Only complexity 0-1 now uses the plain non-delayed-decision NSQ since its
quality is pretty bad. Previously 0-3 were using that plain NSQ.

Previously, the bandwidth detection could (e.g.) change the bandwidth
from fullband to superwideband, and the hysteresis would then cause
bandwidth to be stuck in superwideband.

Removes unused 64-bit version

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

Computes a pseudo SNR as the ratio between the running peak signal
energy to instantaneous noise energy. DTX is turned on only if this is
higher than a pre-determined threshold.

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

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

A frame is marked as valid for DTX if it contains noise or only digital
silence. As before, there is an overhang period of 200 ms and a maximum
consecutive DTX period of 400 ms.

If the new DTX cannot be used because of the complexity setting and
sampling frequency chosen, the SILK DTX will be used instead.

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

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

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

Koen Vos authored 9 years ago and Jean-Marc Valin committed 8 years ago

JMV: edited to be a little more conservative

Slightly slower on x86, about the same speed on ARMv7, should be faster on
DSPs.

Should be able to make better mode/bandwidth decisions when CBR/complexity
are involved.

The new attenuation is much smaller than previously.

New MLP doesn't attempt to classify silence as speech/music

This saves bits and makes more sense since alloc_trim_analysis()
mostly looks at the lower bands that are coded with SILK

Gradual downmix now happens in the 24 kb/s to 32 kb/s range

Turns out that even low bitrates benefit from VBR

No values outside of +/-1 detected now.

Jonathan Lennox authored 8 years ago and Timothy B. Terriberry committed 8 years ago

Notably, ARM floating-point build.

This maintains the invariant that we don't use later instruction
 sets if the OS claims an earlier one is not available.
However, it does not update configure to ensure that there is
 build support for all earlier instruction sets if NEON build
 support is enabled (though I am not aware of a build toolchain
 where this is actually a problem).

Signed-off-by: Timothy B. Terriberry <tterribe@xiph.org>