to any given sample in a FLAC stream without a seek table, but the delay can be unpredictable since the bitrate may vary widely within a stream.  By adding seek points to a stream, this delay can be significantly reduced.  Each seek point takes 18 bytes, so 1% resolution within a stream adds less than 2k.  There can be only one SEEKTABLE in a stream, but the table can have any number of seek points.  There is also a special 'placeholder' seekpoint which will be ignored by decoders but which can be used to reserve space for future seek point insertion.</LI>
+                       </UL>
+               </LI></P>
+               <P><LI>
+                       The audio data is composed of one or more audio frames.  Each frame consists of a frame header, which contains a sync code, info about the frame like the block size, sample rate, number of channels, et cetera, and an 8-bit CRC.  The frame header also contains either the sample number of the first sample in the frame (for variable-blocksize streams), or the frame number (for fixed-blocksize streams).  This allows for fast, sample-accurate seeking to be performed.  Following the frame header are encoded subframes, one for each channel, and finally, the frame is zero-padded to a byte boundary.  Each subframe has its own header that specifies how the subframe is encoded.
+               </LI></P>
+               <P><LI>
+                       Since a decoder may start decoding in the middle of a stream, there must be a method to determine the start of a frame.  A 14-bit sync code begins each frame.  The sync code will not appear anywhere else in the frame header.  However, since it may appear in the subframes, the decoder has two other ways of ensuring a correct sync.  The first is to check that the rest of the frame header contains no invalid data.  Even this is not foolproof since valid header patterns can still occur within the subframes.  The decoder's final check is to generate an 8-bit CRC of the frame header and compare this to the CRC stored at the end of the frame header.
+               </LI></P>
+               <P><LI>
+                       Again, since a decoder may start decoding at an arbitrary frame in the stream, each frame header must contain some basic information about the stream because the decoder may not have access to the STREAMINFO metadata block at the start of the stream.  This information includes sample rate, bits per sample, number of channels, etc.  Since the frame header is pure overhead, it has a direct effect on the compression ratio.  To keep the frame header as small as possible, FLAC uses lookup tables for the most commonly used values for frame parameters.  For instance, the sample rate part of the frame header is specified using 4 bits.  Eight of the bit patterns correspond to the commonly used sample rates of 8/16/22.05/24/32/44.1/48/96 kHz.  However, odd sample rates can be specified by using one of the 'hint' bit patterns, directing the decoder to find the exact sample rate at the end of the frame header.  The same method is used for specifying the block size and bits per sample.  In this way, the frame header size stays small for all of the most common forms of audio data.
+               </LI></P>
+               <P><LI>
+                       Individual subframes (one for each channel) are coded separately within a frame, and appear serially in the stream.  In other words, the encoded audio data is NOT channel-interleaved.  This reduces decoder complexity at the cost of requiring larger decode buffers.  Each subframe has its own header specifying the attributes of the subframe, like prediction method and order, residual coding parameters, etc.  The header is followed by the encoded audio data for that channel.
+               </LI></P>
+               <P><LI>
+                       <A NAME="subset">FLAC specifies a subset of itself as the Subset format.  The purpose of this is to ensure that any streams encoded according to the Subset are truly "streamable", meaning that a decoder that cannot seek within the stream can still pick up in the middle of the stream and start decoding.  It also makes hardware decoder implementations more practical by limiting the blocking such that decoder buffer sizes can be easily determined.  "flac" generates Subset streams by default unless the "--lax" command-line option is used.  The Subset makes the following limitations on what may be used in the stream:
+                       <UL>
+                       <LI>
+                               The blocksize bits in the <A HREF="#frame_header">frame header</A> must be 0001-0101 or 1000-1111, specifying a fixed-blocksize stream (the exception being the last block as described in the table).  This also means that the STREAMINFO metadata block must specify equal mininum and maximum blocksizes.
+                       </LI>
+                       <LI>
+                               The bits-per-sample bits in the <A HREF="#frame_header">frame header</A> must be 001-110.
