On Sat, 22 Nov 2003 [EMAIL PROTECTED] wrote:
> I came across the lofty claims made on www.kvcd.org using their magic
> matrices. Putting 180 mins of VCD material on a single CD-R is impressive
That is a bit too lofty (optimistic) in most cases - the actual
amount will depend on how noisy the source data is.
> (tmpgenc, etc). I am a Linux user. I was glad to see that the latest
> version of mjpeg 1.6.1-90 has added this -K kvcd feature to mpeg2enc.
Works quite well but there is another option you'll need to use...
> I am encoding to XVCD (using a manual bitrate of 2000kbps) and the filesize
> of my resulting MPG1 is not at all different than if I did NOT use the
> -K kvcd.
> What is the recommended way to use this feature? By specifying a manual
> bitrate higher than the VCD standard, am I negating this custom matrix?
VCD MPEG-1 is CBR (Constant Bit Rate) by default - thus if you
specify "-b 2000" then the resulting file will be *padded* to
achieve the specified bitrate.
Requesting a nonstandard bitrate does not negate the use of the
custom matrices.
When encoding for XVCD MPEG-1 you need to request the VBR (Variable
Bit Rate) when multiplexing - use the "-V" (--vbr) with "mplex":
mplex -V -f 2 ...
> What is a good -q factor to use? I think dvd::rip calls mpeg2enc with the
> default value for this, which is 8?
-q 8 (the default is reasonable) for VCD use but -q 7 or perhaps
-q 6 are also good choices depending on how clean the source is.
If the original is noisy then a slightly higher -q (7 or 8) is
better.
Add the --vbr to the mplex run and you should be all set.
Cheers,
Steven Schultz
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Mjpeg-users mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/mjpeg-users
