Gitiles
Code Review
Sign In
gerrit-public.fairphone.software
/
platform
/
external
/
libopus
/
0668b8ea5e6608101d6faeb28fa3a483c4a52475
0668b8e
fixed-point: converted normalise_bands (had to split implementation)
by Jean-Marc Valin
· 17 years ago
cefc5ca
Revert "compute_band_energies() merged with normalised_bands()"
by Jean-Marc Valin
· 17 years ago
f8eb420
Changed the pulse spreading rotations so that the number of iterations is
by Jean-Marc Valin
· 17 years ago
ba8e2f7
fixed-point: done converting quant_bands() and unquant_bands()
by Jean-Marc Valin
· 17 years ago
c813308
Add option to #include "custom_support.h"
by Jean-Marc Valin
· 17 years ago
c4a40c6
doing spreading function and gain quantisation in-place
by Jean-Marc Valin
· 17 years ago
6441379
fixed-point: part of stereo_mix() converted.
by Jean-Marc Valin
· 17 years ago
ab69667
oops, unb0rked testcelt output
by Jean-Marc Valin
· 17 years ago
9d312b9
Defining RADIX_TWO_ONLY removes all butterflies, except for radix 2 and 4.
by Jean-Marc Valin
· 17 years ago
f93747c
fixed-point: changed find_spectral_pitch() to use single-precision (16-bit) FFT.
by Jean-Marc Valin
· 17 years ago
4024444
nothing to see here
by Jean-Marc Valin
· 17 years ago
b624cdd
fixed-point: masking curve computation now converted. ***Fixed a bug in the
by Jean-Marc Valin
· 17 years ago
b0c93c6
oops, pgain_table now has half the number of elements
by Jean-Marc Valin
· 17 years ago
afa5eb3
fixed-point: quantised the pitch gain table.
by Jean-Marc Valin
· 17 years ago
9237669
fixed-point: converted the pitch gain quantisation, except for the codebook
by Jean-Marc Valin
· 17 years ago
a78dc94
Making 16-bit compilers happy
by Jean-Marc Valin
· 17 years ago
d9de593
Fixed a bunch of warnings
by Jean-Marc Valin
· 17 years ago
a1bc18a
compute_band_energies() merged with normalised_bands()
by Jean-Marc Valin
· 17 years ago
069c3de
Removed deprecated mode interface and added missing include
by Jean-Marc Valin
· 17 years ago
7137c26
fixed-point: exp_rotation() now fully converted, using an approximation of the
by Jean-Marc Valin
· 17 years ago
7e94cc4
fixed-point: converted denormalise_bands()
by Jean-Marc Valin
· 17 years ago
a31899f
added suffix to real fft. The celtclient can now be built with an unmodified
by Jean-Marc Valin
· 17 years ago
300cb5b
First attempt at adding a suffix to kiss_fft functions so we can have multiple
by Jean-Marc Valin
· 17 years ago
bc60f8b
making sure __GNUC_PREREQ() is only used when __GNUC__ is defined
by Jean-Marc Valin
· 17 years ago
0758689
pre-emphasis coef now a constant
by Jean-Marc Valin
· 17 years ago
3121327
A bunch of const qualifyers and a few comments
by Jean-Marc Valin
· 17 years ago
730c6e3
Applying the window properly in find_spectral_pitch()
by Jean-Marc Valin
· 17 years ago
ddb181b
A bit more reordering in stack memory allocation -- saved 1024 bytes on peak use
by Jean-Marc Valin
· 17 years ago
137ec8e
applying the pitch windowing directly in find_spectral_pitch()
by Jean-Marc Valin
· 17 years ago
05080b4
No longer storing the zero parts of the in[] array in the encoder
by Jean-Marc Valin
· 17 years ago
632f218
Pitch predictor now uses a larger range of offset by not being limited to the
by Jean-Marc Valin
· 17 years ago
0141723
compute_mdcts() no longer reads the part of the input that should be zeros
by Jean-Marc Valin
· 17 years ago
3e08a88
Only the (rising) overlap part of the window is stored now. No need to hold the
by Jean-Marc Valin
· 17 years ago
f28062f
Don't bother multiplying when the window's value is one or zero
by Jean-Marc Valin
· 17 years ago
227ad2f
cwrs32-test should be 16-bit clean now.