+                       </LI>
+                       <LI>
+                               The sample rate bits in the <A HREF="#frame_header">frame header</A> must be 0001-1011.
+                       </LI>
+                       </UL>
+               </LI></P>
+       </UL>
+
+       <P>
+               The following tables constitute a formal description of the FLAC format.  Numbers in angle brackets indicate how many bits are used for a given field.
+       </P>
+
+       </FONT>
+       </TD></TR>
+       </TABLE>
+       <TABLE WIDTH="100%" CELLPADDING="0" CELLSPACING="0" BORDER="0"><TR BGCOLOR="#000000"><TD><IMG SRC="images/1x1.gif" WIDTH="1" HEIGHT="1" ALT=""></TD></TR></TABLE>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="stream"><FONT SIZE="+1"><B>STREAM</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;32&gt;
+                       </TD>
+                       <TD>
+                               "fLaC", the FLAC stream marker in ASCII, meaning byte 0 of the stream is 0x66, followed by 0x4C 0x61 0x43
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#metadata_block_streaminfo"><I>METADATA_BLOCK</I></A>
+                       </TD>
+                       <TD>
+                               This is the mandatory STREAMINFO metadata block that has the basic properties of the stream
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#metadata_block"><I>METADATA_BLOCK</I></A>*
+                       </TD>
+                       <TD>
+                               Zero or more metadata blocks
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#frame"><I>FRAME</I></A>+
+                       </TD>
+                       <TD>
+                               One or more audio frames
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="metadata_block"><FONT SIZE="+1"><B>METADATA_BLOCK</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#metadata_block_header"><I>METADATA_BLOCK_HEADER</I></A>
+                       </TD>
+                       <TD>
+                               A block header that specifies the type and size of the metadata block data.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#metadata_block_data"><I>METADATA_BLOCK_DATA</I></A>
+                       </TD>
+                       <TD>
+                               &nbsp;
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="metadata_block_header"><FONT SIZE="+1"><B>METADATA_BLOCK_HEADER</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;1&gt;
+                       </TD>
+                       <TD>
+                               Last-metadata-block flag: '1' if this block is the last metadata block before the audio blocks, '0' otherwise.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;7&gt;
+                       </TD>
+                       <TD>
+                               BLOCK_TYPE<BR>
+                               <UL>
+                               <LI>
+                                       <TT>0</TT> : STREAMINFO
+                               </LI>
+                               <LI>
+                                       <TT>1</TT> : PADDING
+                               </LI>
+                               <LI>
+                                       <TT>2</TT> : APPLICATION
+                               </LI>
+                               <LI>
+                                       <TT>3</TT> : SEEKTABLE
+                               </LI>
+                               <LI>
+                                       <TT>4</TT> : VORBIS_COMMENT
+                               </LI>
+                               <LI>
+                                       <TT>5-127</TT> : reserved
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;24&gt;
+                       </TD>
+                       <TD>
+                               Length (in bytes) of metadata to follow (does not include the size of the METADATA_BLOCK_HEADER)
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="metadata_block_data"><FONT SIZE="+1"><B>METADATA_BLOCK_DATA</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#metadata_block_streaminfo"><I>METADATA_BLOCK_STREAMINFO</I></A><BR>|| <A HREF="#metadata_block_padding"><I>METADATA_BLOCK_PADDING</I></A><BR>|| <A HREF="#metadata_block_application"><I>METADATA_BLOCK_APPLICATION</I></A><BR>|| <A HREF="#metadata_block_seektable"><I>METADATA_BLOCK_SEEKTABLE</I></A><BR>|| <A HREF="#metadata_block_vorbis_comment"><I>METADATA_BLOCK_VORBIS_COMMENT</I></A>
+                       </TD>
+                       <TD>
+                               The block data must match the block type in the block header.
