 Jean-Marc Valin
|
18a380a7c2
Make it possible to ignore inverted phase stereo for downmix purposes
|
il y a 8 ans |
|
Jean-Marc Valin
|
6db0908db3
Bitstream change: prevent noise injection in hybrid mode
|
il y a 8 ans |
|
Jean-Marc Valin
|
12fb15f42b
Add configure option --enable-update-draft
|
il y a 8 ans |
 Mark Harris
|
7d0e33f7f1
Correct multistream variable frame size stride
|
il y a 8 ans |
|
Mark Harris
|
3421f6d2ba
Correct analysis downmix scaling factor
|
il y a 8 ans |
|
Jean-Marc Valin
|
9f3883a955
Properly process both channels for DTX
|
il y a 8 ans |
|
Jean-Marc Valin
|
096b2207ba
Make the DC rejection filter Q14 instead of Q15
|
il y a 8 ans |
|
Jean-Marc Valin
|
8ab3e80ec4
Fixes an overflow in LPC_inverse_pred_gain_QA()
|
il y a 8 ans |
|
Jean-Marc Valin
|
18335e8e4d
Removed a float operation that sneaked in in the fixed-point code
|
il y a 8 ans |
|
Jean-Marc Valin
|
b4d0df8dde
Oops, fixes the sign in dc_reject()
|
il y a 8 ans |
|
Jean-Marc Valin
|
adf94f2c61
Implementing compute_frame_energy() from celt_inner_prod() for float
|
il y a 8 ans |
|
Jean-Marc Valin
|
b7c7653941
Speeding up PVQ search by allocating even more pulses in the projection.
|
il y a 8 ans |
|
Jean-Marc Valin
|
fb19ba9126
Reducing the overhead due to dependency chains in dc_reject() for stereo
|
il y a 8 ans |
|
Jean-Marc Valin
|
9ac942660d
Reducing the dependency chain in dc_reject()
|
il y a 8 ans |
|
Jean-Marc Valin
|
11dba8d902
Apply deemphasis to both channels in the same loop for the simple case
|
il y a 8 ans |
|
Jean-Marc Valin
|
cd159fd1ec
Making gcc use SSE directly for float->int conversion when available
|
il y a 8 ans |
|
Jean-Marc Valin
|
9f7e502e06
Reducing dependencies in deemphasis()
|
il y a 8 ans |
|
Jean-Marc Valin
|
7f3fb20185
Fixes problem with M_PI
|
il y a 8 ans |
|
Jean-Marc Valin
|
2ff6556f1f
Making stereo_itheta() use the same atan2() approximation as tonality_analysis()
|
il y a 8 ans |
|
Jean-Marc Valin
|
76674feae2
SSE2 implementation of the PVQ search
|
il y a 8 ans |
|
Jean-Marc Valin
|
e806d6a741
Making signx[] an int in alg_quant() and removes unnecessary sign copying
|
il y a 8 ans |
|
Jean-Marc Valin
|
c7bbc3e31f
Speeding up PVQ using unlikely() and moving first position out of the loop
|
il y a 8 ans |
|
Jean-Marc Valin
|
09b5352928
Getting gcc to use cmovs rather than branches in alg_quant()
|
il y a 8 ans |
|
Jean-Marc Valin
|
416611cfa1
biasing quantization
|
il y a 8 ans |
|
Jean-Marc Valin
|
a01feba028
Fixes "mixed declarations and code"
|
il y a 8 ans |
|
Jean-Marc Valin
|
10b2978de3
Making it clearer to Coverity that nStates cannot exceed NLSF_QUANT_DEL_DEC_STATES
|
il y a 8 ans |
|
Jean-Marc Valin
|
928dcffd98
Weighting theta rdo based on channel energy
|
il y a 8 ans |
|
Jean-Marc Valin
|
5da0498a3e
Fixes overflow in CNG
|
il y a 8 ans |
|
Jean-Marc Valin
|
abe8b78513
Don't do theta RDO on intensity-stereo-coded bands
|
il y a 8 ans |
|
Jean-Marc Valin
|
eb61c31c14
Saving the state rather than re-compute the best option
|
il y a 8 ans |
