Audio-Video delay on Finding Nemo

[Harm]

1. DVD-RB Version: 0.52 (but on previous versions as well)
2. Encoder being used: CCE SP 2.67.00.23
3. Using eclCCE? Yes
4. DVD: Finding Nemo
5. Bug encountered: Slight audio delay on DC track

I tried to report this bug on the big thread but got no reaction there. wmansir asked to open a new thread for it.

On this DVD there are 4 audio tracks. The 4th track being the Director's comment which is 2 channels (the other 3 are 6 channels).
Although it's difficult to check for a delay on the other tracks I think they are right. The 4th track is definitely off (slightly too early). You can see that e.g. with the introduction where the director is talking.
I ran this one with DTS track (track 2) deselected.

[Joergen]

I dont think they sync the DC tracks so precisely, are you sure you're just not hearing things?

Like I watched a backup I made of an old western, and could swear the sound was off when people talked, but all the SFX like gunshots, walking, falling were in prefect sync. It's cause they DUBBED all the voices when it was too difficult to record live on-set.

[Harm]

Originally posted by Joergen
I dont think they sync the DC tracks so precisely, are you sure you're just not hearing things?
No I am not hearing things..... :)
The original DVD doesn't have this delay! And the director's mouth movements are perfectly in sync with the audio there.
I also checked the files after I ran DVD decrypter (input for Rebuilder) and they are OK too!
So I guess the delay is introduced by Rebuilder......

[Harm]

Bump.......... in the hope jdobbs will notice it.... :sly:

[jdobbs]

Originally posted by Harm
No I am not hearing things..... :)
The original DVD doesn't have this delay! And the director's mouth movements are perfectly in sync with the audio there.
I also checked the files after I ran DVD decrypter (input for Rebuilder) and they are OK too!
So I guess the delay is introduced by Rebuilder...... What are you using to preprocess before running DVD-RB? Anything?

[Harm]

Originally posted by jdobbs
What are you using to preprocess before running DVD-RB? Anything?
Just DVD decrypter 3.2.2.0 in filemode and then rebuilder right away in 3 step CCE mode.

Maybe wmansir has found the time to run it because he also has the same DVD.

[wmansir]

I still haven't found time to do it yet. Hopefully I will soon.

I should also point out that Finding Nemo's DC isn't just a normal alternate audio track, but a visual commentary that contains extra footage (behind the scenes, storyboards, roughcuts, ect) inserted between scenes. DVD-RB is able to handle it because it uses duplicate VOBids. (IIRC) The DC track is actually one long PGC and the theatrical cut uses select VOBids from that track to show just the movie.

I still don't understand why this would cause any sync problems, but it's something to factor in.

[Harm]

Originally posted by wmansir
I still haven't found time to do it yet. Hopefully I will soon.

I should also point out that Finding Nemo's DC isn't just a normal alternate audio track, but a visual commentary that contains extra footage (behind the scenes, storyboards, roughcuts, ect) inserted between scenes. DVD-RB is able to handle it because it uses duplicate VOBids. (IIRC) The DC track is actually one long PGC and the theatrical cut uses select VOBids from that track to show just the movie.

I still don't understand why this would cause any sync problems, but it's something to factor in.
Already found some time to test it wmansir? :)

[wmansir]

I tried looking for my Nemo DVD but I think I let someone borrow it (at least I hope so), but I did find a DVD Shrink backup I made. So I ran that thru DVD-RB v0.54 but unfortunately (or not) I couldn't confirm this problem. I'm not very good at spotting lip sync problems, but I compared the Shrink and DVD-RB output and they seemed identical. If it was there it was minor enough to drive me crazy looking for it in lip sync and I didn't notice any non-speech sound effects were off (from the movie sound that was potted down).

Since I didn't do the original this isn't definitive of course, I'll try and get my original back this week and see if that turns out any different.

[Harm]

You looked at the DC's comment? The difference isn't hard to spot. As soon as the directors start to talk (beginning of the track) you see it right away. Especially when the third director gets his close-up (only his mouth is shown) it is very obvious.
Maybe I should run it myself again with version 0.54?

