[MEncoder-users] High quality encoding of anime and broken mencoder options

Nicolas George nicolas.george at normalesup.org
Wed Mar 5 11:56:54 CET 2008


Le sextidi 16 ventôse, an CCXVI, Jorge Peixoto de Morais Neto a écrit :
> Hi. I seek high quality encoding of anime. I have tested
> many encoding options, always with vcodec=mpeg4 and 2-pass encoding. 
> I have found that even with vbitrate, the actual bitrate depends on
> the options used. Therefore, to keep the comparison fair, I have
> measured encoding quality by Q=PSNR/bitrate, with bitrate in Mbits/s.

I believe that your Q value is not relevant: why PSNR/bitrate rather than
PSNR²/bitrate or PSNR/bitrate², or PSNR/log(bitrate) or... PSNR itself is a
logarithm of mean square error, there is no reason that a simple division
should be the relevant measure.

There is a case where the results are certain; if bitrate(A) <= bitrate(B)
while quality(A) >= quality(B), then it is certain that the settings for A
are better (quality-wise) that the settings for B.

Fortunately, in your tests, you are precisely in this case.

If this is not true, then you have to vary your target bitrate until you can
compare, or plot curves and see which one is above the other.

> Excellent quality:
> secs  1201 frames PSNR: Y:35.48, Cb:38.85, Cr:38.26, All:36.28

I would not call that "excellent quality": I like my videos well above
PSNR 40.

> 1) Do you think the above settings are wise? Any suggestions?

Do you insist on using simple MPEG-4? H.264 gives much better results.

Shameless plug: You may be interesting by a look at <URL:
http://www.normalesup.org/~george/comp/video_quality/
 >. I have a second part in this experiment waiting to be published, where I
explore the effect on quality and time of the various encoding options, but
it still requires some work.

Regards,

-- 
  Nicolas George



More information about the MEncoder-users mailing list