Log in

View Full Version : parsing IFO with a strange result


madhatter300871
17th August 2009, 15:27
Hi.

Not sure if the MPEG2 encoding section is the correct place but couldn't see anywhere else more obvious, so here goes.

I'll use '"The Matrix" as my example here, although I have experienced this problem before, although not too often.

The Matrix has a duration of 2 hours, 10 minutes and so many seconds. Everything is completely decrypted and sitting on my hard drive. It plays perfectly and media player shows a time of 2 hours 10 minutes and so many seconds.

If I run the files through DVDshrink and pull out only the main movie, same thing. I have a much smaller set of files (obviously) with a duration of 2 hours, 10 minutes and so many seconds.

However, when I parse the IFO file with either vstrip (I still love to use this) or more recently as a test with dvd decryptor (remember the files I have are already decrypted so I'm only using these apps as IFO parsers) , I select the main PGC of VTS 2 (which shows as 2 hours 10 minutes and so many seconds) but the output stream is only 1 hour 57 minutes in length.

What am I missing here ? Does anybody know what's happening ? Short of playing both movies side by side and keenly observing I don't know what to do. I'm now wondering if my backups are not complete after all and don't want to have to load everything onto my media server again.... I've got tons of stuff !

I can only presume there is 13 minutes(ish) of video from another angle maybe... but why would DVD shrink extract the whole 2 hours 10 mins while the IFO parsers will not.

Any help would be genuinely appreciated as I cannot get to the bottom of this.

Many thanks.

Ghitulescu
18th August 2009, 07:36
The authoring of Matrix was very complicated with lots of branches and decisions (you can skip scenes, you can add scenes, different flowcharts, the whole stuff). I have the PAL version at home, but I can reach it only in the weekend :(
Meanwhile I suggest to check the output of Shrink, to see if it correctly picked up the right movie.
There is also a guide on the net (was?) concerning the reauthoring of Matrix (with scenarist), wherein you'll find a better explanation of how Matrix was authored, maybe it contains your answer, too.
This is one link, but it should be another better one -> http://www.doom9.org/index.html?/mpg/ButtonOverVideo.htm

madhatter300871
18th August 2009, 13:17
Thanks ghitulescu

I read the link that you sent and indeed the Matrix is a complicated disk !!

Im pretty clued up on DVD structures but by no means an expert, but this link has made me think that probably the 2 hour 10 minutes main movie might actually skip cells during normal playback if the white rabbit isn't clicked.....

The first white rabbit happens about 3 or 4 minutes into the film so tonight Ill watch both versions (the main 2 hour 10 and the parsed 1 hour 57) side by side and see if they play exactly the same but the timer on the longer film jumps if the white rabbit isn't selected.

If this is the case (and I'm hoping it is) then I'm happy. If not, I'll read that guide again and keep playing.

Interestingly I only have PGC1 and PGC6 in VTS2 as the full main movie, all the others are 10ms in length. The guide by seyton says PGC13, 14 and 15 are the main movie also. I have the PAL version, was this authored differently ?? I only ask out of interested, I don't suppose it really matters.

Thanks for the reply, I'll post back what I find.

madhatter300871
18th August 2009, 23:35
So I played them back side by side and the DVD showed a time of 2 hours 10 and the hard disc vstripped main PGC had a time of 1 hour 57.

When playing, the DVDs displayed time did start to creep ahead of the PGC stripped diplayed time by a few seconds after about 8 minutes.
I then randomly seeked through each one and the time stamps where wildly different for any given scene.

However the strangest thing was this :-

While watching them side by side they had drifted by about 3 or 4 seconds after 8 minutes but where still on the same scene as each other (give or take a frame). So I paused each movie at the same point, seeked the DVD forward then seeked it back to the same frame and the DVD time stamp was 3 minutes ahead (so was showing 12 minutes instead of 8) !!!

So Im convinced that my vstripped PGC has nothing at all missed out, but the DVD simply displays a longer run time for some reason, and is doing something screwy with the time stamps. Could it for some reason be playing slower ?

Tomorrow I'll run each one through an avisynth script to count the frames, my guess is they will both have the same number of frames.

This is strange hey !