by Jean-Marc Valin
· 17 years ago
4c2787a
Laplace encoder now works with 16-bit frequencies
by Jean-Marc Valin
· 17 years ago
eb5a245
Hadn't realised the bitr* stuff wasn't needed.
by Jean-Marc Valin
· 17 years ago
7ff2cbc
oops, fixed the Laplace testcase
by Jean-Marc Valin
· 17 years ago
3794264
Saved 4 kB of stack usage in find_spectral_pitch() by doing the FFT in-place
by Jean-Marc Valin
· 17 years ago
7cdc5a3
Making it easier to debug laplace testcase
by Jean-Marc Valin
· 17 years ago
a556694
Stereo pitch search no longer requires twice the memory/complexity.
by Jean-Marc Valin
· 17 years ago
86cb7e8
oops. Another 16-bit fix
by Jean-Marc Valin
· 17 years ago
9e006b6
16-bit fix
by Jean-Marc Valin
· 17 years ago
d748cd5
Another C90-fying pass. Fixed some warnings in the process.
by Jean-Marc Valin
· 17 years ago
81b38c2
Moved mdct state and window to the mode so it can be shared between multiple
by Jean-Marc Valin
· 17 years ago
37131e3
Just realised half the twiddles in the real FFT weren't even used. Removing
by Jean-Marc Valin
· 17 years ago
8d4ac15
Saved on stack usage by changing the order of the allocation
by Jean-Marc Valin
· 17 years ago
9b9e986
Bit of memory reduction in the pitch search. More needed.
by Jean-Marc Valin
· 17 years ago
cac91ec
Fixed temp arrays that were allocated too large (no change in peak mem)
by Jean-Marc Valin
· 17 years ago
877b197
fixed-point: converted intra prediction and folding, unb0rked mixed-precision
by Jean-Marc Valin
· 17 years ago
2aaa0fe
fixed-point: overflow debugging now works again.
by Jean-Marc Valin
· 17 years ago
8600f69
Initial support for a managed stack/scratchpad. Still needs some work.
by Jean-Marc Valin
· 17 years ago
8b2a592
fixed-point: energy quantisation is entirely converted now.
by Jean-Marc Valin
· 17 years ago
adf87e2
fixed-point: more energy quantisation stuff.
by Jean-Marc Valin
· 17 years ago
47bf175
fixed-point: a few more bits converted from the energy quantisation. Almost done
by Jean-Marc Valin
· 17 years ago
68b02b1
fixed-point: copied the exp2 implementation from Speex, using it for dB2Amp()
by Jean-Marc Valin
· 17 years ago
2eaba8a
fixed-point: log approximation
by Jean-Marc Valin
· 17 years ago
9819347
fixed-point: unquant_energy_mono() has received the fixed-point code from
by Jean-Marc Valin
· 17 years ago
d031620
fixed-point: second check-point on quant_energy_mono()
by Jean-Marc Valin
· 17 years ago
43b3537
fixed-point: first check-point on quant_energy_mono() conversion
by Jean-Marc Valin
· 17 years ago
4e1328b
fixed-point: alg_quant() entirely converter -- may require revisiting...
by Jean-Marc Valin
· 17 years ago
8b158f5
fixed-point: (hopefully) last check-point for alg_quant() conversion
by Jean-Marc Valin
· 17 years ago
642ff94
fixed-point: The cross-products in alg_quant() are now all converted.
