[MPlayer-users] one-pass encoding results better than two-pass

Matyas Sustik mplayer.list at sustik.com
Wed Jan 7 06:04:36 CET 2009


Jason Garrett-Glaser wrote:
> By default, x264 uses psy optimizations which lower SSIM and PSNR but

I am using dev-SVN-r26940 which does not support psy yet.

> improve visual quality.  Your first pass settings probably use turbo,

I do not use turbo.

> But more importantly, your first pass is 200kbps higher than the
> second--of course it can be higher quality.  You can't compare quality
> between two encodes at different bitrates.

I am not following you there.  The one pass encoding had the smaller file
size and higher ssim/psnr.  That I found unexpected.  Incidentally, I do not
know how the reported kb/s:3259.63 is calculated from the filesize.  I trust
the file system reported size though.

The bitrate was set to 3000 on both encodes.

Since the sample was short (20sec) I thought maybe this was just a fluke.

So I rerun on a 60 seconds sample.  This time I tried a 1 pass a two pass and
a three pass encoding with the same source, bitrate, yada-yada:

Single pass encoding:
File size: 21902932
x264 [info]: SSIM Mean Y:0.9822603
x264 [info]: PSNR Mean Y:44.466 U:47.888 V:49.304 Avg:45.431 Global:44.716
kb/s:2734.11

Two passes:
File size: 23831479
x264 [info]: SSIM Mean Y:0.9838375
x264 [info]: PSNR Mean Y:44.659 U:48.036 V:49.531 Avg:45.624 Global:45.243
kb/s:2990.10

3 passes:
File size: 23903141
x264 [info]: SSIM Mean Y:0.9837665
x264 [info]: PSNR Mean Y:44.587 U:47.986 V:49.520 Avg:45.557 Global:45.237
kb/s:2999.61

The first two run results I can accept without much suspicion.  However the
benefit of the second pass is non-existent for me: the increase of less than
0.002 SSIM costs almost 10% in file size. But the third seems again to be
suspicious featuring a larger file with lower SSIM.

So the question: should I bother with multipass encodings?
Matyas
-
Every hardware eventually breaks.  Every software eventually works.



More information about the MPlayer-users mailing list