This patch does 4 things, all of which interact and thus it
woudln't be possible to commit them separately without causing
either quality regressions or assertion failures.
Fate comparison targets don't all reflect improvements in
quality, yet listening tests show substantially improved quality
and stability.
1. Increase SF range utilization.
The spec requires SF delta values to be constrained within the
range -60..60. The previous code was applying that range to
the whole SF array and not only the deltas of consecutive values,
because doing so requires smarter code: zeroing or otherwise
skipping a band may invalidate lots of SF choices.
This patch implements that logic to allow the coders to utilize
the full dynamic range of scalefactors, increasing quality quite
considerably, and fixing delta-SF-related assertion failures,
since now the limitation is enforced rather than asserted.
2. PNS tweaks
The previous modification makes big improvements in twoloop's
efficiency, and every time that happens PNS logic needs to be
tweaked accordingly to avoid it from stepping all over twoloop's
decisions. This patch includes modifications of the sort.
3. Account for lowpass cutoff during PSY analysis
The closer PSY's allocation is to final allocation the better
the quality is, and given these modifications, twoloop is now
very efficient at avoiding holes. Thus, to compute accurate
thresholds, PSY needs to account for the lowpass applied
implicitly during twoloop (by zeroing high bands).
This patch makes twoloop set the cutoff in psymodel's context
the first time it runs, and makes PSY account for it during
threshold computation, making PE and threshold computations
closer to the final allocation and thus achieving better
subjective quality.
4. Tweaks to RC lambda tracking loop in relation to PNS
Without this tweak some corner cases cause quality regressions.
Basically, lambda needs to react faster to overall bitrate
efficiency changes since now PNS can be quite successful in
enforcing maximum bitrates, when PSY allocates too many bits
to the lower bands, suppressing the signals RC logic uses to
lower lambda in those cases and causing aggressive PNS.
This tweak makes PNS much less aggressive, though it can still
use some further tweaks.
Also update MIPS specializations and adjust fuzz
Also in lavc/mips/aacpsy_mips.h: remove trailing whitespace
As noted in a comment, pe.min in the reference encoder
is centered around current pe. The bit reservoir algo
needs pe.min to be a local minimum, because it can only
account for local PE variations. If it's set to a global
minimum as was being done, bit reservoir logic doesn't
work as efficiently.
This patch tries to forget old minimums and converge to
a local minimum without losing the stability of the
previous solution. Listening tests until now suggest this
solves numerous RC issues.
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.
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).
Avoid clipping due to quantization noise to produce audible
artifacts, by detecting near-clipping signals and both attenuating
them a little and encoding escape-encoded bands (usually the
loudest) rounding towards zero instead of nearest, which tends to
decrease overall energy and thus clipping.
Currently fate tests measure numerical error so this change makes
tests using asynth (which are near clipping) report higher error
not less, because of window attenuation. Yet, they sound better,
not worse (albeit subtle, other samples aren't subtle at all).
Only measuring psychoacoustically weighted error would make for
a representative test, so that will be left for a future patch.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
This commit modifies 02dbed6 to use band->active_lines to better gauge how much information is contained within a single band and thus allow the perceptual noise subsitution to more accurately determine which bands to code as noise. The spread[w+g] used before this patch behaved more like a low-pass filter for PNS band_types, which could mistakingly mark some low frequency bands as noise.
Reviewed-by: Claudio Freire <klaussfreire@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This commit adds the energy spread to the struct for each band and removes 2 unused fields.
distortion and perceptual_weight were not referenced in any file nor were they set to any value,
so it was safe to remove them. The energy spread is currently only used in the aac psy model.
It's defined as being proportional to the tonality of each band.
Reviewed-by: Claudio Freire <klaussfreire@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
The minimum of the ath(x, ATH_ADD) function depends on ATH_ADD.
This patch uses the first order approximation to determine it.
For ATH_ADD = 4 this results in the value at 3407.06812 (-5.24241638)
not the one at 3410 (-5.24237967).
CC: libav-stabl@libav.org
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
The minimum of the ath(x, ATH_ADD) function depends on ATH_ADD.
This patch uses the first order approximation to determine it.
For ATH_ADD = 4 this results in the value at 3407.06812 (-5.24241638)
not the one at 3410 (-5.24237967).
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
Approved-by: Claudio Freire <klaussfreire@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
If band->thr is 0.0f, the division is undefined, making norm_fac not a
number or infinity, which causes psy_band->threshold to become NaN.