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="metadata_block_streaminfo"><FONT SIZE="+1"><B>METADATA_BLOCK_STREAMINFO</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;16&gt;
+                       </TD>
+                       <TD>
+                               The minimum block size (in samples) used in the stream.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;16&gt;
+                       </TD>
+                       <TD>
+                               The maximum block size (in samples) used in the stream.  (Minimum blocksize == maximum blocksize) implies a fixed-blocksize stream.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;24&gt;
+                       </TD>
+                       <TD>
+                               The minimum frame size (in bytes) used in the stream.  May be 0 to imply the value is not known.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;24&gt;
+                       </TD>
+                       <TD>
+                               The maximum frame size (in bytes) used in the stream.  May be 0 to imply the value is not known.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;20&gt;
+                       </TD>
+                       <TD>
+                               Sample rate in Hz.  Though 20 bits are available, the maximum sample rate is limited by the structure of frame headers to 1048570Hz.  Also, a value of 0 is invalid.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;3&gt;
+                       </TD>
+                       <TD>
+                               (number of channels)-1.  FLAC supports from 1 to 8 channels
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;5&gt;
+                       </TD>
+                       <TD>
+                               (bits per sample)-1.  FLAC supports from 4 to 32 bits per sample.  Currently the reference encoder and decoders only support up to 24 bits per sample.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;36&gt;
+                       </TD>
+                       <TD>
+                               Total samples in stream.  'Samples' means inter-channel sample, i.e. one second of 44.1Khz audio will have 44100 samples regardless of the number of channels.  A value of zero here means the number of total samples is unknown.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;128&gt;
+                       </TD>
+                       <TD>
+                               MD5 signature of the unencoded audio data.  This allows the decoder to determine if an error exists in the audio data even when the error does not result in an invalid bitstream.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1">NOTES</FONT><BR>
+                               <UL>
+                               <LI>
+                                       FLAC specifies a minimum block size of 16 and a maximum block size of 65535, meaning the bit patterns corresponding to the numbers 0-15 in the minimum blocksize and maximum blocksize fields are invalid.
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="metadata_block_padding"><FONT SIZE="+1"><B>METADATA_BLOCK_PADDING</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;n&gt;
+                       </TD>
+                       <TD>
+                               n '0' bits (n must be a multiple of 8)
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="metadata_block_application"><FONT SIZE="+1"><B>METADATA_BLOCK_APPLICATION</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;32&gt;
+                       </TD>
+                       <TD>
+                               Registered application ID.  (Visit the <A HREF="id.html">registration page</A> to register an ID with FLAC.)
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;n&gt;
+                       </TD>
+                       <TD>
+                               Application data (n must be a multiple of 8)
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="metadata_block_seektable"><FONT SIZE="+1"><B>METADATA_BLOCK_SEEKTABLE</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#seekpoint"><I>SEEKPOINT</I></A>+
+                       </TD>
+                       <TD>
+                               One or more seek points.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1">NOTE</FONT><BR>
+                               <UL>
+                               <LI>
+                                       The number of seek points is implied by the metadata header 'length' field, i.e. equal to length / 18.
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="seekpoint"><FONT SIZE="+1"><B>SEEKPOINT</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;64&gt;
+                       </TD>
+                       <TD>
+                               Sample number of first sample in the target frame, or 0xFFFFFFFFFFFFFFFF for a placeholder point.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;64&gt;
+                       </TD>
+                       <TD>
+                               Offset (in bytes) from the first byte of the first frame header to the first byte of the target frame's header.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;16&gt;
+                       </TD>
+                       <TD>
+                               Number of samples in the target frame.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1">NOTES</FONT><BR>
+                               <UL>
+                               <LI>
+                                       For placeholder points, the second and third field values are undefined.
+                               </LI>
+                               <LI>
+                                       Seek points within a table must be sorted in ascending order by sample number.
+                               </LI>
+                               <LI>
+                                       Seek points within a table must be unique by sample number, with the exception of placeholder points.
+                               </LI>
+                               <LI>
+                                       The previous two notes imply that there may be any number of placeholder points, but they must all occur at the end of the table.
