Log - 3e6691a703c5139d3052b99c1d9e6331c84e1b7e - platform/external/libopus

b7d5531 Change Visual Studio files to use PACKAGE_VERSION. by Ralph Giles · 11 years ago
2f2f9d6 Fully automate version updating by Ron · 11 years ago
c152d60 Use dynamic stack allocation in the SILK encoder. by Timothy B. Terriberry · 11 years ago
f77410d VBR bitrate calibration for variable duration. by Jean-Marc Valin · 11 years ago
3252bf2 Variable framesize improvements by Jean-Marc Valin · 11 years ago
f96fc8c Merge branch 'exp_analysis' by Jean-Marc Valin · 12 years ago
f55b0ea Documentation fixes by Jean-Marc Valin · 12 years ago
69c3dcd Fixes a bug introduced in 851f803 by Jean-Marc Valin · 12 years ago
851f803 Moves mono downmixing and upsampling to the compute_mdcts() by Jean-Marc Valin · 12 years ago
541a472 Adds code to revise the transient decision using the band energy by Jean-Marc Valin · 12 years ago
2a5f056 Running transient_analysis() even for 2.5 ms frames by Jean-Marc Valin · 12 years ago
48ac122 Makes analysis usable for all frame sizes by Jean-Marc Valin · 12 years ago
ebdfbfb Fixes stupid preprocessor-related bug introduced in e368e6209. by Jean-Marc Valin · 12 years ago
fa28dea Fixes a seed issue introduced in 5367dac by Jean-Marc Valin · 12 years ago
e368e62 Remove condition in pre/de-emphasis when not building with custom modes by Jean-Marc Valin · 12 years ago
5367dac seed and remaining_bits moved to band context by Jean-Marc Valin · 12 years ago
a9d6286 Using a band context to reduce the number of arguments being passed around. by Jean-Marc Valin · 12 years ago
f18acee more bands.c cleaning up by Jean-Marc Valin · 12 years ago
ddd0d21 collapse mask cleanup by Jean-Marc Valin · 12 years ago
414fd27 Removes useless parameters and re-indents the code by Jean-Marc Valin · 12 years ago
3d6c341 Splitting off the recursion in quant_partition() by Jean-Marc Valin · 12 years ago
4ffbf21 stereo part of quant_band() moved to quant_band_stereo() by Jean-Marc Valin · 12 years ago
113a5e1 Moves the split angle code to compute_theta() by Jean-Marc Valin · 12 years ago
dfe3bf9 First step in quant_band() cleanup: N=1 case. by Jean-Marc Valin · 12 years ago
d683c76 Fixes MSVC warnings for double->float and float->int conversion by Jean-Marc Valin · 12 years ago
5fb50ad skip bands with no energy by Jean-Marc Valin · 12 years ago
7509fdb New bandwidth detection code by Jean-Marc Valin · 12 years ago
90bac9d Minor tuining of complexity thresholds by Jean-Marc Valin · 12 years ago
28ef277 VBR code simplification by Jean-Marc Valin · 12 years ago
dae16fb Cleanup: tf_estimate now ranges from 0 to 1 by Jean-Marc Valin · 12 years ago
f8809dd Dynalloc rate calibration by Jean-Marc Valin · 12 years ago
1122d29 Makes "stereo savings" less aggressive on mono or near-mono. by Jean-Marc Valin · 12 years ago
b33db8f Makes tonality boost less aggressive by Jean-Marc Valin · 12 years ago
41fd7a1 Disables the new spread+tapset estimator by Jean-Marc Valin · 12 years ago
8eb9bb7 Numerous PLC cleanups. by Timothy B. Terriberry · 12 years ago
9283114 Adds explicit valgrind checks when ENABLE_VALGRIND is defined by Jean-Marc Valin · 12 years ago
c093829 Do imdct post-rotate and deshuffle in-place. by Nils Wallménius · 12 years ago
aa3a599 Fixes forward MDCT for overlaps that aren't a multiple of 4 by Jean-Marc Valin · 12 years ago
54f7cb4 Updates MDCT unit test by Jean-Marc Valin · 12 years ago
027a202 More comments in the PLC by Jean-Marc Valin · 12 years ago
cacb566 Place output of imdct post-rotate and deshuffle in out buffer to avoid copying and simplify the code. by Nils Wallménius · 12 years ago
e0884fe Window both sides of overlap when adding them. by Nils Wallménius · 12 years ago
0094c88 Better comments in the PLC by Jean-Marc Valin · 12 years ago
c3273ed Fixes an overflow in the fixed-point celt_sqrt() for large values. by Jean-Marc Valin · 12 years ago
e0491e7 Fixes many many issues in the CELT PLC by Jean-Marc Valin · 12 years ago
83f8012 Remove obsolete #ifdef OPUS_BUILD. by Ralph Giles · 12 years ago
07418d9 Guard _BitScanReverse on MSVC so that MSVC 6 doesn't break. by Gregory Maxwell · 12 years ago
413caa0 Tweaks to the new transient detector by Jean-Marc Valin · 12 years ago
4357e93 Follow up to commit 8b906c102b37ce74571898c8879c01b2a9088fb3 by Nils Wallménius · 12 years ago
702458a Should make celt_maxabs*() slightly faster by Jean-Marc Valin · 12 years ago
