These aren't quite as helpful as the ones in 8bpp, since over there,
we can use pmulhrsw, but here the coefficients have too many bits to
be able to take advantage of pmulhrsw. However, we can still skip
cols for which all coefs are 0, and instead just zero the input data
for the row itx. This helps a few % on overall decoding speed.
It was useful to (accidentally?) spot an overflow in the column pass
of the x86 simple_idct10 implementation.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
omse goes from 0.03060703 (which fails for dct-test) to 0.01663750.
This also actually improve the error of decoding the sample generated
by fate-vsynth3-dnxhd1080i-10bit using simple_idct10 to FAANI, which
goes (when resampled to yuv422p) from:
stddev: 0.06 PSNR: 72.28 MAXDIFF: 1
to identical.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Includes escapes that should now be supported and a few features not yet
fully supported, like comments, regions, classes, ruby, and lang.
All were tested with https://quuz.org/webvtt/ for validation, except
regions because the validator doesn't support them yet, and I couldn't
find any other way to validate WebVTT.
Signed-off-by: Ricardo Constantino <wiiaboo@gmail.com>
It was merged with the iff_ilbm decoder in commit
929a24efff.
Define AV_CODEC_ID_IFF_BYTERUN1 as AV_CODEC_ID_IFF_ILBM for API
compatibility.
Reviewed-by: Ronald S. Bultje <rsbultje@gmail.com>
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
This finalizes merging of the work in the patches in ticket #2686.
Improvements to twoloop and RC logic are extensive.
The non-exhaustive list of twoloop improvments includes:
- Tweaks to distortion limits on the RD optimization phase of twoloop
- Deeper search in twoloop
- PNS information marking to let twoloop decide when to use it
(turned out having the decision made separately wasn't working)
- Tonal band detection and priorization
- Better band energy conservation rules
- Strict hole avoidance
For rate control:
- Use psymodel's bit allocation to allow proper use of the bit
reservoir. Don't work against the bit reservoir by moving lambda
in the opposite direction when psymodel decides to allocate more/less
bits to a frame.
- Retry the encode if the effective rate lies outside a reasonable
margin of psymodel's allocation or the selected ABR.
- Log average lambda at the end. Useful info for everyone, but especially
for tuning of the various encoder constants that relate to lambda
feedback.
Psy:
- Do not apply lowpass with a FIR filter, instead just let the coder
zero bands above the cutoff. The FIR filter induces group delay,
and while zeroing bands causes ripple, it's lost in the quantization
noise.
- Experimental VBR bit allocation code
- Tweak automatic lowpass filter threshold to maximize audio bandwidth
at all bitrates while still providing acceptable, stable quality.
I/S:
- Phase decision fixes. Unrelated to #2686, but the bugs only surfaced
when the merge was finalized. Measure I/S band energy accounting for
phase, and prevent I/S and M/S from being applied both.
PNS:
- Avoid marking short bands with PNS when they're part of a window
group in which there's a large variation of energy from one window
to the next. PNS can't preserve those and the effect is extremely
noticeable.
M/S:
- Implement BMLD protection similar to the specified in
ISO-IEC/13818:7-2003, Appendix C Section 6.1. Since M/S decision
doesn't conform to section 6.1, a different method had to be
implemented, but should provide equivalent protection.
- Move the decision logic closer to the method specified in
ISO-IEC/13818:7-2003, Appendix C Section 6.1. Specifically,
make sure M/S needs less bits than dual stereo.
- Don't apply M/S in bands that are using I/S
Now, this of course needed adjustments in the compare targets and
fuzz factors of the AAC encoder's fate tests, but if wondering why
the targets go up (more distortion), consider the previous coder
was using too many bits on LF content (far more than required by
psy), and thus those signals will now be more distorted, not less.
The extra distortion isn't audible though, I carried extensive
ABX testing to make sure.
A very similar patch was also extensively tested by Kamendo2 in
the context of #2686.
Currently only 2 profiles are evaluated because they are the only 2
with distributed test sequences.
- CID 1260: YUV 4:2:2 10 bits with block-adaptive interlace coding,
from ticket 4876;
- CID 1270: YUV 4:4:4 10 bits (HR), 1920x839, from ticket 4581.
