Log - 2f6c5feae86a1648e20e251f19d6eb5ccb7a47c2 - platform/external/libopus

2f6c5fe Oops, fixed an uninitialized var error by Jean-Marc Valin · 14 years ago
71ae6d4 Some C90-ication by Jean-Marc Valin · 14 years ago
60ff999 Mode cleanup (removed redundant mdctSize field) by Jean-Marc Valin · 14 years ago
f477d58 Think I got static modes working again by Jean-Marc Valin · 14 years ago
aadaafd Updating static modes (not done yet) by Jean-Marc Valin · 14 years ago
a57e924 Removing some warnings by Jean-Marc Valin · 14 years ago
a011065 Improved quality of small frame sizes at low bitrate. by Jean-Marc Valin · 14 years ago
281c8b7 Preventing ridiculous bit depths > 16 bit/sample by Jean-Marc Valin · 14 years ago
6b95d8f Improved code for reusing ec structures by Jean-Marc Valin · 14 years ago
30d5125 No need to memset() the byte buffer by Jean-Marc Valin · 14 years ago
bb4f63f Compile fix for MSVC by Benjamin Jemlich · 14 years ago
c239bd6 removing some warnings by Jean-Marc Valin · 14 years ago
6c3788c Removing old PCL, added plc.h by Jean-Marc Valin · 14 years ago
dc68eb2 Comments, code cleanup by Jean-Marc Valin · 14 years ago
feca095 fixed-point: New PLC now enabled by default by Jean-Marc Valin · 14 years ago
7b7f071 fixed-point: More work on the PLC by Jean-Marc Valin · 14 years ago
0da0d91 fixed-point: PLC auto-correlation converted by Jean-Marc Valin · 14 years ago
3a9699e minor tweaks to frac_div32() by Jean-Marc Valin · 14 years ago
bd82ca8 Fixed-point version of frac_div32() by Jean-Marc Valin · 14 years ago
456eab2 More work on fixed-point Levinson-Durbin by Jean-Marc Valin · 14 years ago
890a9c0 Implementing two choices for the tf resolution by Jean-Marc Valin · 14 years ago
9a33c5c Remove unnecessary free of global stack on error by Jean-Marc Valin · 14 years ago
e2d5b33 Completely disabling the tf code at low bitrate by Jean-Marc Valin · 14 years ago
0a571ef Encoding the tf parameters after coarse energy by Jean-Marc Valin · 14 years ago
47ad4a0 Doing the log2Amp() just in one place by Jean-Marc Valin · 14 years ago
ffe5061 nothing to see here by Jean-Marc Valin · 14 years ago
01b54b9 Completely removed the old allocation table. by Jean-Marc Valin · 14 years ago
c51e98b Base everything on the "standard allocation table" by Jean-Marc Valin · 14 years ago
12e851d Using allocation table for "standard" frame sizes by Jean-Marc Valin · 14 years ago
bb8fa1f Allocation table now in bits/sample by Jean-Marc Valin · 14 years ago
6c37fb1 Changing the allocator resolution to 1/8 bit by Jean-Marc Valin · 14 years ago
6202c74 Fix for PLC crash when using large frames by Jean-Marc Valin · 14 years ago
32589cd Trying not to crash on bit errors by Jean-Marc Valin · 14 years ago
816f893 removed stupid code in transient_analysis() by Jean-Marc Valin · 14 years ago
4f0de0b Fixed stupid bug for single-sample bands by Jean-Marc Valin · 14 years ago
43e9406 Change ec_{enc|dec}_bit_prob to take probabilities in units of 1/65536 instead of 1/256. This allows them to use a single 16x16->32 multiply instead of a 24x8->32 multiply. Also change the time-frequency resolution flag coding to ensure that "0" is always the most-probable symbol (i.e., that prob("1")<50%), as that's where all the rounding error accumulates. by Timothy B. Terriberry · 14 years ago
299747e Provide direct implementations ec_{enc|dec}_bit_prob() that do not require a division instead of using the normal entropy coder path. This should be exactly equivalent to the existing code. by Timothy B. Terriberry · 14 years ago
8cc945c Cleaning up VBR to be frame-size independent by Jean-Marc Valin · 14 years ago
5246238 Folding tuning by Jean-Marc Valin · 14 years ago
7331977 RD optimisation now takes bitrate into account by Jean-Marc Valin · 14 years ago
bdb5871 Simplification to the transient detection code by Jean-Marc Valin · 14 years ago
8823261 More simplifications to the Viterbi code by Jean-Marc Valin · 14 years ago
0ef39c9 Simplifications to the tf_res RD code by Jean-Marc Valin · 14 years ago
9d420c5 Viterbi-based RD optimization of the tf_res decisions by Jean-Marc Valin · 14 years ago
581fdba changing probability of first tf resolution symbol by Jean-Marc Valin · 14 years ago
163b76e Adaptive time-frequency resolution by Jean-Marc Valin · 14 years ago
8035b65 Adds a range coder call to encode a single bit with arbitrary probability by Jean-Marc Valin · 14 years ago
c4ac570 Fixed a long-standing rare mismatch by Jean-Marc Valin · 14 years ago
