This isn't quite comskip related, but I don't know where else I can ask this.
I have acquired several DVD-Rs of shows recorded on a TiVo and burned to DVD. I found a program called VOB2MPG that converted the files back to MPG, and started trying to tune comskip to them.
When I feed the mpg file to comskip, it is unable to detect sound. In the .csv file, the volume is -1 for all frames. I get sound when I play them back normally, so I know it is there in some form. Can someone out there recommend a way to get these DVDs back into an MPG, in a way where the sound is readable by comskip? I do have the original .VOB files still available.
I have comskip working to some extent, catching maybe 3 minutes out of 9 of commercials. There are no black frames in between show and commercial, and there appears to be no closed captioning. Without black frames, closed captioning, or volume, I'm unable to produce acceptable results. I can't fix the first two, but hopefully I can get volume working with comskip.
Thanks a lot for a great program. I discovered it last week and have it mostly working on the shows I record myself, and am still tweaking it a bit.
help with TiVo recordings
I found 79.39 today, and it seems to do the trick as far as detecting volume. Will you still want a sample?
I'll probably be posting again shortly with more general tuning help. I'm at least getting the raw data in now, and can exhaust a few more options first. I'm having trouble generating cutpoints though, the channel uses a lot of whitescreen instead of blackscreen breaks, and doesn't seem to always break well enough to detect a potential cutpoint using scene change *or* silence. I'll have a .csv and a hand-edited .txt before I post again though.
Thanks!
I'll probably be posting again shortly with more general tuning help. I'm at least getting the raw data in now, and can exhaust a few more options first. I'm having trouble generating cutpoints though, the channel uses a lot of whitescreen instead of blackscreen breaks, and doesn't seem to always break well enough to detect a potential cutpoint using scene change *or* silence. I'll have a .csv and a hand-edited .txt before I post again though.
Thanks!
If you find a recording impossible to detect I will be glad to help you with the tuning if you follow the procedure described here
http://mk.kaashoek.com/comskip/viewtopic.php?t=34
http://mk.kaashoek.com/comskip/viewtopic.php?t=34