 Jean-Marc Valin
|
80232ff51f
CBR: set gains to their previons values where we're busting the budget
|
8 years ago |
|
Jean-Marc Valin
|
e0d66856cd
Set pulses to zero if we can't meet rate target
|
8 years ago |
|
Jean-Marc Valin
|
291fd14eae
Now that we're preventing huge gains, we no longer need to pretend there's
|
8 years ago |
|
Jean-Marc Valin
|
9b89223ac5
CBR: lock the gain on a subframe when the number of pulses stops going down
|
8 years ago |
 Koen Vos
|
2dffa214c1
merge back in the LTP gain safety limiter
|
8 years ago |
|
Jean-Marc Valin
|
5025c296e9
CBR: Using the same gain increase for fixed-point as for float
|
8 years ago |
|
Jean-Marc Valin
|
b0f625b1fb
Oops, edited the wrong comment in the previous commit
|
8 years ago |
|
Jean-Marc Valin
|
5372bdabbe
Re-tuning compute_equiv_rate()
|
8 years ago |
|
Jean-Marc Valin
|
b19c474bba
Getting rid of the SILK penalty for hybrid CBR.
|
8 years ago |
|
Jean-Marc Valin
|
fbdbafe036
Improving CBR for hybrid mode
|
8 years ago |
|
Jean-Marc Valin
|
8ea37b0b03
Making the encoder more aggressive at using wider bandwidths
|
8 years ago |
|
Koen Vos
|
208e4bf1bd
comment about reduced codelength component in LTP quantizer error metric
|
8 years ago |
|
Koen Vos
|
f13d512449
avoid double if statement for finding best quantized LTP coefs
|
8 years ago |
|
Koen Vos
|
3ede0a1ee6
increase NSQ decision delay from 32 to 40
|
8 years ago |
|
Koen Vos
|
17cee886e4
reduced noise floor in noise shaping analysis by 2 dB
|
8 years ago |
|
Koen Vos
|
58e3070f8d
increase order of noise shaping filter
|
8 years ago |
|
Koen Vos
|
50417c22e1
more conservative scaling of LTP corrs; remove assert that checks for negative residual energy; discard CB entries leading to such negative energies
|
8 years ago |
|
Koen Vos
|
44423724ff
Clean up: alignment of comments
|
9 years ago |
|
Koen Vos
|
e7a53f74d5
Clean up: replace tabs by spaces
|
9 years ago |
|
Koen Vos
|
8668661ebf
removed prefilter
|
9 years ago |
|
Koen Vos
|
69f247f421
slight clean up
|
9 years ago |
|
Koen Vos
|
d01175b56a
undo some loop unrolling
|
9 years ago |
|
Koen Vos
|
804ef131bf
pitch analysis (in FIX) now scales the input down at the start, instead of at every stage
|
9 years ago |
|
Koen Vos
|
81060bb4d1
minor clean up of pointer offsets
|
9 years ago |
|
Koen Vos
|
79a28f4ef0
simplified computation of LTP coefs
|
9 years ago |
 Felicia Lim
|
48c78d80ed
Attenuate SILK PLC gain only for unvoiced speech
|
8 years ago |
|
Jean-Marc Valin
|
a16fb953a7
We should now be able to be less conservative when using CBR
|
8 years ago |
|
Jean-Marc Valin
|
1dc5380727
Makes the encoder more aggressive about meeting the rate target
|
8 years ago |
|
Jean-Marc Valin
|
262a2f499b
Using more fine-grained complexity settings for SILK
|
8 years ago |
|
Jean-Marc Valin
|
10fdd00c56
Fixing bandwidth selection hysteresis
|
8 years ago |
