2002-10-27 |
jm | The temp stack is now void* instead of float* |
blob | commitdiff |
2002-10-27 |
jm | ... |
blob | commitdiff | diff to current |
2002-10-26 |
jm | Stack allocation cleanup... |
blob | commitdiff | diff to current |
2002-10-25 |
jm | Code cleanup |
blob | commitdiff | diff to current |
2002-10-23 |
jm | fixed a double_codebook bug and prevented pitch from... |
blob | commitdiff | diff to current |
2002-10-23 |
jm | Compute impulse response of "perceptual synthesis filte... |
blob | commitdiff | diff to current |
2002-10-23 |
jm | Continuing cleanup, removed unused functions, ... |
blob | commitdiff | diff to current |
2002-10-23 |
jm | Merged split_cb_search_nogain and split_cb_search_shape... |
blob | commitdiff | diff to current |
2002-10-23 |
jm | Re-wrote the signed search to be like the unsigned... |
blob | commitdiff | diff to current |
2002-10-23 |
jm | Removed un-necessary copies and re-ordered some calcula... |
blob | commitdiff | diff to current |
2002-10-22 |
jm | Slight optimization to the way the target is updated... |
blob | commitdiff | diff to current |
2002-10-21 |
jm | Filter optimizations, cleanup, removed the stack POP... |
blob | commitdiff | diff to current |
2002-10-11 |
jm | Changed license to BSD |
blob | commitdiff | diff to current |
2002-08-22 |
jmvalin | Win32 support, thanks to john33 (Hydrogen Audio) |
blob | commitdiff | diff to current |
2002-08-16 |
jmvalin | stupid VBR bug corrected + improvements |
blob | commitdiff | diff to current |
2002-08-15 |
jmvalin | oops... |
blob | commitdiff | diff to current |
2002-08-14 |
jmvalin | Cleaned up vocoder mode... |
blob | commitdiff | diff to current |
2002-07-30 |
jmvalin | Fixed post-filter producing NaN's, started writing... |
blob | commitdiff | diff to current |
2002-07-29 |
jmvalin | added encoder complexity option |
blob | commitdiff | diff to current |
2002-07-28 |
jmvalin | Allow more than one frame per packet |
blob | commitdiff | diff to current |
2002-07-20 |
jmvalin | Implemented new "global" search for the whole sub-frame... |
blob | commitdiff | diff to current |
2002-07-20 |
jmvalin | Removed redundent searches |
blob | commitdiff | diff to current |
2002-07-20 |
jmvalin | Added an n-best VQ search function in order to simplify... |
blob | commitdiff | diff to current |
2002-07-06 |
jmvalin | Cleanup |
blob | commitdiff | diff to current |
2002-07-05 |
jmvalin | code cleanup |
blob | commitdiff | diff to current |
2002-07-04 |
jmvalin | Trained some codebooks... |
blob | commitdiff | diff to current |
2002-07-02 |
jmvalin | Cleaned up unused stuff and licensed all codebooks... |
blob | commitdiff | diff to current |
2002-06-26 |
jmvalin | Added a comfort noise mode... |
blob | commitdiff | diff to current |
2002-06-07 |
jmvalin | Implemented an n-best open-loop pitch search to speed... |
blob | commitdiff | diff to current |
2002-06-07 |
jmvalin | Removed some warnings, fixed posfilter mode for wideband |
blob | commitdiff | diff to current |
2002-06-07 |
jmvalin | High-band (for wideband) is now a shape-sign codebook... |
blob | commitdiff | diff to current |
2002-05-18 |
jmvalin | FrameBits renamed to SpeexBits and a "lost" argument... |
blob | commitdiff | diff to current |
2002-05-14 |
jmvalin | Big changes in both narrowband and wideband. Retrained... |
blob | commitdiff | diff to current |
2002-05-09 |
jmvalin | Last cleanup for 0.0.3 Speex-0.0.3 |
blob | commitdiff | diff to current |
2002-05-06 |
jmvalin | Finished cleaning up the sub-band mode |
blob | commitdiff | diff to current |
2002-05-03 |
jmvalin | speex.[ch] renamed to nb_celp.[ch] for consistency |
blob | commitdiff | diff to current |
2002-04-29 |
jmvalin | removed unused variable |
blob | commitdiff | diff to current |
2002-04-29 |
jmvalin | Decoder back in sync with the encoder |
blob | commitdiff | diff to current |
2002-04-23 |
jmvalin | Fixed warnings |
blob | commitdiff | diff to current |
2002-04-23 |
jmvalin | Many, many updates. Better split-VQ search, better... |
