 Jean-Marc Valin
|
48033bcd79
Include config.h in mlp.c
|
8 年之前 |
|
Jean-Marc Valin
|
0342071379
Adds some smoothing to the energy quantization
|
8 年之前 |
|
Jean-Marc Valin
|
bf4acba483
Tune SILK rate for hybrid with FEC enabled
|
8 年之前 |
|
Jean-Marc Valin
|
77c8b41f9b
Moving FEC decision to the Opus encoder level (from SILK level)
|
8 年之前 |
|
Jean-Marc Valin
|
e600b72dd1
Taking loss rate into account in compute_equiv_rate()
|
8 年之前 |
|
Jean-Marc Valin
|
6b98f15849
Retuning the tf_analysis() lambda
|
9 年之前 |
|
Jean-Marc Valin
|
02bdc22e5b
Fixes equiv_rate for CBR
|
9 年之前 |
|
Jean-Marc Valin
|
717a036991
Minor TF cleanup (tf_sum was useless), plus comments
|
9 年之前 |
|
Jean-Marc Valin
|
0934d5d44b
Use SPREAD_AGGRESSIVE on non-transient hybrid frames
|
9 年之前 |
|
Jean-Marc Valin
|
4d4eb59e3a
Don't use too much temporal resolution on hybrid frames at low rate
|
9 年之前 |
|
Jean-Marc Valin
|
631dc27d61
Disable patch_transient_decision() on hybrid since it's causing artifacts
|
9 年之前 |
|
Jean-Marc Valin
|
00f4df3694
CBR: set gains to their previons values where we're busting the budget
|
9 年之前 |
|
Jean-Marc Valin
|
54c6703079
Set pulses to zero if we can't meet rate target
|
9 年之前 |
|
Jean-Marc Valin
|
158a051c8a
Now that we're preventing huge gains, we no longer need to pretend there's
|
9 年之前 |
|
Jean-Marc Valin
|
19bb2dbe58
CBR: lock the gain on a subframe when the number of pulses stops going down
|
9 年之前 |
 Koen Vos
|
b9b738e78e
merge back in the LTP gain safety limiter
|
9 年之前 |
|
Jean-Marc Valin
|
80d1a6a3fb
CBR: Using the same gain increase for fixed-point as for float
|
9 年之前 |
|
Jean-Marc Valin
|
06b232f3da
Oops, edited the wrong comment in the previous commit
|
9 年之前 |
|
Jean-Marc Valin
|
bd3a85e64c
Re-tuning compute_equiv_rate()
|
9 年之前 |
|
Jean-Marc Valin
|
f695b0121e
Getting rid of the SILK penalty for hybrid CBR.
|
9 年之前 |
|
Jean-Marc Valin
|
43a91a28c1
Improving CBR for hybrid mode
|
9 年之前 |
|
Jean-Marc Valin
|
6cacbfe632
Making the encoder more aggressive at using wider bandwidths
|
9 年之前 |
|
Koen Vos
|
0d3e6df03d
comment about reduced codelength component in LTP quantizer error metric
|
9 年之前 |
|
Koen Vos
|
8a47532387
avoid double if statement for finding best quantized LTP coefs
|
9 年之前 |
|
Koen Vos
|
df32cf8e02
increase NSQ decision delay from 32 to 40
|
9 年之前 |
|
Koen Vos
|
195ddd8420
reduced noise floor in noise shaping analysis by 2 dB
|
9 年之前 |
|
Koen Vos
|
dccc568718
increase order of noise shaping filter
|
9 年之前 |
|
Koen Vos
|
e4aff1ddb3
more conservative scaling of LTP corrs; remove assert that checks for negative residual energy; discard CB entries leading to such negative energies
|
9 年之前 |
|
Koen Vos
|
879a91a80f
Clean up: alignment of comments
|
9 年之前 |
|
Koen Vos
|
a1bbe2988b
Clean up: replace tabs by spaces
|
9 年之前 |
