Ricardo Constantino (:RiCON) authored 8 years ago and Jean-Marc Valin committed 8 years ago

Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>

This makes it possible to use folding rather than LCG noise in the second CELT
band (9.6 to 12 kHz) in hybrid mode.

The multistream encoder input has st->layout.nb_channels channels.
The number of channels actually encoded is st->layout.nb_streams +
st->layout.nb_coupled_streams, which may be fewer, so ideally it
would only analyze those when computing the optimal frame size, but
there is no code to do that currently.

Divide by the actual number of channels mixed

Adding a "channels" argument to compute_frame_energy() and is_digital_silence()

This avoids an overflow on extreme signals

We now declare that anything that would overflow is not stable enough

Thanks to Mark Harris for pointing it out.

Gives another ~2% speedup

Speeds up the encoder by about 1%

This makes the decoder ~2.5% faster on x86 because the stereo loop
takes the same processing time as one mono loop due to the dependency chain

Reordering the add with VERY_SMALL changes the dependencies cycle from 2 add + 1 mul
(11 cycles on haswell) to 1 add + 1 mul (8 cycles). This makes the entire decoder about
1.5% faster.

We used the SSE reciprocal square root instruction to vectorize the serch rather
than compare one at a time with multiplies. Speeds up the entire encoder by 8-10%.

No measurable speed change.

Speeds up encoding by another ~1-2%

Speeds up CELT encoding by around 5% on x86

(found through fuzzing)

No point in minimizing the rate too since it's almost constant.

Felicia Lim authored 8 years ago and Jean-Marc Valin committed 8 years ago

Signed-off-by: Jean-Marc Valin <jmvalin@jmvalin.ca>

Ricardo Constantino (:RiCON) authored 8 years ago and Ralph Giles committed 8 years ago

This lets projects which depend on opus build against our
test results.

Signed-off-by: Ralph Giles <giles@mozilla.com>