Auto correct ac3/dts delay feature - How does it work ?

[rca29]

I've made several backups over the last year, and as i always like to get the best possible image, i usually stay away from 'One click DVD backup solutions'.
But i always had an issue with DTS negative delay...
...and my question is: How does this work on DoItFast4U ?
How is the DTS delay corrected ?

Thanks and sorry if this has been posted before

[Eyes`Only]

if the delay is negative, it removes frames from the .dts to correct the delay, if the delay is positive, it pads the .dts with blank audio. The issue is that Scenarist (and many other authoring apps) ignore the delay in the header of the ac3 files when it imports them. Therefore, changing the bytes that specify the delay in the header won't do a thing. You could change the delay to -123456 and it wouldn't matter a bit, since it would just be ignored. So we found a solution, pad or cut the stream to match the delay.

[Matthew]

Of course, in the case of AC3 (not sure about DTS), 1 frame = 32 ms so no delay correction is exact unless it is a multiple of 32. In Scenarist positive delays can be created, but only in increments of 1 frame (40 ms in the case of PAL). So a reauthored DVD may be out by a max of 16 ms compared to the original.

This is one area where Maestro is superior to Scenarist - positive audio delays can be created in 2 ms intervals (I've found anyway) so a/v desync will only be out a max of 1 ms compared to the original DVD.

Of course, as the delay is not noticeable at 16 ms anyway, this greater accuracy isn't exactly brimming with practical advantage. But it makes some people feel better :p

[Trahald]

eyes - i have a ? does ac3 corrector add padding to the end of the file when it removes frames.. or delete frames from the end when it adds to the front?

[Eyes`Only]

no

[D3s7]

Originally posted by Eyes`Only
The issue is that Scenarist (and many other authoring apps) ignore the delay in the header of the ac3 files when it imports them.

I thought the delay was in the VOB not the ac3 header... if I could extract delay from the ac3 header I'd be a happy camper