| 2006-05-07 |
jm | Removed the whole LPC enhancement -- wasn't really... |
tree | commitdiff |
| 2006-05-07 |
jm | Oops! Defined MAC16_16_P13 for float and fixed-point... |
tree | commitdiff |
| 2006-05-07 |
jm | fixed-point accuracy improvements in both compute_rms16... |
tree | commitdiff |
| 2006-05-07 |
jm | Slight accuracy improvement in Levinson-Durbin algo... |
tree | commitdiff |
| 2006-05-04 |
jm | Gapless works again (I think). Bit of cleaning up as... |
tree | commitdiff |
| 2006-05-04 |
jm | Fixed a wideband regression (innovation wasn't saved... |
tree | commitdiff |
| 2006-05-04 |
jm | added missing call to jitter_buffer_tick() in speex_jit... |
tree | commitdiff |
| 2006-05-03 |
jm | reverted to original SpeexJitter API |
tree | commitdiff |
| 2006-05-02 |
jm | Modify the open-loop gain depending on pitch gain |
tree | commitdiff |
| 2006-05-01 |
jm | Simple test for saturation. Saturated samples all get... |
tree | commitdiff |
| 2006-04-30 |
jm | Approximation for pitch search in new enhancer. Trying... |
tree | commitdiff |
| 2006-04-30 |
jm | cleaning up new enhancer |
tree | commitdiff |
| 2006-04-30 |
jm | blackfin: proper rounding in 16-bit filters |
tree | commitdiff |
| 2006-04-29 |
jm | fir_mem16 implemented using filter_mem16 on Blackfin |
tree | commitdiff |
| 2006-04-29 |
jm | blackfin assembly for iir_mem16 |
tree | commitdiff |
| 2006-04-29 |
jm | Need to set buffer to NULL after free()ing. Thanks... |
tree | commitdiff |
| 2006-04-29 |
jm | Blackfin assembly version of filter_mem16 |
tree | commitdiff |
| 2006-04-29 |
jm | Better data alignment for Blackfin (i.e. no bus error... |
tree | commitdiff |
| 2006-04-29 |
jm | big kludge for braindead (non C99) compilers. |
tree | commitdiff |
| 2006-04-29 |
jm | patch by Brian Retford to remove the weird struct manip... |
tree | commitdiff |
| 2006-04-29 |
jm | Patch for two memory leaks and a read overflow in the... |
tree | commitdiff |
| 2006-04-28 |
jm | pcm_wrapper was a dumb idea in the first place. |
tree | commitdiff |
| 2006-04-28 |
jm | No longer needed (was made for symbian before we figure... |
tree | commitdiff |
| 2006-04-27 |
jm | new enhancer now fully fixed-point |
tree | commitdiff |
| 2006-04-27 |
jm | two more variables in fixed-point. Oh, and un-b0rked... |
tree | commitdiff |
| 2006-04-27 |
jm | more fixed-point in new enhancer |
tree | commitdiff |
| 2006-04-27 |
jm | three more vars in fixed-point (four to go). |
tree | commitdiff |
| 2006-04-26 |
jm | oops. Previous checking broke the floating point. Think... |
tree | commitdiff |
| 2006-04-26 |
jm | enhancer fixed-point work |
tree | commitdiff |
| 2006-04-25 |
jm | intermediate fixed-point work |
tree | commitdiff |
| 2006-04-25 |
jm | simpler code, similar results |
tree | commitdiff |
| 2006-04-25 |
jm | some tuning, replaced conditional constants with QCONST16. |
tree | commitdiff |
| 2006-04-25 |
jm | TI C5x and C6x patch sent by Jim Crichton. |
tree | commitdiff |
| 2006-04-22 |
jm | fixed an old bug in QCONST (and made it 16-bit clean) |
tree | commitdiff |
| 2006-04-22 |
jm | QCONST should now be 16-bit clean |
tree | commitdiff |
| 2006-04-21 |
jm | Fixed a major regression in the 4 kbps mode by adding... |
tree | commitdiff |
| 2006-04-21 |
jm | 16-bit cleanliness |
tree | commitdiff |
| 2006-04-20 |
jm | Vorbis-psy works again |
tree | commitdiff |
| 2006-04-20 |
jm | this makes the _mem16 version of the filters use the... |
tree | commitdiff |
| 2006-04-20 |
jm | Make sure nobody breaks the code by changing the analys... |
tree | commitdiff |
| 2006-04-19 |
jm | 24.6 kbps mode now entirely in fixed-point. |
tree | commitdiff |
| 2006-04-19 |
jm | oops, re-enabled lpc enhancement |
tree | commitdiff |
| 2006-04-19 |
jm | fixed narrowband mode 7 |
tree | commitdiff |
| 2006-04-19 |
jm | energy scaling is a bit more accurate (especially in... |
tree | commitdiff |
| 2006-04-19 |
jm | Added some EXTEND32 for 16-bit machines |
tree | commitdiff |
| 2006-04-18 |
jm | Fixed ring buffer (not, it's not the same as a stack... |
tree | commitdiff |
| 2006-04-17 |
jm | C89-correctness |
tree | commitdiff |
| 2006-04-17 |
jm | fixed enhancement for wideband (sort of) and a bit... |
tree | commitdiff |
| 2006-04-16 |
jm | this should fix the ARM assembly (but I can't test... |
tree | commitdiff |
| 2006-04-16 |
