View Full Version : Does Trellis affect VAQ?
mahsah
29th February 2008, 02:33
Sorry, I'm a newbie with x264, and I just want to reduce banding and blocking in dark areas in an anime I am encoding. The new VAQ seems to be a good way to do this, but since I normally use the AE-Maxquality profile in megui (which uses Trellis 2), will this effect the VAQ at all (they are both quantization methods, right?)
Also what bitrate would be good for a 720x480 anamorphic encode with VAQ enabled? I understand it is better to have a bit more bitrate with VAQ?
Dark Shikari
29th February 2008, 02:37
VAQ chooses the quantizer--trellis performs the quantization. They're two separate things, and neither will conflict with the other.
mahsah
29th February 2008, 02:40
Alright, thanks. Also you wrote the VAQ right? Great job, I get awesome results with it.
Dark Shikari
29th February 2008, 02:41
Alright, thanks. Also you wrote the VAQ right? Great job, I get awesome results with it.Correct, and welcome :)
bluebebe
26th March 2008, 22:38
how can i enable VAQ in the encodings? uses this the params --aq-strength and threshold?
Dark Shikari
26th March 2008, 22:45
how can i enable VAQ in the encodings? uses this the params --aq-strength and threshold?If you have a patched build, AQ is on by default. Strength can be raised with --aq-strength.
bluebebe
27th March 2008, 18:12
thanks for the hint, and "--aq-strength 0.0" will completly disable the adaptive quantizer of x264 = means best quality results at cost of higher bitrate?
Dark Shikari
27th March 2008, 18:19
thanks for the hint, and "--aq-strength 0.0" will completly disable the adaptive quantizer of x264 = means best quality results at cost of higher bitrate?Huh? Generally AQ is beneficial for quality.
Sharktooth
27th March 2008, 18:20
the old AQ was sacrificing quality to remove blocking in some areas, Dark Shikari's new V(ariance)AQ is completely different and helps quality in almost every situation.
bluebebe
27th March 2008, 18:52
i learned in xvid encoding lession that AQ will compress complete dark or white areas with a lower bitrate and i thought it was so in x264 too. it is not so okay, what is the min / max range for the --aq-strength param and esp. the best setting which will produce the best results? it is better to use a low strength under 1.0 or is it better to set a higher number of it?
gav1577
27th March 2008, 19:06
@bluebebe I would not recommend going over --aq-strength 1.0 personally i leave it at 1.0 always
i find best at this level for combating the effects of banding :)
Sharktooth
27th March 2008, 20:11
i learned in xvid encoding lession that AQ will compress complete dark or white areas with a lower bitrate and i thought it was so in x264 too. it is not so okay, what is the min / max range for the --aq-strength param and esp. the best setting which will produce the best results? it is better to use a low strength under 1.0 or is it better to set a higher number of it?
there are several kinds of possible "adaptive" quantizations, ppl keep inventing new algos every day... adaptive is just a generic term. xvid's AQ is lumimasking... that's completely different from Haali's AQ and Dark Shikari's Variance AQ.
TheRyuu
27th March 2008, 20:13
there are different kinds of possible "adaptive" quantizations. adaptive is just a generic term. xvid's AQ is lumimasking... that's completely different from Haali's AQ and Dark Shikari's Variance AQ.
Unless you're using an xvid built with VAQ.... :)
Sharktooth
27th March 2008, 20:14
of course
Dark Shikari
27th March 2008, 20:14
Unless you're using an xvid built with VAQ.... :)Though even then its quite a bit different both because of the quantizer field smoothing Xvid uses, along with the fact that MPEG-4 ASP's AQ ability is somewhat restricted.
vBulletin® v3.8.11, Copyright ©2000-2026, vBulletin Solutions Inc.