SubRip 0.98b bug, not process VobSub subtitles correctly

[kxy]

To be more exact, the timing is WAY off.

I use subrip to process the vobsub *.sub file, this is the result

1
00:00:27,395 --> 00:00:31,695
DJ: What's up? This is K107,
broadcasting on New Year's Eve.

2
00:00:31,733 --> 00:00:35,499
Why don't y'all scribble down
those resolutions and call me?

<skip>

2183
00:03:57,800 --> 00:03:59,358
Yeah

Just to make sure, I mux the subtitle and notice the timing of the subtitle is off. Notice the hr is doesn't exist(bug number 2). The correct timing should be

1
00:00:27,193 --> 00:00:31,493
DJ: What's up? This is K107,
broadcasting on New Year's Eve.

2
00:00:31,531 --> 00:00:35,297
Why don't y'all scribble down
those resolutions and call me?

<skip>

2183
02:03:21,519 --> 02:03:23,077
Yeah

[gabest]

The timing resets after vob id changes.

MPEG is a streamable format, so sequential playback isn't a problem, but converting it into a random accessable, indexed one can be problematic sometimes...

[vxvas]

have you tried ticking the small box just
above start ???
It sais somthing like "if the timing gets
zero at the middle of the dvd ..."

[kxy]

@gabest, I understand that would be a problem if I rip the subtitle into a seperate vob, then of course the id wouldn't match up. I can produce what you mentioned by using smartripper then use either subripper 0.44 or even the new version of SubRip, 0.98b. But the fact is I am loading the *.sub file directly using SubRip_0.98b like the way I would load it up using subresync. The timing works on Subresync but with subrip it won't report it correctly.

Why don't I use subresync then? Well, the reason I am trying other software is that the subresync currently doesn't store character matrix info, and one of the perks of subrip is that it stores the character matrix and I can always go back and fix the OCR mistakes(less work), instead of using a text editor(more work).
MPEG is a streamable format, so sequential playback isn't a problem, but converting it into a random accessable, indexed one can be problematic sometimes...
Ripping the sub by indexing it never worked once, I tried loading different subs. So it is not a problematic sometimes, it is always the problem. But if you were talking about something else, then never mind :)


@vxvas, forgive me if I am being harsh. But this is a bug report not a help request. Which means that I spend a lot of time testing this bug on different subtitles with various options. I understand that you are trying to help. It would be okay if it is a help thread, but let me repeat, this is simply a bug report. So if you are interest in helping the development, you can try out a subtitle on your own and see if you can reproduce the same bug OR tick the mark like what you suggested and see if works. Then you can come back to me and tell me that I report a false bug if you were right. ;) So, to answer you question, yes I have tick that, even tho I know that has nothing to do with this bug, and of course it didn't work.

[zuggy]

Thanks for bugreport, kxy.

Bug2: I've tried it and it seems, that timecode resets after VOB id changes (vide subresync.exe) in VobSub file. Have you the original VOB file to test, if the same happens, when you rip subtitles from VOB?

The only thing i can do for it yet is, that i'll use start timecodes from idx files instead of VobSub "VOB". Which will prevent the desynch too (Bug1).

[kxy]

Originally posted by zuggy
Thanks for bugreport, kxy.
Bug2: I've tried it and it seems, that timecode resets after VOB id changes (vide subresync.exe) in VobSub file. Have you the original VOB file to test, if the same happens, when you rip subtitles from VOB?

If I rip the entire movie to vob, then the time stamp will be correct. But if I just rip the subtitle into seperate vob without the movie parts or seperate sub files via smartripper, or load the *.sub(ripped either from vobsub or smartriper) into the subrip, then the timing stamp is way off.


The only thing i can do for it yet is, that i'll use start timecodes from idx files instead of VobSub "VOB". Which will prevent the desynch too (Bug1). [/B]
I don't know what Gabest did with his subresync. You can download it and try it, not sure if the source is included. When I use his subresync program, all the timing is correct, both bug 1 and bug 2 are gone. Maybe you can look at his source?

BTW, good job with your edit Character matrix, that is the one reason that I keep coming back to subrip.

[zuggy]

gabest stores the start times to *.idx file - look into it.

I'll make some kind of workaround, that will work for VobSub titles, and that is, as i wrote before, that subrip will use these correct start times, if *.idx file will be present.

In current time, I don't know, if ripped stream loses (resp. not contain) some info about first timecode related to A/V/subtitle muxing, but I'm sure gabest knows :-)

And BTW I'm not author of SubRip, that is Brain. I just made 0.97-to-0.98 enhancements.