[wmansir]

There was no obvious difference between the input and output, either in that scene or later scenes, even parts of the visual commentary where crew people spoke on camera. It's possible DVD Shrink changes something that was causing the error, or something was fixed in the last few versions. I'll have to try again with the original to know for sure.

Did you see this problem on software and standalone players?

[Harm]

The difference can be seen on both software and standalone player.
You ran the 3 step CCE with rebuilder?

[wmansir]

I ran a OPV with CCE to save time, though I don't see how that could make a difference.

[Harm]

Originally posted by wmansir
I ran a OPV with CCE to save time, though I don't see how that could make a difference.
And what version of CCE did you use?
Maybe the video gets shifted some how in the CCE encoding?
I'll run the OPV too on "my version" and let you know the results....

[wmansir]

I used 2.5.

I don't see how the video would get shifted, at least not in any way that is dependent on the video mode or version.

[Harm]

I ran OPV but got a run time error '6' overflow when starting step after Q=28786....
Should OPV run up to this Q amount? The result of every Q amount is 1644548 sectors.

Am I doing something wrong? I never used OPV until now.....

In the meanwhile I while run the normal 3 step CCE again.

And BTW I deselected the DTS track.........!

[jdobbs]

Originally posted by Harm
I ran OPV but got a run time error '6' overflow when starting step after Q=28786....
Should OPV run up to this Q amount? The result of every Q amount is 1644548 sectors.

Am I doing something wrong? I never used OPV until now.....

In the meanwhile I while run the normal 3 step CCE again.

And BTW I deselected the DTS track.........! Q=28786?

[Harm]

Originally posted by jdobbs
Q=28786?
Yep! This was version 0.54. It started out with a low number (I think it was 9) and kept on running each time raising the Q number and displaying the same result sectorwise.....
Can you make sense of that?

[jdobbs]

Originally posted by Harm
Yep! This was version 0.54. It started out with a low number (I think it was 9) and kept on running each time raising the Q number and displaying the same result sectorwise.....
Can you make sense of that? Can you tell me what the target number of sectors was? As a matter of fact, can you abort as it's climbing and post the status screen?

[Harm]

Originally posted by jdobbs
Can you tell me what the target number of sectors was? As a matter of fact, can you abort as it's climbing and post the status screen?
Sure I can........ at what number Q do you want me to abort?
And are you interested in a specific part of the status screen?

As I said the Q number climbs but the result of predicted sectors is always the same (1644548).

[wmansir]

Check your D2VAVS folder and see if you have both a .mpv and a .m2v file being generated (after the SECOND Q test). If so try reinstalling the latest ECLCCE (delete your Eclcce.ini file and set it up from scratch too).

[Harm]

First: I ran the CCE 3 step with version 0.54 and the result is still out of sync (clearly visible on the DC introductory).

Second: if I look in the dir where the results of the OPV are saved there are both m2v and mpv files there. Does this mean I need to follow your advice or is this the way it should be? I use version 1.7b. Is this the correct version?

[Sir Didymus]

Last release of EclCCE is 1.81...
You may get it in the Installation, Setup, and Usage sticky thread...

Hope it is helpful...

[Harm]

Originally posted by jdobbs
Can you tell me what the target number of sectors was? As a matter of fact, can you abort as it's climbing and post the status screen?
Well I tried aborting it when Q reach more than 5 figures.....
(at that time of course the next Q was running....)
I got a popup telling me only 1 instance of ECL was allowed running and after that run time error "53". When I clicked them both rebuilder disappeared leaving me no status screen to copy..... :(

I think it started with Q=49 and raised itself continually. As I said the predicted amount of sectors is always 1644548.

Do you want me to run it again? If so how can I abort it without losing the status screen? Or can I enable some kind of logging?

[wmansir]

You need ECL 1.81 for OPV to work. Earlier versions would revert the output extension back to .mpv after the first test, so DVD-RB get's caught in a loop looking at the original .m2v file from the first test. That's why the results don't change.