jm | minor tweaks to cheb_poly_eva() |
tree | commitdiff |
| 2006-04-14 |
jm | minor simplification |
tree | commitdiff |
| 2006-04-14 |
jm | fixed warnings, added const where appropriate |
tree | commitdiff |
| 2006-04-14 |
jm | comments only |
tree | commitdiff |
| 2006-04-14 |
jm | added speex_echo_capture() and speex_echo_playback... |
tree | commitdiff |
| 2006-04-14 |
jm | minor type cleanup |
tree | commitdiff |
| 2006-04-14 |
jm | minor tweak |
tree | commitdiff |
| 2006-04-13 |
jm | Just a bunch of (scalar) float ops left in the new... |
tree | commitdiff |
| 2006-04-13 |
jm | fixed some NaNs |
tree | commitdiff |
| 2006-04-13 |
jm | version change |
tree | commitdiff |
| 2006-04-13 |
jm | oops, wasn't scaling the right thing... |
tree | commitdiff |
| 2006-04-13 |
jm | simplified packet loss handling |
tree | commitdiff |
| 2006-04-12 |
jm | new enhancer compiles and (sort of) works with FIXED_PO... |
tree | commitdiff |
| 2006-04-12 |
jm | removed unused stuff |
tree | commitdiff |
| 2006-04-12 |
jm | OK, jitter_buffer_get() now returns the offset, which... |
tree | commitdiff |
| 2006-04-12 |
jm | yet another change to the jitter buffer API (passing... |
tree | commitdiff |
| 2006-04-11 |
jm | open-loop excitation gain computed in 16 bits |
tree | commitdiff |
| 2006-04-11 |
jm | rewrote new enhancer pitch tracker using the open-loop... |
tree | commitdiff |
| 2006-04-11 |
jm | added explicit errors for those compiling with stupid... |
tree | commitdiff |
| 2006-04-10 |
jm | added -no-undefined to the libspeex options as suggeste... |
tree | commitdiff |
| 2006-04-10 |
jm | innovation is saved directly to the sb_celp en/decoder... |
tree | commitdiff |
| 2006-04-10 |
jm | ringing computation now done with 16-bit accuracy. |
tree | commitdiff |
| 2006-04-09 |
jm | moved some temporary variables from decoder state to... |
tree | commitdiff |
| 2006-04-08 |
jm | Got rid of ~64 divides per sub-frame and saved 512... |
tree | commitdiff |
| 2006-04-08 |
jm | minor RAM tweak: delayed stack allocation to reduce... |
tree | commitdiff |
| 2006-04-08 |
jm | weighted signal now in 16-bit precision, saving 500... |
tree | commitdiff |
| 2006-04-08 |
jm | removed unused variables and cleaned up some non-C89ness |
tree | commitdiff |
| 2006-04-07 |
jm | oops. Fixed an off-by-one that could (at least theoreti... |
tree | commitdiff |
| 2006-04-07 |
jm | LPC analysis window now stored as constant in code... |
tree | commitdiff |
| 2006-04-07 |
jm | another 640 bytes removed from the encoder state (using... |
tree | commitdiff |
| 2006-04-07 |
jm | removed the decoder output buffer -- saving 640 bytes... |
tree | commitdiff |
| 2006-04-07 |
jm | Fixed several fixed-point regressions: |
tree | commitdiff |
| 2006-04-04 |
jm | accuracy improvement in QMF |
tree | commitdiff |
| 2006-04-04 |
jm | Got fine timing adjustments to work (tick size instead... |
tree | commitdiff |
| 2006-04-03 |
jm | cleaner implementation of the buffer margin (now segfau... |
tree | commitdiff |
| 2006-04-01 |
jm | cleanup and saved 480 bytes of RAM in the decoder |
tree | commitdiff |
| 2006-04-01 |
jm | bit of cleaning up. More to come. |
tree | commitdiff |
| 2006-04-01 |
jm | non-causal comb filter |
tree | commitdiff |
| 2006-03-24 |
jm | Added a desired buffer size feature. |
tree | commitdiff |
| 2006-03-24 |
jm | Jitter buffer is more robust. |
tree | commitdiff |
| 2006-03-23 |
jm | another major reorganisation, generalisation in the... |
tree | commitdiff |
| 2006-03-22 |
jm | Implemented the old (speex-dependent) jitter buffer... |
tree | commitdiff |
| 2006-03-22 |
jm | cleaning up |
tree | commitdiff |
| 2006-03-21 |
jm | Reworking the jitter buffer. Still very incomplete. |
tree | commitdiff |
| 2006-03-20 |
jm | Fix the Vorbis curve for the smaller LPC window. |
tree | commitdiff |
| 2006-03-19 |
jm | Getting rid of 25600 divide operations/sec in the pitch... |
tree | commitdiff |
| 2006-03-18 |
jm | new experimental comb filter code |
tree | commitdiff |
| 2006-03-15 |
jm | Making it easier to play with the excitation as I see... |
tree | commitdiff |
| 2006-02-21 |
jm | adjust group delay |
tree | commitdiff |
| 2006-02-21 |
jm | adjust group delay |
tree | commitdiff |
| 2006-02-20 |
jm | reduced lpc lookahead from 10ms to 5ms. No significant... |
tree | commitdiff |
| next |
xiph / speexdsp.git / history