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="metadata_block_vorbis_comment"><FONT SIZE="+1"><B>METADATA_BLOCK_VORBIS_COMMENT</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;n&gt;
+                       </TD>
+                       <TD>
+                               The contents of a vorbis comment packet as specified <A HREF="http://www.xiph.org/ogg/vorbis/doc/v-comment.html">here</A>, including the vendor string.  Note that the vorbis comment spec allows for on the order of 2 ^ 64 bytes of data where as the FLAC metadata block is limited to 2 ^ 24 bytes.  Given the stated purpose of vorbis comments, i.e. human-readable textual information, this limit is unlikely to be restrictive.  Also note that the 32-bit field lengths are little-endian coded according to the vorbis spec, as opposed to the usual big-endian coding of fixed-length integers in the rest of FLAC.
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="frame"><FONT SIZE="+1"><B>FRAME</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#frame_header"><I>FRAME_HEADER</I></A>
+                       </TD>
+                       <TD>
+                               &nbsp;
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#subframe"><I>SUBFRAME</I></A>+
+                       </TD>
+                       <TD>
+                               One SUBFRAME per channel.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;?&gt;
+                       </TD>
+                       <TD>
+                               Zero-padding to byte alignment.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#frame_footer"><I>FRAME_FOOTER</I></A>
+                       </TD>
+                       <TD>
+                               &nbsp;
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="frame_header"><FONT SIZE="+1"><B>FRAME_HEADER</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;14&gt;
+                       </TD>
+                       <TD>
+                               Sync code '<TT>11111111111110</TT>'
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;2&gt;
+                       </TD>
+                       <TD>
+                               Reserved:<BR>
+                               <UL>
+                               <LI>
+                                       <TT>00</TT> : mandatory value
+                               </LI>
+                               <LI>
+                                       <TT>01-11</TT> : reserved for future use
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;4&gt;
+                       </TD>
+                       <TD>
+                               Block size in inter-channel samples:<BR>
+                               <UL>
+                               <LI>
+                                       <TT>0000</TT> : get from STREAMINFO metadata block
+                               </LI>
+                               <LI>
+                                       <TT>0001</TT> : 192 samples
+                               </LI>
+                               <LI>
+                                       <TT>0010-0101</TT> : 576 * (2^(n-2)) samples, i.e. 576/1152/2304/4608
+                               </LI>
+                               <LI>
+                                       <TT>0110</TT> : get 8 bit (blocksize-1) from end of header
+                               </LI>
+                               <LI>
+                                       <TT>0111</TT> : get 16 bit (blocksize-1) from end of header
+                               </LI>
+                               <LI>
+                                       <TT>1000-1111</TT> : 256 * (2^(n-8)) samples, i.e. 256/512/1024/2048/4096/8192/16384/32768
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;4&gt;
+                       </TD>
+                       <TD>
+                               Sample rate:<BR>
+                               <UL>
+                               <LI>
+                                       <TT>0000</TT> : get from STREAMINFO metadata block
+                               </LI>
+                               <LI>
+                                       <TT>0001-0011</TT> : reserved
+                               </LI>
+                               <LI>
+                                       <TT>0100</TT> : 8kHz
+                               </LI>
+                               <LI>
+                                       <TT>0101</TT> : 16kHz
+                               </LI>
+                               <LI>
+                                       <TT>0110</TT> : 22.05kHz
+                               </LI>
+                               <LI>
+                                       <TT>0111</TT> : 24kHz
+                               </LI>
+                               <LI>
+                                       <TT>1000</TT> : 32kHz
+                               </LI>
+                               <LI>
+                                       <TT>1001</TT> : 44.1kHz
+                               </LI>
+                               <LI>
+                                       <TT>1010</TT> : 48kHz
+                               </LI>
+                               <LI>
+                                       <TT>1011</TT> : 96kHz
+                               </LI>
+                               <LI>
+                                       <TT>1100</TT> : get 8 bit sample rate (in kHz) from end of header
+                               </LI>
