View Single Post
Old 20th October 2010, 20:50   #7  |  Link
yetanotherid
Banned
 
Join Date: Apr 2008
Posts: 723
Quote:
Originally Posted by Marshy View Post
And this based on what; hard fact of every high or low motion scene or just your subjective visual opinion? Doesn't 2pass allocate bitrate better in scenes whch need it or those that not. Your comment is vaque, back it up with proof and show us.
No hard facts. Just my subjective visual opinion. Is there another measurement of quality I should be using?

I've converted movies each way, stopped them on various identical frames and looked for differences... even after zooming in. I've watched them at normal speed side by side on identical monitors, and any differences I could see were only when pausing on identical frames and zooming in, but if there were differences they were so tiny I really had to look to see them, and I still couldn't say one method consistently gave better results than the other. They were just sometimes very, very slightly different. When watching the movies playing side by side if there really was a quality difference I doubt I could spot it even with bionic eyes.

I couldn't say whether 2 pass allocates bitrate better than single pass. I suspect it might sometimes allocate it more efficiently but that possibly doesn't equate to a difference in perceived quality. There must be some difference as I've encoded video to a certain quality (say AutoGK's 75%) then used the resulting file size for a 2 pass encode. I've done that quite a few times and the resulting quality of the 2 pass encode is never exactly 75%. It's only a little different, say 74% or 76% percent at the most, but that's not a quality difference anyone's ever going to be able to see and as I said maybe it's more a difference in encoding efficiency rather than perceived quality itself.
Thinking about it, it may even be partly due to the fact that a single pass encode just stops when it's done, whereas when I pick the same file size for the 2 pass encode I can only do so to the nearest MB.

As I said in an earlier post 2 pass encodes with the standalone player compatibility option enabled keeps the maximum bitrate in check whereas single pass doesn't, but even then I've compared encodes done both ways and even during high bitrate moments I can't see a difference.

I guess the fact that I can encode to a certain quality, then do a 2 pass encode using the first one's file size, and when more often than not AutoGK reports a first pass quality as being no more than around 1% different to the single pass encode, it helps me confidently advise someone there's no real quality difference between the two methods.

Hopefully that's taken some of the vagueness out of my reply for you. If you know of a way to measure perceived quality that's better than subjectively perceiving it, please let me know. I'd be keen to try it.

Last edited by yetanotherid; 20th October 2010 at 21:06.
yetanotherid is offline