[Harm]

Originally posted by wmansir
You need ECL 1.81 for OPV to work. Earlier versions would revert the output extension back to .mpv after the first test, so DVD-RB get's caught in a loop looking at the original .m2v file from the first test. That's why the results don't change.
OK. That was the reason of those errors in the OPV run.
I now ran it but.......... alas the sync problem is still there!
And I rechecked: it's not there in the input files!

So now what? Do I have to run it again with the DTS file still selected? Or can't that be the reason?

And jdobbs: I can't seem to select the lines in the status window at all --- should I be able to do that?

[Sir Didymus]

Ok. Please, take it easy, and consider the application is in beta testing...

I know many other people reported to have done a successful backup of Finding Nemo, so it is a good idea to ask them to check for the DC track...

Just few comments on my side:

First: Deselecting the DTS tracks is a must, (imho, of course, except for the music DVDs...) so keep it unchecked.

Second: For copying the status windows, just do: File --> Copy Status to Clipboard. You may then paste it into whatever editor you have.

Third: I know it takes some time, but could you please verify what happens using ReJig (and maybe also QuEnc) instead of CCE ? This should provide some hints for understanding if the "guilty" is really the authoring engine of DVD-RB...

Cheers,
SD

[Harm]

Originally posted by Sir Didymus
Ok. Please, take it easy, and consider the application is in beta testing...

I know many other people reported to have done a successful backup of Finding Nemo, so it is a good idea to ask them to check for the DC track...

Just few comments on my side:

First: Deselecting the DTS tracks is a must, (imho, of course, except for the music DVDs...) so keep it unchecked.

Second: For copying the status windows, just do: File --> Copy Status to Clipboard. You may then paste it into whatever editor you have.

Third: I know it takes some time, but could you please verify what happens using ReJig (and maybe also QuEnc) instead of CCE ? This should provide some hints for understanding if the "guilty" is really the authoring engine of DVD-RB...

Cheers,
SD
I know the application is in beta and I am taking it easy. I just registered this problem in the continuing process of getting it perfect! I am not really bothered by the DC being out of sync.

Next time I need to copy the status window I'll try your suggestion. I also thicked "echo status to log" and will see what happens next time.

I am willing to do a Rejig pass but I have never used it so are there special considerations when using it through rebuilder?

As I understood from wmansir is Finding Nemo a special kind of DVD. I think that's the main reason I put in all this effort to give the builder(s) enough information for them to find the process in error.

If someone has other suggestions I gladly will follow them..... :)

[jdobbs]

Originally posted by Harm
I know the application is in beta and I am taking it easy. I just registered this problem in the continuing process of getting it perfect! I am not really bothered by the DC being out of sync.

Next time I need to copy the status window I'll try your suggestion. I also thicked "echo status to log" and will see what happens next time.

I am willing to do a Rejig pass but I have never used it so are there special considerations when using it through rebuilder?

As I understood from wmansir is Finding Nemo a special kind of DVD. I think that's the main reason I put in all this effort to give the builder(s) enough information for them to find the process in error.

If someone has other suggestions I gladly will follow them..... :) Echo status to log does nothing. I often work on different functions and have partial code in them, keeping them hidden when publishing a version -- I forgot to hide that one...

[Harm]

Originally posted by jdobbs
Echo status to log does nothing. I often work on different functions and have partial code in them, keeping them hidden when publishing a version -- I forgot to hide that one...
OK. No problem!
Would it be a good idea to check the version of ECLCCE before running a OPV run? You see what kind of strange problems occur when running it through the incorrect ECLCCE version!
;)

[Harm]

Has any of you seen this post in the big bug thread?
It's by user the_eza

"One more audio out of synch report:

DVD: Criterion's Hidden Fortress
SW: DVDRebuilder 0.54 + CCE 2.67
-One click mode
-dynamically assign cell bitrates enabled
-Steal Space from Extras: 50%

The extras in the result have audio badly out of synch. The movie itself is OK."

Could this be the same thing as on Finding Nemo?