2 quick questions - HeadAC3he downmixing and delay times

[brett]

Question 1 - surround downmixing an AC3 surround source:

This should apply to using GraphEdit, BeSweet, HeadAC3he, or however you decode your audio. You always have the option to do a Dolby Surround downmix (or DS2 in HeadAC3he). This makes perfect sense when you're dealing with a 5.1 source, but what about source that's already in 2.0 surround? You can't exactly "downmix" a 2.0 source to 2.0 surround, can you? What is the correct thing to do? Will selecting Surround downmixing preserve the surround encoding correctly, or will it screw it up? Should I just be selecting Stereo? I looked through all the limited HeadAC3he documentation and all of Doom9's guides, and did a search for "AC3 2.0" on the forums without coming up with anything, so try not to yell at me too much for asking such a dumb question.



Question 2 - what gives with the delay times in DVD2AVI and SmartRipper?

First off, when you demux the same audio track, DVD2AVI and SmartRipper will give you different delay times. They're only different by 1ms, but it'd still be nice to know which one is right.

Anyway, I'm confused about what HeadAC3he does with delay times. Doom9's guide says, "HeadAC3he will automatically adjust the delay if you demux in DVD2AVI," but later it says that when multiplexing the MP3 with the AVI that you should "enter the delay value from your AC3 file in the Delay audio track by XY ms field" in NanDub. Is it correct that HeadAC3he has already adjusted the delay in the output, so I can just enter 0 here?



Thanks for any help.

[DSPguru]

1. downmix isn't relevant for 2.0 source.
2. delay can be either asserted on the encoding process or at the muxing process. no need to compensate for the a/v delay twice.
dvd2avi known to have better a/v delay accuracy.

[kdiddy]

No I htink he means...with Head, do you adjust the time to zero..thus that cures the delay...or do you leave the delay time at what it says, and headache does it automatically??

[valnar]

Sorry to bring this thread back, but I had the same question.

If I demux with DVD2AVI and resulting filename has a delay shown in the filename description, has that delay already been applied so no further post-delaying is necessary?

** OR **

re: HeadAC3he. Does it take the filename with the DELAY statement and apply that correction automatically? What if the filename was changed so it no longer had that DELAY in there? DoitFast4U changes the name, so would I still need to manually account for the delay, even with HeadAC3he?

Thanks,
Robert

[Eyes`Only]

Due to this popular request, starting with 1.0.9.1 (still beta) I've added the delay time to the filenames when you demux with DoItFast4U!

[moon1234]

Question 1:
Applying a DPL II downmix on an 2 channel AC3 will help preserve the DPL2 encoding of the file. If it wasn't encoded DPL2 then you gain nothing. If you are going directly to another format (mp2, mp3, etc.)then just use BeSweet. There is no separate output to WAV. Just make sure that you encode using a PSY model of 2 if going to MP2. This will maximize the DPL2 encoding. I have tested this myself on some sample conversions and I can tell the difference. The rest of my tone deaf family can't.

Question 2:
The delay is not applied to the de-multiplexd file.

This was the delay that was set in the program stream. If you want to multiplex this audio stream once you convert the file to DIVX (Or whatever), then you either need to apply an audio preload skew when you multiplex OR add or delete samples to the audio file itself.

Using an audio preload skew is the suggested solution as it doesn;t modify the file. Not all players read this tag correctly or apply it correctly. If your player is one of these then you need to add or delete samples.

The skew that you see in the file name is informational only. Nothing has been done to the file.

-Moon1234