[Ffmpeg-cvslog] [ffmpeg]: r5417 - in /trunk: libavcodec/snow.c libavcodec/snow.h tests/ffmpeg.regression.ref tests/rotozoom.regression.ref

[Loren Merritt]

On Sun, 28 May 2006, Corey Hickey wrote:
>
> Wow, two big snow patches in one day. It must be Christmas already. :)
>
> Loren mentioned that this is very content dependent, but here are the
> PSNR and user times for the second pass of my usual test clip. There
> wasn't any significant variation of video stream size, so I'm omitting
> that line to keep it clean.
>
> refs=1
> PSNR: Y:41.66, Cb:44.04, Cr:44.63, All:42.38
> user    1m5.722s
[...]
> refs=8
> PSNR: Y:41.81, Cb:44.08, Cr:44.66, All:42.51
> user    1m42.654s
>
> Looks like the encoding time increases linearly and the the PSNR
> increases logarithmically. The speed reduction isn't all that bad,
> except when using vme=8, which compounds the speed reduction:
>
> refs=1:vme=8
> PSNR: Y:42.35, Cb:44.55, Cr:45.11, All:43.03
> user    7m2.596s
>
> refs=8:vme=8
> PSNR: Y:42.46, Cb:44.56, Cr:45.11, All:43.11
> user    23m20.679s

I get just as much slowdown for me=epzs as for me=iter ...

kb/s:1123.75 psnr:42.87 time: 50.41 refs=1
kb/s:1107.34 psnr:42.98 time: 65.48 refs=2
kb/s:1101.14 psnr:43.02 time: 80.08 refs=3
kb/s:1104.14 psnr:43.05 time: 94.51 refs=4
kb/s:1106.04 psnr:43.06 time:108.52 refs=5
kb/s:1106.71 psnr:43.07 time:122.19 refs=6
kb/s:1106.61 psnr:43.08 time:135.48 refs=7
kb/s:1106.52 psnr:43.08 time:148.25 refs=8
kb/s:1025.59 psnr:43.10 time:272.62 refs=1:vme=8
kb/s:1013.71 psnr:43.20 time:402.03 refs=2:vme=8
kb/s:1008.98 psnr:43.24 time:507.26 refs=3:vme=8
kb/s:1010.21 psnr:43.27 time:594.24 refs=4:vme=8
kb/s:1010.41 psnr:43.28 time:667.33 refs=5:vme=8
kb/s:1009.93 psnr:43.28 time:731.47 refs=6:vme=8
kb/s:1009.82 psnr:43.29 time:787.55 refs=7:vme=8
kb/s:1011.12 psnr:43.30 time:836.36 refs=8:vme=8

--Loren Merritt