by Jean-Marc Valin
· 17 years ago
c9d606f
fixed-point: third alg_quant() checkpoint
by Jean-Marc Valin
· 17 years ago
ed9e423
fixed-point: second check-point in the conversion of alg_quant()
by Jean-Marc Valin
· 17 years ago
f675adc
fixed-point: First check-point in alg_quant() conversion
by Jean-Marc Valin
· 17 years ago
a02ca1e
fixed-point: Added a ROUND() operator, no real change to the code
by Jean-Marc Valin
· 17 years ago
6ba6978
Oops. Fixed a missing \ in a Makefile.am
by Jean-Marc Valin
· 17 years ago
79fa8b6
Just in case sizeof(char)!=1
by Jean-Marc Valin
· 17 years ago
5e17149
Removed some unused parameters and fixed a couple -W warnings
by Jean-Marc Valin
· 17 years ago
3ca9b1d
fixed-point: Moved sqrt and cos approximations to mathops.h
by Jean-Marc Valin
· 17 years ago
a847b77
fixed-point: done with mix_pitch_and_residual() though a bit of cleaning up
by Jean-Marc Valin
· 17 years ago
1ca0722
fixed-point: mix_pitch_and_residual() check-point #3
by Jean-Marc Valin
· 17 years ago
b50c541
fixed-point: another mix_pitch_and_residual() check-point
by Jean-Marc Valin
· 17 years ago
d17edd3
fixed-point: half-way converting mix_pitch_and_residual() -- just check-pointing
by Jean-Marc Valin
· 17 years ago
9d8d9b3
fixed-point: compression factor (alpha) now a 16-bit value (still internally
by Jean-Marc Valin
· 17 years ago
9fbb56a
fixed-point: converted pitch_quant_bands() -- that one was an easy one-liner
by Jean-Marc Valin
· 17 years ago
ff74e39
fixed-point: converted compute_pitch_gain() and removed the energy-based
by Jean-Marc Valin
· 17 years ago
3f382ca
Fixing pi again before another massive public outcry
by Jean-Marc Valin
· 17 years ago
387a20d
fixed-point: exp_rotation() mostly converted (still need to convert the cos/sin)
by Jean-Marc Valin
· 17 years ago
e40a19c
fixed-point: no float vars left (and nearly no float ops left) in celt.c
by Jean-Marc Valin
· 17 years ago
5d56183
fixed-point: log-energy for previous frame now a 16-bit value. This currently
by Jean-Marc Valin
· 17 years ago
1d6ad10
fixed-point: converted window to 16-bit value
by Jean-Marc Valin
· 17 years ago
4207438
fixed-point: pre/de-emphasis converted.
by Jean-Marc Valin
· 17 years ago
d4018c3
Reduced code duplication in vq.c by adding mix_pitch_and_residual() which
by Jean-Marc Valin
· 17 years ago
2f5ccf6
C90-fying ectest
by Jean-Marc Valin
· 17 years ago
7a455d4
Another 16-bit fix.
by Jean-Marc Valin
· 17 years ago
b30dd9e
fixed-point: got stereo to work again by fixing renormalise_bands()
by Jean-Marc Valin
· 17 years ago
2805a14
fixed-point: celt_pgain_t now a 16-bit value (Q15)
by Jean-Marc Valin
· 17 years ago
5f09ea5
celt_mask_t for masking curves
by Jean-Marc Valin
· 17 years ago
8835a4d
Removed useless masking curve for current frame and ignored Fs argument to
by Jean-Marc Valin
· 17 years ago
017d445
pitch gain is now celt_pgain_t
by Jean-Marc Valin
· 17 years ago
b60340f
fixed-point: band energy now a 32-bit value. It might have (barely) fix into
by Jean-Marc Valin
· 17 years ago
e901fe3
fixed-point: added a celt_ener_t type for band energy.
by Jean-Marc Valin
· 17 years ago
e046c18
Minus a bunch of warnings when enabling alloca()
by Jean-Marc Valin
· 17 years ago
e995c31
Put the channel count outside of the energy calculation. Increased the allowed
by Jean-Marc Valin
· 17 years ago
2fa8aff
fixed-point: celt_norm_t now a 16-bit value.
by Jean-Marc Valin
· 17 years ago
9a5f3d2
16-bit fixes and warning fixes
by Jean-Marc Valin
· 17 years ago
508de38
Trying to be nice with 16-bit chips.
by Jean-Marc Valin
· 17 years ago
65d79e4
Corrected the last digit of pi after massive public outcry :-)
by Jean-Marc Valin
· 17 years ago
Next »