delta/opus.git, branch exp_lbr_tune gitlab.xiph.org: xiph/opus.git Adds some smoothing to the energy quantization 2016-07-19T20:11:50+00:00 Jean-Marc Valin jmvalin@jmvalin.ca 2016-07-04T05:06:11+00:00 6fccb4b6159c9f54be8db8daaaa8d305bdfcbc33 When the energy is stable, we slightly bias energy quantization towards the previous error to make the gain more stable (a constant offset is better than fluctuations). We reduce the bitrate by about 0.2% to 1% at low bitrate for the same quality. 
When the energy is stable, we slightly bias energy quantization towards
the previous error to make the gain more stable (a constant offset is
better than fluctuations).

We reduce the bitrate by about 0.2% to 1% at low bitrate for the same quality.
Tune SILK rate for hybrid with FEC enabled 2016-07-19T20:11:50+00:00 Jean-Marc Valin jmvalin@jmvalin.ca 2016-06-21T17:37:29+00:00 b3c9e79a8291e85c0cd5f91b55d1e856edd95ac1 






Moving FEC decision to the Opus encoder level (from SILK level)
2016-07-19T20:11:50+00:00

Jean-Marc Valin
jmvalin@jmvalin.ca

2016-06-21T16:38:25+00:00

2b011b64d8ca5f3f7b26806aeb77939ff7703c06












Taking loss rate into account in compute_equiv_rate()
2016-07-19T20:11:50+00:00

Jean-Marc Valin
jmvalin@jmvalin.ca

2016-06-20T22:54:06+00:00

099fc05ebf443e2bcc8c5ab9c627c86ec1e743e0












Retuning the tf_analysis() lambda
2016-07-19T20:11:50+00:00

Jean-Marc Valin
jmvalin@jmvalin.ca

2016-06-08T19:19:03+00:00

7780d4a6b16822498723235c7e34d9c0f62809fb

Increasing the value at low rate seems to help a bit.





Increasing the value at low rate seems to help a bit.







Fixes equiv_rate for CBR
2016-07-19T20:11:50+00:00

Jean-Marc Valin
jmvalin@jmvalin.ca

2016-06-08T06:33:56+00:00

1484591656533c36b67615f7dd89e82c467b5f14












Minor TF cleanup (tf_sum was useless), plus comments
2016-07-19T20:11:50+00:00

Jean-Marc Valin
jmvalin@jmvalin.ca

2016-06-08T06:15:21+00:00

2ca6df03df62154c9b76220a94108b56e4dfb34b












Use SPREAD_AGGRESSIVE on non-transient hybrid frames
2016-07-19T20:11:50+00:00

Jean-Marc Valin
jmvalin@jmvalin.ca

2016-06-06T17:43:13+00:00

f705e9b5bd0b4c60b9934962aef7af3c0865272f












Don't use too much temporal resolution on hybrid frames at low rate
2016-07-19T20:11:50+00:00

Jean-Marc Valin
jmvalin@jmvalin.ca

2016-06-06T20:54:29+00:00

8229f07d3bfbef53f1d89cfd371b4deed6f92227

Otherwise, we risk having "temporal holes" in the HF that anti-collapse
can't always fill in.





Otherwise, we risk having "temporal holes" in the HF that anti-collapse
can't always fill in.







Disable patch_transient_decision() on hybrid since it's causing artifacts
2016-07-19T20:11:50+00:00

Jean-Marc Valin
jmvalin@jmvalin.ca

2016-06-06T20:52:50+00:00

92d2492bfdbea8267faf0e6a6fbba09a5ffa7b1c

It would trigger on the second frame of "S"s, causing holes in the spectrum





It would trigger on the second frame of "S"s, causing holes in the spectrum