On Thu, 28 Oct 2004, Dik Takken wrote:
> > By the time the encoder detects that the rate is too high and adjusts
> > the effective -q the rate spike has already been passed to the output...
>
> Ah, that's an important detail. It might be a good thing if the man page
> would be a bit more clear about this.
Actually it is mentioned in the toplevel (mjpegtools) manpage:
"A quality factor should be chosen that way that the mplex output of
Peak bit-rate and average bit-rate differ by about 20-25%\&.
If the difference is very small (less than < 10%) it is likely
that you will begin to see artifacts in high motion scenes. The most common
cause of the average rate being too close (or equal) to the maximum rate
is wrong value for the maximal bitrate or a quality factor that is too
high."
The 20-25% is too conservative and should probably be changed to
15-20%.
> Perhaps it's a good idea to have mpeg2enc print a warning when the actual
> bitrate exceeds 9700 kbit saying that some hardware DVD players may not be
> able to play the stream perfectly.
Ideally what should be done is get rid of the "-q" concept completely
and use a "minimum", "desired average" and "thou shall not exceed"
set of bitrates. "desired average" and "absolute maximum"
might be enough. The encoder would internally adjust its parameters
to meet the specified rates.
Another caution is that some DVD players have difficulty placing high
bitrate recordable (DVD+R/-R/+RW/-RW) media. I've seen recommendations
in Apple's documentation about not using higher than 8000Kb/s in order
to assure portability of recorded discs.
> > mplex -f 8 -o /dev/null input.m2v
>
> Another *very* useful hint. Thanks! :)
You may need to give "-r" and "-b" options if you're overriding
portions of the defaults set by "-f 8".
You're welcome - glad you found that little tip useful.
> Yes, I did. Shows nothing special. The actual q value never exceeds 1.25.
That's still 25% over what you gave. I'd try -q 3 and see what happens,
just for curiosity's sake ;)
> Smoothing the frames a bit is a good idea indeed. I like soft images
> better than compression artifacts. :)
yuvmedianfilter will do that. yuvdenoise can also be tried. The
cvs version has 'y4mspatialfilter' which performs bandwidth limiting
function that can, with aggressive settings, soften/blur the image.
Numerous ways to process the data - have fun! :)
Cheers,
Steven Schultz
-------------------------------------------------------
This Newsletter Sponsored by: Macrovision
For reliable Linux application installations, use the industry's leading
setup authoring tool, InstallShield X. Learn more and evaluate
today. http://clk.atdmt.com/MSI/go/ins0030000001msi/direct/01/
_______________________________________________
Mjpeg-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/mjpeg-users