blob | commitdiff | diff to current |
2002-04-14 |
jmvalin | Got rid of a couple mallocs... |
blob | commitdiff | diff to current |
2002-04-12 |
jmvalin | Speed improvements (got rid of a couple divide ops... |
blob | commitdiff | diff to current |
2002-04-12 |
jmvalin | Speed improvements: faster FIR filter and better algori... |
blob | commitdiff | diff to current |
2002-04-10 |
jmvalin | Big code cleanup, some minor bug fixed too |
blob | commitdiff | diff to current |
2002-04-10 |
jmvalin | Think both encoder and decoder for SB-CELP work... |
blob | commitdiff | diff to current |
2002-04-09 |
jmvalin | SB-CELP decoder (continued) |
blob | commitdiff | diff to current |
2002-04-09 |
jmvalin | SB-CELP work... more to go |
blob | commitdiff | diff to current |
2002-04-09 |
jmvalin | More decoder work |
blob | commitdiff | diff to current |
2002-04-08 |
jmvalin | Removed useless stuff |
blob | commitdiff | diff to current |
2002-04-08 |
jmvalin | Re-wrote the gain quantization for split-VQ excitation... |
blob | commitdiff | diff to current |
2002-03-27 |
jmvalin | code cleanup, some optimizations Speex-0.0.1 |
blob | commitdiff | diff to current |
2002-03-22 |
jmvalin | Multi-pulse seems to work not too bad (but disabled... |
blob | commitdiff | diff to current |
2002-03-22 |
jmvalin | Modified WB perceptual filter, starting a multi-pulse... |
blob | commitdiff | diff to current |
2002-03-20 |
jmvalin | Added joint optimization of excitation gains |
blob | commitdiff | diff to current |
2002-03-19 |
jmvalin | crappy wideband codec at 28.5 kbps... |
blob | commitdiff | diff to current |
2002-03-18 |
jmvalin | Wideband almost done but buggy... |
blob | commitdiff | diff to current |
2002-03-15 |
jmvalin | wideband coexists with narrowband now |
blob | commitdiff | diff to current |
2002-03-15 |
jmvalin | OK, back to normal (narrowband codec works) |
blob | commitdiff | diff to current |
2002-03-15 |
jmvalin | More wideband stuff... |
blob | commitdiff | diff to current |
2002-03-15 |
jmvalin | We're going wideband... |
blob | commitdiff | diff to current |
2002-03-14 |
jmvalin | Think the modularity stuff is mostly done... |
blob | commitdiff | diff to current |
2002-03-14 |
jmvalin | Cleanup mostly done for the encoder. All functions... |
blob | commitdiff | diff to current |
2002-03-14 |
jmvalin | More cleanup in codebook search... |
blob | commitdiff | diff to current |
2002-03-13 |
jmvalin | Decoder seems to work |
blob | commitdiff | diff to current |
2002-03-13 |
jmvalin | More decoder stuff |
blob | commitdiff | diff to current |
2002-03-13 |
jmvalin | Code cleanup, removed warnings |
blob | commitdiff | diff to current |
2002-03-11 |
jmvalin | The encoder (testenc) now produces a bitstream |
blob | commitdiff | diff to current |
2002-03-11 |
jmvalin | Working demo at 14.5 kbps (fully quantized) |
blob | commitdiff | diff to current |
2002-03-08 |
jmvalin | Working demo... |
blob | commitdiff | diff to current |
2002-03-07 |
jmvalin | The code is getting horribly messy, but there's too... |
blob | commitdiff | diff to current |
2002-03-05 |
jmvalin | Big cleanup... filter memories, stack allocation, ... |
blob | commitdiff | diff to current |
2002-02-28 |
jmvalin | Pitch prediction stuff... |
blob | commitdiff | diff to current |
2002-02-27 |
jmvalin | Started the decoder part, I think we now update filters... |
blob | commitdiff | diff to current |
2002-02-27 |
jmvalin | Added 3-tap pitch predictor by analysis by synthesis |
blob | commitdiff | diff to current |
2002-02-26 |
jmvalin | Think the filter stuff works with W(z)=A(z/.9)/A(z/.5) |
blob | commitdiff | diff to current |
2002-02-26 |
jmvalin | This time the filters really work with W(z)=A(z/.9) |
blob | commitdiff | diff to current |
2002-02-26 |
jmvalin | Think I figuring out the filter stuff |
blob | commitdiff | diff to current |
2002-02-19 |
jmvalin | Excitation codebook stuff, but it doesn't work;-( |
blob | commitdiff | diff to current |
|