Hi -
After the mail items about -Q yesterday I did a couple SVCD
encoding runs and the results are not what one would expect.
I'll leave out the yuvdenoise, etc stages - the basic mpeg2enc line used
(for the -Q 0 case) was:
mpeg2enc -M 2 -f 4 -q 6 -K kvcd -4 2 -2 1
Then -Q 2.0 was added:
mpeg2enc -M 2 -Q 2.0 -f 4 -q 6 -K kvcd -4 2 -2 1
and finally -X was used (with a value of 500, the default is 100):
mpeg2enc -M 2 -X 500 -Q 2.0 -f 4 -q 6 -K kvcd -4 2 -2 1
The resulting files are, in order:
-rw-rw-r-- 1 sms wse 117495516 Sep 17 18:17 bugs8-Q0.m2v
-rw-rw-r-- 1 sms wse 117998355 Sep 17 16:29 bugs8-Q2.m2v
-rw-rw-r-- 1 sms wse 119481253 Sep 17 22:42 bugs8-X500Q2.m2v
Fine tuning is right ;) ~1% difference overall. I guess the next
thing to try is with the default or tmpegenc quantization matrices.
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
