On Sun, 21 Sep 2003, Richard Ellis wrote:
> The only difference between these files was different -N values. All
> other denoise/mpeg2enc parameters were identical.
What value are you using for -q? What I am finding is that
the effect of -N changes as -q is varied - especially for
the lower values of -N
> The setting of -N does work nicely to fine tune a files resultant
> size, but clearly it's also a non-linear function as you already
> suggested it may be in an earlier email.
I had hoped to have some numbers (and I do have a few but the
runs are not complete) by now. The thought was to speed up
the process by using multiple machines. However, the machines
are running different operating systems and the average bitrate
numbers for identical runs come out different - they are not
comparable if mpeg2enc was run on different OSs! In one case
(-q 7 -N 0.5) the BSD system came out about 6% lower than the
otherwise identical run on a SuSE 8.2 system. As soon as I
have the numbers (5 different -q values, and 16 -N values for
each -q) I'll tabulate them and post. I think it'll be a
useful "quick reference quide" - given a -q you should be able
to pick a -N value to suit the average bitrate reduction.
Cheers,
Steven Schultz
-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Mjpeg-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/mjpeg-users
