On Wed, Sep 22, 2004 at 04:04:02PM -0600, James Bigler wrote:
> This was suspicious as the original source images are fine. I even did:
>
> convert nocrack.00{6,6,6,6,6,6,6,6,6,6}0.ppm nocrack.0* crack.0*
> -thumbnail '640x512>' -bordercolor black -border 50 -gravity center
> -crop 640x512+0+0 ppm:- | ppmtoy4m -F 25:1 -r -n 150 > frames.raw
>
> cat frames.raw | y4mtoppm | pnmsplit - "t1/%d.ppm"
>
> ls -1 t1/*.ppm | sort -n | xargs xv
>
> And everything looked fine.
>
> I did this to encode:
>
> cat frames.raw | mpeg2enc -f 0 -o o1.m1v
>
> And then to inspect the frames:
>
> mpeg2dec -o pgm o1.m1v
> ls -1 *.pgm | sort -n | xargs xv
>
> And there are a number of frames in the wrong place.
>
> The only reason I can think that the data is wrong, but the time step is
> correct is splicing more than one frame. I've seen this on a couple of
> frames where a frame is divided horizontally with one frame on top and
> another on the bottom. For the particular frame in question, the split
> is not visible, but the top part of the correct frame was on top and the
> bottom was for a future frame.
>
> Any way to compensate for out of order frames? I believe this is where
> the "popping" or "pulsing" in my movies is from.
Just for the sake of a comparisson, have you tried generating mpeg1 stream
using the latest CVS ffmpeg ?
I'd recommend something like this:
$ ./ffmpeg -i %d.ppm -q 2 -r <desired frame rate> test.mpeg
Thanks,
Roman.
-------------------------------------------------------
This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170
Project Admins to receive an Apple iPod Mini FREE for your judgement on
who ports your project to Linux PPC the best. Sponsored by IBM.
Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php
_______________________________________________
Mjpeg-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/mjpeg-users