+                               <LI>
+                                       <TT>1101</TT> : get 16 bit sample rate (in Hz) from end of header
+                               </LI>
+                               <LI>
+                                       <TT>1110</TT> : get 16 bit sample rate (in tens of Hz) from end of header
+                               </LI>
+                               <LI>
+                                       <TT>1111</TT> : invalid, to prevent sync-fooling string of 1s
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;4&gt;
+                       </TD>
+                       <TD>
+                               Channel assignment
+                               <UL>
+                               <LI>
+                                       <TT>0000-0111</TT> : (number of independent channels)-1.  when == 0001, channel 0 is the left channel and channel 1 is the right
+                               </LI>
+                               <LI>
+                                       <TT>1000</TT> : left/side stereo: channel 0 is the left channel, channel 1 is the side(difference) channel
+                               </LI>
+                               <LI>
+                                       <TT>1001</TT> : right/side stereo: channel 0 is the side(difference) channel, channel 1 is the right channel
+                               </LI>
+                               <LI>
+                                       <TT>1010</TT> : mid/side stereo: channel 0 is the mid(average) channel, channel 1 is the side(difference) channel
+                               </LI>
+                               <LI>
+                                       <TT>1011-1111</TT> : reserved
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;3&gt;
+                       </TD>
+                       <TD>
+                               Sample size in bits:<BR>
+                               <UL>
+                               <LI>
+                                       <TT>000</TT> : get from STREAMINFO metadata block
+                               </LI>
+                               <LI>
+                                       <TT>001</TT> : 8 bits per sample
+                               </LI>
+                               <LI>
+                                       <TT>010</TT> : 12 bits per sample
+                               </LI>
+                               <LI>
+                                       <TT>011</TT> : reserved
+                               </LI>
+                               <LI>
+                                       <TT>100</TT> : 16 bits per sample
+                               </LI>
+                               <LI>
+                                       <TT>101</TT> : 20 bits per sample
+                               </LI>
+                               <LI>
+                                       <TT>110</TT> : 24 bits per sample
+                               </LI>
+                               <LI>
+                                       <TT>111</TT> : reserved
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;1&gt;
+                       </TD>
+                       <TD>
+                               Zero bit padding, to prevent sync-fooling string of 1s
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;?&gt;
+                       </TD>
+                       <TD>
+                               if(variable blocksize)<BR>
+                               &nbsp;&nbsp;&nbsp;&lt;8-56&gt;:"UTF-8" coded sample number (decoded number is 36 bits)<BR>
+                               else<BR>
+                               &nbsp;&nbsp;&nbsp;&lt;8-48&gt;:"UTF-8" coded frame number (decoded number is 31 bits)
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;?&gt;
+                       </TD>
+                       <TD>
+                               if(blocksize bits == 011x)<BR>
+                               &nbsp;&nbsp;&nbsp;8/16 bit (blocksize-1)
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;?&gt;
+                       </TD>
+                       <TD>
+                               if(sample rate bits == 11xx)<BR>
+                               &nbsp;&nbsp;&nbsp;8/16 bit sample rate
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;8&gt;
+                       </TD>
+                       <TD>
+                               CRC-8 (polynomial = x^8 + x^2 + x^1 + x^0, initialized with 0) of everything before the crc, including the sync code
+                       </TD>
+               </TR>
+               <TR>
+                       <TD>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1">NOTES</FONT><BR>
+                               <UL>
+                               <LI>
+                                       The blocksize bits 0000-0101 and 1000-1111 may only be used if the blocksize is fixed throughout the entire stream.  Blocksize bits 0110-0111 may be used in any case but the decoder will have to pessimistically guess that it is a variable-blocksize stream.  There is only one special case: the encoder may use blocksize bits 0110-0111 on the last frame of a fixed-blocksize stream, as long as the blocksize is not greater than the stream blocksize.
+                               </LI>
+                               <LI>
+                                       The "UTF-8" coding used for the sample/frame number is the same variable length code used to store compressed UCS-2, extended to handle larger input.
