RSS feed
31348844
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/14 06:55
Speeded up cwrsi and icwrs by at least an order of magnitude. Now using
06c67dbc
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/13 12:41
Simplified exp_spreading by unifying both directions. Have to admit I don't
c9cc6d3e
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/13 06:37
Introducing a (very) crude budget for the energy encoder.
cc4d3dda
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/12 19:08
Fixed stereo regression introduced in 05686a5d6e366d3a067c39f1b8567def7baa450d
cab576ec
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/12 12:21
Updating only the L-best entries in alg_quant() that are useful.
fe419836
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/12 10:05
Copying pointers is faster than copying arrays (who knew!).
846d4e24
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/12 08:48
Allowing the quantiser serch to put more than one pulse at one,
7d40730e
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/12 06:32
optional fast ncwrs64() implementation (copied from ncwrs) requiring
05686a5d
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/11 12:54
As extra safety, make sure not to use pitch prediction when it would make
7e8a487e
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/11 11:55
No longer encoding the pitch index if the gain is zero anyway.
cb7a2a3d
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/11 11:44
Re-ordered the parameters in the stream: [energy, pitch index, pitch gains]
0e20ca09
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/11 10:33
Enabling "plain folding" in case we don't even have enough bits for intra-frame
8f0f4b94
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/11 08:52
Limiting intra-frame prediction codebook to 32 entries (plus sign)
5b50f4bf
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/11 06:50
Decays corresponding to the psychoacoustic spreading function are now
b7ed44ba
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/11 06:15
minor tweak to pitch weighting function, disabled some code that doesn't
66fa639c
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/11 05:53
More flexible energy quantisation with nearly no penalty.
c35883d7
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/11 05:16
Remove multiplier-free version since we don't need it
7351e286
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/08 10:09
Everything converted to use kiss-fft. Got rid of smallft and fftwrap.
6211c90d
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/08 09:21
Split the radix functions into forward and backward versions, removed the
d7dfb008
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/08 08:25
Made pre-computed twiddles the same for forward and inverse FFT
711ad251
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/08 07:22
Changed ordering of real FFT freq data to something more aligned (it's not
b3756709
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/08 06:50
Fixed incorrect assumption about the number of bytes returned by the
4d0a7d0f
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/07 07:24
Now using an MDCT implementation I can actually understand.
01240776
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/07 16:14
Re-enabled intra-frame prediction, which seems to have exposed a few issues
cb0956d0
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/07 15:21
Add a test for tell(). Turns out we can make it fail by writing zeros.
276e6afc
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/02 15:15
Fixed a typo and removed an old warning from speexenc. Also, using more useful
e623312f
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/01 11:27
Got the Ogg encoder and decoder to (barely) work.
17e5b805
– Jean-Marc Valin <
[email protected]>
authored
on 2008/02/01 04:02
clean up... oh and actually committed the header definition
2fecbdf0
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/31 18:43
Fixed the default int32 type which was wrong on amd64 (and added testcase).
4c6bc884
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/31 12:34
"make check" now actually tried to encode and decode to see if things work
abe043f0
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/31 09:26
Moving everything to the same type abstraction (sort of).
ccc988b8
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/31 04:59
type abstraction header and header definition (still incomplete)
4618eee0
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/29 03:17
Added a check to make sure the encoder signal matches that of the decoder
38b6d479
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/29 03:04
remove prob{enc,dec,mod} that were no longer needed
6238bc0e
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/28 17:28
Moved the content of libentcode into libcelt to reduce dependencies,
94d4ea93
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/27 12:34
Bit of cleaning up in the byte dumping part. Making use of any remaining bit(s)
bab4fb19
– tterribe <tterribe@0101bb08-14d6-0310-b084-bc0e0c8e3800>
authored
and Jean-Marc Valin <
[email protected]>
committed
on 2008/01/23 18:10
Actually free test data vectors in the new random stream tests.
7e3293f7
– tterribe <tterribe@0101bb08-14d6-0310-b084-bc0e0c8e3800>
authored
and Jean-Marc Valin <
[email protected]>
committed
on 2008/01/23 18:04
Fix the bug that was causing the last byte of the stream to be returned incorrectly, as well as undo jm's reversion, which can cause ec_dec_tell() to operate incorrectly at the end of the stream.
77ff83a2
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/22 08:16
Making the library usable by other projects (install things properly, added
388f1d6c
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/18 11:28
Tweaks for slightly improving the handling of the last byte (still not
e5ec0f8d
– jm@0101bb08-14d6-0310-b084-bc0e0c8e3800 <jm@0101bb08-14d6-0310-b084-bc0e0c8e3800>
authored
and Jean-Marc Valin <
[email protected]>
committed
on 2008/01/17 23:24
Reverted some of Timothy's changes to the range decoding that were causing decode random decode problems at the end of the stream (but only once in a while).
3cfcd059
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/18 08:56
Big mode cleanup. Keeping only one mono and one stereo mode
a76a0b23
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/17 17:43
Further simplified the API by passing the rate directly to the
649d5cc2
– tterribe <tterribe@0101bb08-14d6-0310-b084-bc0e0c8e3800>
authored
and Jean-Marc Valin <
[email protected]>
committed
on 2008/01/17 02:51
Update ec_dec_bits64() to correspond to the new ec_enc_bits64(). Not sure why this wasn't committed before.
b35536a3
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/17 11:57
Enabled pure CBR, though there's still some decoder issues.
f51ca493
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/17 05:58
Looks like the bit allocation code is mostly working. Just need to actually
3eff11d8
– tterribe <tterribe@0101bb08-14d6-0310-b084-bc0e0c8e3800>
authored
and Jean-Marc Valin <
[email protected]>
committed
on 2008/01/11 00:51
Add ec_dec_tell() to report the same information as ec_enc_tell() decode side. Remove ec_enc_tell() and replace it with ec_enc_tellf(): just request 0 fractional bits if you want an integer value.
fad779ca
– tterribe <tterribe@0101bb08-14d6-0310-b084-bc0e0c8e3800>
authored
and Jean-Marc Valin <
[email protected]>
committed
on 2008/01/11 00:12
Add ec_enc_tellf, which can return the number of bits used to fractional precision.
06390d08
– tterribe <tterribe@0101bb08-14d6-0310-b084-bc0e0c8e3800>
authored
and Jean-Marc Valin <
[email protected]>
committed
on 2008/01/10 22:13
Add oc_enc_tell to report an estimate of the number of bits used so far.
45018cbf
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/11 09:02
Doing stereo mixing adaptively based on amplitude difference.
a2482ace
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/11 06:24
Conversion to 18-band split and fixed a few related bugs.
43e22481
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/10 12:26
oops, trying not to overflow the 64-bit max allocation.
f8dda66a
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/10 12:02
Tuning of the stereo mode, including conversion to the 18-band split.
a4833ffa
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/10 10:34
Stereo decoding working again (fixed a few issues in the encoder at the same
021478e2
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/09 10:31
More stereo work (doing mid-side quantisation of the energy)
95c59ea6
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/08 11:22
Updated the 18-band mode (mode0) to use the low-overlap MDCT
825ec60b
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/05 17:49
Completed low-overlap MDCT support by reducing the synthesis
2875f6bf
– Jean-Marc Valin <
[email protected]>
authored
on 2008/01/05 08:32
Now using 8 bands for the pitch gain, with a 128-entry codebook.
Next