Hi, Thanks for test results. Yes, now I get also good length when I choose 24 instead of 30 which is reported by mplayer (which is quite confusing...) However, I have still theese (and other mentioned before) problems with some particular vobs (of unknown origin), maybe it is just them that are broken... well I just have to get the originals somehow and check with streamdump. And still I get some NTSC DVDs play/rip on one dvdrom, but not on another one... But this one is finally fixed! So I dont have to worry about region encoding, woaw, this is really great, there are quite some nice ntsc-released-only dvds out there! Thanks to all who helped DS
Hi Dominik !!!!!
I haven't the old files, but I made some test for you with a newer. In
http://mate.dm.uba.ar/~mpavon/tmp
You can find the following files:
sample.vob (20mb) test.sh 23.976.log 29.97.log
The first is the vob sample for the test. I encoded it with a target size of 2mb. I estimated the videobitrate with the formula
target_size*1024^2*8/(duration_in_sec*1000)
Duration is 26.7 sec so I got 628 approx.
For the test I used test.sh first with OFPS="-ofps 23.976" and output ofps-23.976.avi and later without it (so I believe mencoder used 29.97 as it said) with output ofps-29.97.avi.
The size of the resulting files is:
-rw-r--r-- 1 m users 2104052 Nov 8 23:19 ofps-23.976.avi -rw-r--r-- 1 m users 1663710 Nov 8 23:21 ofps-29.97.avi
As you can apreciate, 23.976 reach the target size and 29.97 is noticiablely smaller.
23.976.log and 29.97.log are a cut & paste from the Eterm after the second pass. You can see the the difference in the duplicate frames between 23.97 and 29.97, I guess this is a sign the that 29.97 fps is wrong
Regards,
MartÃn