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="frame_footer"><FONT SIZE="+1"><B>FRAME_FOOTER</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;16&gt;
+                       </TD>
+                       <TD>
+                               CRC-16 (polynomial = x^16 + x^15 + x^2 + x^0, initialized with 0) of everything before the crc, back to and including the frame header sync code
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="subframe"><FONT SIZE="+1"><B>SUBFRAME</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#subframe_header"><I>SUBFRAME_HEADER</I></A>
+                       </TD>
+                       <TD>
+                               &nbsp;
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#subframe_constant"><I>SUBFRAME_CONSTANT</I></A><BR>|| <A HREF="#subframe_fixed"><I>SUBFRAME_FIXED</I></A><BR>|| <A HREF="#subframe_lpc"><I>SUBFRAME_LPC</I></A><BR>|| <A HREF="#subframe_verbatim"><I>SUBFRAME_VERBATIM</I></A>
+                       </TD>
+                       <TD VALIGN="TOP">
+                               The SUBFRAME_HEADER specifies which one.
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="subframe_header"><FONT SIZE="+1"><B>SUBFRAME_HEADER</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;1&gt;
+                       </TD>
+                       <TD>
+                               Zero bit padding, to prevent sync-fooling string of 1s
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;6&gt;
+                       </TD>
+                       <TD>
+                               Subframe type:
+                               <UL>
+                               <LI>
+                                       <TT>000000</TT> : <A HREF="#subframe_constant">SUBFRAME_CONSTANT</A>
+                               </LI>
+                               <LI>
+                                       <TT>000001</TT> : <A HREF="#subframe_verbatim">SUBFRAME_VERBATIM</A>
+                               </LI>
+                               <LI>
+                                       <TT>00001x</TT> : reserved
+                               </LI>
+                               <LI>
+                                       <TT>0001xx</TT> : reserved
+                               </LI>
+                               <LI>
+                                       <TT>001xxx</TT> : if(xxx <= 4) <A HREF="#subframe_fixed">SUBFRAME_FIXED</A>, xxx=order ; else reserved
+                               </LI>
+                               <LI>
+                                       <TT>01xxxx</TT> : reserved
+                               </LI>
+                               <LI>
+                                       <TT>1xxxxx</TT> : <A HREF="#subframe_lpc">SUBFRAME_LPC</A>, xxxxx=order-1
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;1+k&gt;
+                       </TD>
+                       <TD>
+                               'Wasted bits-per-sample' flag:
+                               <UL>
+                               <LI>
+                                       <TT>0</TT> : no wasted bits-per-sample in source subblock, k=0
+                               </LI>
+                               <LI>
+                                       <TT>1</TT> : k wasted bits-per-sample in source subblock, k-1 follows, unary coded; i.e. k=3 => 001 follows, k=7 => 0000001 follows.
+                               </LI>
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="subframe_constant"><FONT SIZE="+1"><B>SUBFRAME_CONSTANT</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;n&gt;
+                       </TD>
+                       <TD>
+                               Unencoded constant value of the subblock, n = frame's bits-per-sample.
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="subframe_fixed"><FONT SIZE="+1"><B>SUBFRAME_FIXED</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;n&gt;
+                       </TD>
+                       <TD>
+                               Unencoded warm-up samples (n = frame's bits-per-sample * predictor order).
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#residual"><I>RESIDUAL</I></A>
+                       </TD>
+                       <TD>
+                               Encoded residual
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="subframe_lpc"><FONT SIZE="+1"><B>SUBFRAME_LPC</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;n&gt;
+                       </TD>
+                       <TD>
+                               Unencoded warm-up samples (n = frame's bits-per-sample * lpc order).
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;4&gt;
+                       </TD>
+                       <TD>
+                               (Quantized linear predictor coefficients' precision in bits)-1 (1111 = invalid).
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;5&gt;
+                       </TD>
+                       <TD>
+                               Quantized linear predictor coefficient shift needed in bits (NOTE: this number is signed two's-complement).
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;n&gt;
+                       </TD>
+                       <TD>
+                               Unencoded predictor coefficients (n = qlp coeff precision * lpc order) (NOTE: the coefficients are signed two's-complement).