144b6e6 Oops, fixes transient_analysis() for the right channel by Jean-Marc Valin · 12 years ago
10b30e7 Fixes a VBR bug caused by maxDepth not to be computed without dynalloc by Jean-Marc Valin · 12 years ago
c823120 Add the new source files to the MSVC projects. by Ralph Giles · 12 years ago
d47308d Remove long-outdated comment by Jean-Marc Valin · 12 years ago
9ed3d02 Fixes a quant_coarse_energy() hybrid mode bug by Jean-Marc Valin · 12 years ago
1fd1d7d Fixes a bug where transient frames would use previous spreading by Jean-Marc Valin · 12 years ago
a6d663c Disables tf_analysis() for hybrid mode by Jean-Marc Valin · 12 years ago
1ecb7ea Fixes issues with multiple files defining CELT_C by Jean-Marc Valin · 12 years ago
6906210 Splits out the CELT encoder and decoder by Jean-Marc Valin · 12 years ago
8b906c1 Changed denormalize_bands() to only work between "start" and "end" by Jean-Marc Valin · 12 years ago
7aa593a Reduces stack in quant_all_bands() for hybrid mode by Jean-Marc Valin · 12 years ago
fd314f4 More stack usage reduction in quant_all_bands() by Jean-Marc Valin · 12 years ago
5055cf8 Removes the need for lowband_scratch in quant_all_bands() by Jean-Marc Valin · 12 years ago
06bab45 Moved up the second encoder mdct to reduce stack usage by Jean-Marc Valin · 12 years ago
cc83f6b Reduces decoder stack usage by Jean-Marc Valin · 12 years ago
4ea3ae9 Further cleanup of the MDCT code, fixes PLC bug by Jean-Marc Valin · 12 years ago
21b7164 Avoid copying imdct output by Nils Wallménius · 12 years ago
5ab07e6 comb_filter() bypass for the case where the gain is zero. by Jean-Marc Valin · 12 years ago
fedfe91 Oops, put back the "static" for transient_analysis() by Jean-Marc Valin · 12 years ago
36890bc Fixes a fixed-point overflow in the new transient detector by Jean-Marc Valin · 12 years ago
f2729de Fixes a fixed-point divide-by-zero issue by Jean-Marc Valin · 12 years ago
fac68ce New transient detection algorithm by Jean-Marc Valin · 12 years ago
00898f2 Adds missing license headers on some scripts and test files by Jean-Marc Valin · 12 years ago
c8ce2ef Keeping the deemphasis coefficients in local variable to avoid aliasing issues by Jean-Marc Valin · 12 years ago
aaf40ef Moves the pitch prefiltering code to a separate function by Jean-Marc Valin · 12 years ago
6930d90 oops... fixes a C89 issue by Jean-Marc Valin · 12 years ago
1cdfd1b Moved pre-emphasis to a separate function and cleaned it up by Jean-Marc Valin · 12 years ago
cc9f28c Optimizations to comb_filter() by Jean-Marc Valin · 12 years ago
d3fad97 Removes more indirections in celt.c by Jean-Marc Valin · 12 years ago
297195b Removes a bunch of indirections from the CELT encoder and decoder by Jean-Marc Valin · 12 years ago
4be7fe7 This should speed up deemphasis() quite a bit. by Jean-Marc Valin · 12 years ago
8ccaefd Take the PLC out of the stack size critical path by Jean-Marc Valin · 12 years ago
027ec51 Fix MSVC format conversion warnings. by Ralph Giles · 12 years ago
2572c1e Merge inverse mdct post-rotate and de-shuffle loops by Nils Wallménius · 12 years ago
ed627f0 Attempt to have at least two bands that are coded by Jean-Marc Valin · 12 years ago
3aa2647 Fix an indentation issue. by Ralph Giles · 12 years ago
b7b5872 Fix minor issues reported by scan-build by Philip Jägenstedt · 12 years ago
1b61303 Fixes a bug in the LSB_DEPTH code introduced in cf1053dc by Jean-Marc Valin · 12 years ago
02dcf6a Fixes a fixed-point stereo savings issue causing a celt_ilog2(0) by Jean-Marc Valin · 12 years ago
0e49059 Gets rid of unused boost analysis code that never worked anyway. by Jean-Marc Valin · 12 years ago
7315b35 Merge branch 'exp_analysis7' by Jean-Marc Valin · 12 years ago
2c3b5ee Saves channels*overlap*4 bytes off the state encoder by Jean-Marc Valin · 12 years ago
454d3d5 s/SHL/SHL16/ by Jean-Marc Valin · 12 years ago
97a4847 Makes dynalloc more conservative for CBR and CVBR by Jean-Marc Valin · 12 years ago
ed5989b Bit allocation fix for 16-bit platforms by Jean-Marc Valin · 12 years ago
26dbb8e Adds anslysis code for tf_select by Jean-Marc Valin · 12 years ago
79a6f4d Makes two static tables const by Jean-Marc Valin · 12 years ago
5685bd3 Replace long long in celt/ with opus_int64. by Gregory Maxwell · 12 years ago
de74fc1 Remove C99ism in celt/tests/test_unit_mathops.c w/ fixed point build. by Gregory Maxwell · 12 years ago
e2e1fe9 Use 64-bits in fixed point debug for _P16, fixes spurious failures. by Gregory Maxwell · 12 years ago