[x264-devel] CALL FOR MPEG4-AVC/H.264 CODECS COMPARISON

[Dmitriy Vatolin]

Dear Jason,

Monday, February 16, 2009, 12:37:31 PM, you wrote:

>> * Objective quality measurements (PSNR, SSIM, Average Advantage, etc.)

JGG> Since x264's recent psy optimizations have proven quite
JGG> definitively that one can achieve greatly improved visual quality
JGG> while decreasing both PSNR and SSIM, will their be a subjective
JGG> comparison as well as an objective one?

     We think about this, but for 6-9 codecs this is to complex even
     on 1-2 sequences (if make all necessary things accurate).

JGG> In addition, what about the fact that encoders have different
JGG> settings optimized for PSNR, SSIM, and visual quality,
JGG> respectively? x264 can optimize for PSNR (AQ off, psy-RD off),
JGG> SSIM (AQ on, psy-RD off), or visual quality (AQ on, psy-RD on).

     We will use SSIM and recommend optimization for this measure for
     comparison.

JGG> It would be surely be unfair to ask for an encoder to try to
JGG> optimize for all three with one set of settings--one would either
JGG> end up with settings that are suboptimal for all measurements or
JGG> settings that are extremely good in one but bad in the others.

     Sure! Let's use SSIM, such approach allow to test many sequences
     for several usecases as well.

Yours,
 Dr. Vatolin