40e993c Re-organised the special case for N==1 by Jean-Marc Valin · 14 years ago
c5d00a0 Bunch of fixes for frames of 2.5 ms. by Jean-Marc Valin · 14 years ago
9d06984 More recombining "infrastructure" by Jean-Marc Valin · 14 years ago
cee66b5 Code for recombining bands (not enabled yet) by Jean-Marc Valin · 14 years ago
2fcb239 Removing the rest of the >32-bit PVQ code by Jean-Marc Valin · 14 years ago
cccdfd4 Minor code simplifications by Jean-Marc Valin · 14 years ago
a148df1 Doing folding from the highest freq possible by Jean-Marc Valin · 14 years ago
791cfe1 Increasing the temporal resolution of transients by Jean-Marc Valin · 14 years ago
9f89cab Improving the transient detection by Jean-Marc Valin · 14 years ago
67217ee Splitting transients in time domain by Jean-Marc Valin · 14 years ago
6384196 Moving code to quant_band() by Jean-Marc Valin · 14 years ago
4c27552 Tuning/cleanup of angle quantization by Jean-Marc Valin · 14 years ago
cae30df Getting rid of PVQ-level split by Jean-Marc Valin · 14 years ago
4aa1b8d Most 5ms bands now have even width by Jean-Marc Valin · 14 years ago
20639c4 Tuned the 5ms mode to have a band at 4 kHz and one at 8 kHz by Jean-Marc Valin · 14 years ago
5f96146 Implemented CELT_SET_START_BAND by Jean-Marc Valin · 14 years ago
f5e2e32 Support for passing your own ec state to the decoder by Jean-Marc Valin · 14 years ago
4794f2e Support for passing your own ec state to the encoder by Jean-Marc Valin · 14 years ago
c1ca587 Saving ~1 kB by using chars for the alloc table by Jean-Marc Valin · 14 years ago
72c97f5 Sharing more code between encode and decode (bis) by Jean-Marc Valin · 14 years ago
746b2a8 Merged encoding/decoding of mono/stereo by Jean-Marc Valin · 14 years ago
eb92668 Trying to prevent PVQ-level splits by Jean-Marc Valin · 14 years ago
c803ee1 Entropy-coding the new split parameter. by Jean-Marc Valin · 14 years ago
2e8ba54 New band splitting method based on the stereo code by Jean-Marc Valin · 14 years ago
1520f89 Stereo quantization also made more generic by Jean-Marc Valin · 14 years ago
80e56f1 Making stereo code a bit more generic by Jean-Marc Valin · 14 years ago
0607382 Trivial code cleanup by Jean-Marc Valin · 14 years ago
aead79b Proper use of logN and logM for the offsets by Jean-Marc Valin · 14 years ago
c40d6f9 Making stereo code use [un]quant_bands() as well by Jean-Marc Valin · 14 years ago
94216ae Allocation adjustment code in quand_band(). by Jean-Marc Valin · 14 years ago
2123f35 Cleaning up the folding code. by Jean-Marc Valin · 14 years ago
d56c610 API change: optional resynthesis by Jean-Marc Valin · 14 years ago
e949b72 Fixed pitchEnd wrt to variable frame size. It should (sort of) all work now by Jean-Marc Valin · 14 years ago
ce4dd36 More work on variable frame size (getting rid of FRAMESIZE() ) by Jean-Marc Valin · 14 years ago
017fa85 Changing the encoder API to add the frame size by Jean-Marc Valin · 14 years ago
fbd82eb Oops! Fixed the fixed-point build by Jean-Marc Valin · 14 years ago
811f21a Fix for very short frame size (M=1) by Jean-Marc Valin · 14 years ago
073d0bc Yet another step towards variable frame size by Jean-Marc Valin · 14 years ago
0475267 More preparation work for variable frame size by Jean-Marc Valin · 14 years ago
3b918ba replaced instances of mode->nbShortMdcts by the dynamic number of MDCTs (M) by Jean-Marc Valin · 14 years ago
7f1c942 Allocation table stored with BITRES accuracy based on the short block size by Jean-Marc Valin · 14 years ago
5f63354 shortWindow hadn't been used for a while by Jean-Marc Valin · 14 years ago
f7547a4 Shrinking the MDCT's table by separately rotating by the 1/(8N) factor by Jean-Marc Valin · 14 years ago
65ee67a Making the band definition the same at all frame sizes. by Jean-Marc Valin · 14 years ago
be8d125 Minor tweak to the band layout offset by Jean-Marc Valin · 14 years ago
6578ea0 Re-enabled special stereo code for N=2 by Jean-Marc Valin · 14 years ago
4fb9568 Bands are now (again) a multiple of the number of short MDCTs by Jean-Marc Valin · 14 years ago
d6b79ee Skipping transient shape renormalisation in the encoder because by Jean-Marc Valin · 14 years ago
281a63e Bump bitstream version now that we broke everything by Jean-Marc Valin · 14 years ago
8271dc8 Tuning the amount of spreading based on Koen's feedback by Jean-Marc Valin · 14 years ago
9c30de5 Improved transient_analysis() by adding one frame of memory. by Jean-Marc Valin · 14 years ago
b450ed4 Removed unnecessary calls to log2Amp() by Jean-Marc Valin · 14 years ago