View Full Version : Psychovisual setting for best looking video
Element
22nd April 2002, 12:40
I've tried it normal with Rush Hour 2 (91 minutes BTR 991) and Final Fantasy The Spirits Within (101 minutes BTR 823) and the quality is excellent. After, I've encoded Men of Honor (more than two hours) with strong and normal, and the best option was normal, more quality than strong (or I think so) and the same size.
Anyone have tried with light? :confused: IŽve heard the size is bigger, but is bigger also the quality? IŽll continue with my own experiments :cool:
Fox Mulder
23rd April 2002, 07:38
My results: using "light", "normal" or "strong" psychovisuals with DivX 5.0 Pro (in 100% Quality Based, the way you are supossed to test a codec), the results are the same size(byte identical) and with artifacting; and in 5.0.1 Pro the results are still identical in the "normal" or "strong" settings(smaller size than "light"), but without the artifacts.
Selur
23rd April 2002, 17:37
Hmm,.. I did some test a whiel back with psychodelic and quality didn't differ, at least I couldn't see a difference,but strong created a something like 0.5-1% smaller filesize,... (tested in quality mode with a fixed quantizer of 2 and 3)
Cu Selur
Nazgul
23rd April 2002, 19:21
Originally posted by Fox Mulder
My results: using "light", "normal" or "strong" psychovisuals with DivX 5.0 Pro (in 100% Quality Based, the way you are supossed to test a codec), the results are the same size(byte identical) and with artifacting; and in 5.0.1 Pro the results are still identical in the "normal" or "strong" settings(smaller size than "light"), but without the artifacts.
Call me crazy, but quality-based mode doesn't strike me as the best way to test psychovisual enhancements. Psychovisuals are for getting better quality out of a target bitrate by distributing the bits among the image based on how the human eye will perceive that image. Therefore, encoding at a level where it's already putting as many bits as it needs in every part of the scene wouldn't make sense. I would think that a better way to test would be to use 1 or 2-pass bitrate mode, and use a bitrate low enough to cause some noticable image degradation without psychovisual enhancements. That would give you a way to tell if the psychovisuals have any effect.
Nazgul
llemor
24th April 2002, 02:50
I agree with you Nazgul.;)
I have tested using 2-pass mode and I get slightly bigger
size with "light" setting and almost identical size for
"normal" and "strong" settings.
If less bits are applied to very dark/light scenes, so the
encoded stream gets smaller size or bits are distributed to
other scenes that require more? If the later is true, then
encoded file size must be identical :confused:
Nazgul
24th April 2002, 02:58
Originally posted by llemor
I agree with you Nazgul.;)
I have tested using 2-pass mode and I get slightly bigger
size with "light" setting and almost identical size for
"normal" and "strong" settings.
If less bits are applied to very dark/light scenes, so the
encoded stream gets smaller size or bits are distributed to
other scenes that require more? If the later is true, then
encoded file size must be identical :confused:
It's been my understanding that psychovisual enhancements were not implemented to reduce the number of bits used, but rather allocate them more efficiently by using them in areas the human eye would be more sensitive to. Doing so would not automatically use fewer bits, it would just be smarter about how it used them, increasing the quality of the image without increasing the bitrate. Now this has the side effect of being able to lower the bitrate with psychovisual enhancements and get equivalent quality to the original bitrate without psychovisuals(we REALLY need to come up with a standard abbreviation for that word). So in practical terms, yes it can allow you to achieve smaller filesizes. It'll just take some trial and error to figure out how much you can lower the bitrate and keep the quality at the level you want it.
I'm guessing it was one of those features geared towards streaming/online content, where keeping within a bandwidth limit is the primary goal. Or for those of us who like to save as much space as possible. I mean, yeah I've got 650GB of space to use, but there's a LOT of stuff I want to put on that array :)
Nazgul
Fox Mulder
24th April 2002, 06:02
Thanks for the advice, Nazgul. It seems that you have tested this p.v.m. enhancements more in depth.:)
What are the results of your tests?
Hanty
24th April 2002, 11:39
I thought we all agreed that psychovisuals are still too "beta" to be used effectively?
Mac Sidewinder
24th April 2002, 14:35
<<I thought we all agreed that psychovisuals are still too "beta" to be used effectively?>>
Wasn't this with version 5.0 not 5.01? Has psychovisuals changed in the upgrade?
Mac
dragoman
24th April 2002, 16:56
Hi,
Supposedly it is upgraded for 5.01
dragoman
vBulletin® v3.8.11, Copyright ©2000-2026, vBulletin Solutions Inc.