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#residual"><I>RESIDUAL</I></A>
+                       </TD>
+                       <TD>
+                               Encoded residual
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="subframe_verbatim"><FONT SIZE="+1"><B>SUBFRAME_VERBATIM</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;n*i&gt;
+                       </TD>
+                       <TD>
+                               Unencoded subblock; n = frame's bits-per-sample, i = frame's blocksize.
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="residual"><FONT SIZE="+1"><B>RESIDUAL</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;2&gt;
+                       </TD>
+                       <TD>
+                               Residual coding method:<BR>
+                               <UL>
+                               <LI>
+                                       <TT>00</TT> : partitioned rice coding
+                               </LI>
+                               <LI>
+                                       <TT>01-11</TT> : reserved
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#partitioned_rice"><I>RESIDUAL_CODING_METHOD_PARTITIONED_RICE</I></A>
+                       </TD>
+                       <TD>
+                               &nbsp;
+                       </TD>
+               </TR>
+               <TR>
+                       <TD>
+                       </TD>
+                       <TD BGCOLOR="#F4F4CC">
+                               <FONT SIZE="+1">NOTES</FONT><BR>
+                               <UL>
+                               <LI>
+                                       Currently, FLAC specifies only one entropy coding method.
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="partitioned_rice"><FONT SIZE="+1"><B>RESIDUAL_CODING_METHOD_PARTITIONED_RICE</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;4&gt;
+                       </TD>
+                       <TD>
+                               Partition order.
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               <A HREF="#rice_partition"><I>RICE_PARTITION</I></A>+
+                       </TD>
+                       <TD>
+                               There will be 2^order partitions.
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+       <P>
+       <TABLE WIDTH="100%" BORDER="0" CELLSPACING="0" CELLPADDING="0" BGCOLOR="#EEEED4"><TR><TD>
+       <TABLE WIDTH="100%" BORDER="1" BGCOLOR="#EEEED4">
+               <TR>
+                       <TD COLSPAN="2" BGCOLOR="#D3D4C5">
+                               <A NAME="rice_partition"><FONT SIZE="+1"><B>RICE_PARTITION</B></FONT></A>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;4(+5)&gt;
+                       </TD>
+                       <TD>
+                               Encoding parameter:<BR>
+                               <UL>
+                               <LI>
+                                       <TT>0000-1110</TT> : Rice parameter.
+                               </LI>
+                               <LI>
+                                       <TT>1111</TT> : Escape code, meaning the partition is in unencoded binary form using n bits per sample; n follows as a 5-bit number.
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+               <TR>
+                       <TD ALIGN="RIGHT" VALIGN="TOP" BGCOLOR="#F4F4CC">
+                               &lt;?&gt;
+                       </TD>
+                       <TD>
+                               Encoded residual.  The number of samples (n) in the partition is determined as follows:<BR>
+                               <UL>
+                               <LI>
+                                       if the partition order is zero, n = frame's blocksize
+                               </LI>
+                               <LI>
+                                       else if this is not the first partition of the subframe, n = (frame's blocksize / (2^partition order))
+                               </LI>
+                               <LI>
+                                       else n = (frame's blocksize / (2^partition order)) - predictor order
+                               </LI>
+                               </UL>
+                       </TD>
+               </TR>
+       </TABLE>
+       </TD></TR></TABLE>
+       </P>
+
+</TD></TR>
+</TABLE>
+
+
+</CENTER>
+
+<P>&nbsp;Copyright (c) 2000,2001,2002 Josh Coalson</P>
+
+</BODY>
+</HTML>
diff --git a/doc/html/goals.html b/doc/html/goals.html
new file mode 100644 (file)
index 0000000..09b2e0e
--- /dev/null
@@ -0,0 +1,136 @@
+<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
+<!-- Copyright (c)  2000,2001,2002  Josh Coalson -->
+<!-- Permission is granted to copy, distribute and/or modify this document -->
+<!-- under the terms of the GNU Free Documentation License, Version 1.1 -->
+<!-- or any later version published by the Free Software Foundation; -->
+<!-- with no invariant sections. -->
+<!-- A copy