They were generated from the ticket sequences by running the
following kind of command-line;
ffmpeg -i $INPUT -an -sn -vcodec copy -vframes 1 -y $OUTPUT.mov
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
The current one, while correct, does not yield the best possible
results. The specificiations suggest another formula, which results
in quality gains in the decoded output from fate tests. This
justifies changing said formula.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
The System V ABI on x86-64 specifies that the al register contains an upper
bound of the number of arguments passed in vector registers when calling
variadic functions, so we aren't allowed to clobber it.
checkasm_fail_func() is a variadic function so also zero al before calling it.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Tested functions are internally kept in a binary search tree for efficient
lookups. The downside of the current implementation is that the tree quickly
becomes unbalanced which causes an unneccessary amount of comparisons between
nodes. Improve this by changing the tree into a self-balancing left-leaning
red-black tree with a worst case lookup/insertion time complexity of O(log n).
Significantly reduces the recursion depth and makes the tests run around 10%
faster overall. The relative performance improvement compared to the existing
non-balanced tree will also most likely increase as more tests are added.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
The System V ABI on x86-64 specifies that the al register contains an upper
bound of the number of arguments passed in vector registers when calling
variadic functions, so we aren't allowed to clobber it.
checkasm_fail_func() is a variadic function so also zero al before calling it.
Tested functions are internally kept in a binary search tree for efficient
lookups. The downside of the current implementation is that the tree quickly
becomes unbalanced which causes an unneccessary amount of comparisons between
nodes. Improve this by changing the tree into a self-balancing left-leaning
red-black tree with a worst case lookup/insertion time complexity of O(log n).
Significantly reduces the recursion depth and makes the tests run around 10%
faster overall. The relative performance improvement compared to the existing
non-balanced tree will also most likely increase as more tests are added.
This patch tweaks search_for_pns to be both more
aggressive and more careful when applying PNS. On
the one side, it will again try to use PNS on zero
(or effectively zero) bands. For this, both zeroes
and band_type have to be checked (some ZERO bands
aren't marked in zeroes). On the other side, a more
accurate rate-distortion measure avoids using PNS
where it would cause audible distortion.
Also fixed a small bug in the computation of freq
that caused PNS usage on low-frequency bands during
8-short windows. This allows re-enabling PNS during
8-short.
The sample position is made weird and non-nominal to force catching
such issues as default values or specialized operations hiding
issues in corner cases.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This patch modifies the encode frame function to
retry encoding the frame when the resulting bit count
is too far off target, but only adjusting lambda
in small, incremental step. It also makes the logic
more conservative - otherwise it will contend with
bit reservoir-related variations in bit allocation,
and result in artifacts when frame have to be truncated
(usually at high bit rates transitioning from low
complexity to high complexity).
The randomize_buffer() implementation assures that "most of the time",
we'll do a good mix of wide16/wide8/hev/regular/no filters for complete
code coverage. However, this is not mathematically assured because that
would make the code either much more complex, or much less random.
This patch refactors the AAC coders to reuse code
between the MIPS port and the regular, portable C code.
There were two main functions that had to use
hand-optimized versions of quantization code:
- search_for_quantizers_twoloop
- codebook_trellis_rate
Those two were split into their own template header
files so they can be inlined inside both the MIPS port
and the generic code. In each context, they'll link
to their specialized implementations, and thus be
optimized by the compiler.
This approach I believe is better than maintaining
several copies of each function. As past experience has
proven, having to keep those in sync was error prone.
In this way, they will remain in sync by default.
Also, an implementation of the dequantized output
argument for the optimized quantize_and_encode
functions is included in the patch. While the current
implementation of search_for_pred still isn't using
it, future iterations of main prediction probably will.
It should not imply any measurable performance hit while
not being used.
Zhang Rui
Ronald S. Bultje
Christophe Gisquet
Ricardo Constantino
Rostislav Pehlivanov
Andreas Cadhalpun
Claudio Freire
Stefano Sabatini
James Almer
Henrik Gramner
Hendrik Leppkes
Vittorio Giovara
Michael Niedermayer
Rodger Combs