This is passed on to other variables until it finally reaches
sce->sf_idx and is converted to an integer (-2147483648).
This causes a segmentation fault when it is used as array index.
Signed-off-by: Andreas Cadhalpun <Andreas.Cadhalpun@googlemail.com>
Reviewed-by: Claudio Freire <klaussfreire@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This is a small change, but it does have a big impact on bit allocation.
all the regressions marked in the report have no audible
difference (I didn't check them all though), but the improvements can
be heard.
This affects mostly high bit rates. It's related to issue #2686.
In the report, A is the patched version, B is unpatched, all
comparisons show deltas in the form (A-B), so a positive pSNR delta
means a better quality in the patched version, and negative a
regression. Regressions are only considered for pSNR deltas below
-1db, they're considered serious below -6db.
All measurements were done with tiny_psnr.
The summary of the report inline for quick reading:
Files: 58
Bitrates: 6
Tests: 347
Serious Regressions: 0 (0%)
Regressions: 10 (2%)
Improvements: 54 (15%)
Big improvements: 26 (7%)
Worst regression - sine_tester.flac - 384k
- StdDev: 1.68 pSNR: -3.05 maxdiff: -178.00
Best improvement - 07 - Bound.flac - 384k
- StdDev: -1700.05 pSNR: 20.64 maxdiff: -29595.00
Average - StdDev: -55.67 pSNR: 1.20 maxdiff: -1593.00
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
This patch changes existing mathematical functions with faster
ones. Speeds up encoding more than 10%. Tested on x86 and
MIPS platforms.
Signed-off-by: Bojan Zivkovic <bojan@mips.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
There is still are still a few sections missing relating to TNS (not present)
and mid/side (contains other bugs).
Overall this improves quality, and vastly improves rate-control.
Signed-off-by: Martin Storsjö <martin@martin.st>
3GPP:
Remove ffac from and move min_snr out of AacPsyBand.
Rearrange AacPsyCoeffs to make it easier to implement energy spreading.
Rename the band[] array to bands[]
Copy energies and thresholds at the end of analysis.
LAME:
Use a loop instead of an if chain in LAME windowing.
The 3GPP spec uses the following calculation for high spreading:
thr'_spr = max(thr_scaled, s_h(n) * thr_scaled(n-1))
where, n is defined as the current band, and s_h() is defined as "[...] the
distance of adjacent bands in Bark and a constant slope that is 15 dB/Bark
[...]". This is a little ambiguous as you would assume you want the Bark
width of the previous band for this calculation. However, this assumption
appears to be incorrect, and you really want the Bark width of the current
band. Coincidentally this is exactly what the spec calls for! =P
This noticeably improves Tom's Diner at low bitrates (I tested at 64kbps,
with mid/side disabled).
Patch by: Nathan Caldwell <saintdev@gmail.com>
Originally committed as revision 25622 to svn://svn.ffmpeg.org/ffmpeg/trunk
This greatly improves bitrate handling. You will now get within a few
kbps of your requested bitrate instead of 20-40kbps higher.
There is absolutely no analog to this line in the 3GPP spec, that I
can find.
patch by Nathan Caldwell saintdev (at) gmail
Originally committed as revision 25589 to svn://svn.ffmpeg.org/ffmpeg/trunk
Removing the modification vastly improves quality (at a slight bitrate
cost) for some samples. castanets.wav is a good example. The closest
equivalent I see to the modification in the 3GPP spec is a similar
modification (over a specific frequency range) when TNS is used.
This also changes the threshold-in-quiet calculation to match the
3GPP spec.
patch by Nathan Caldwell saintdev (at) gmail
Originally committed as revision 25588 to svn://svn.ffmpeg.org/ffmpeg/trunk
According to the 3GPP spec:
"Thus the pre-echo control is inactive for the first short window (but
not all short windows in a short frame) after a start block and for
all frames with a stop window sequence."
Currently, pre-echo control is only run when the current frame is not
a short frame, and the previous frame is not a short frame.
patch by Nathan Caldwell saintdev (at) gmail
Originally committed as revision 25587 to svn://svn.ffmpeg.org/ffmpeg/trunk
Claudio Freire
Vittorio Giovara
Rostislav Pehlivanov
Michael Niedermayer
Andreas Cadhalpun
Claudio Freire
Diego Biurrun
Bojan Zivkovic
Nathan Caldwell
Mans Rullgard