[FFmpeg-user] Quantify Video Quality

[Tisch Dániel]

2012/2/28 Atish Kathpal <atish.kathpal at gmail.com>

> @ Andrey  yes, thats right. I am interested in finding if there is any
> difference in quality of videos derived from an uncompressed (YUV) master
> as against an AVI master. And I would like to put some numbers against that
> quality, rather than having to manually look at the videos to know how they
> differ in terms of quality.
>
> Is there a tool to help me do that? Can FFmpeg itself report the
> signal:noise ratios you talk about?
>
> Thanks
> Atish
>

I guess FFmpeg can't do this directly, but MSU Video Quality Measurement
Tool from Moscow State University (
http://compression.ru/video/quality_measure/video_measurement_tool_en.html)
is the tool you need. It's not open source, but have a free version which
works perfectly for such a job. Maybe you will need some handwork with
FFmpeg (decode or change container, that does not affect quality) to get
files you can feed into MSU tool. I suggest you go for some kind of SSIM
insted of PSNR.

Cheers,
Daniel



> On Tue, Feb 28, 2012 at 3:11 PM, Andrey Utkin <
> andrey.krieger.utkin at gmail.com> wrote:
>
> > 2012/2/28 Jim Worrall <coniophora at gmail.com>:
> > > It is hard to imagine that the quality of the final transcoded videos
> > > could be improved by going from an uncompressed file through some
> > > (unspecified) codecs in an AVI first.  I would assume that the final
> > > file made directly from the YUV is better.
> > >
> > > To compare the quality, I would just look at a few samples.
> >
> > As i got the idea, the guy asks how can he calculate Signal to Noise
> > Ratio in derived files.
> >
> > --
> > Andrey Utkin
> > _______________________________________________
> > ffmpeg-user mailing list
> > ffmpeg-user at ffmpeg.org
> > http://ffmpeg.org/mailman/listinfo/ffmpeg-user
> >
> _______________________________________________
> ffmpeg-user mailing list
> ffmpeg-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user
>