AC3 rendering workflow

baysidebas wrote on 6/8/2008, 5:54 PM
Once upon a time, in the olden days of DVD Architect, One rendered in Vegas twice, once for video, and once for audio. DVD Architect put it together and authored DVDs.

Is it just that it's the way it's always been done, or is there a technical reason for continuing to workflow it this way? It seems to me that the more straightforward workflow today is to render the project to mpeg with layer 2 audio (one render instead of two, and I'm not concerned with speed here, just pulling the trigger once and having Vegas do its thing with no further attention on my part) and then let DVDA re-encode the audio as part of the DVD preparation process. Again, no additional attention on my part. And I'm talking about plain ole stereo, not 5.1 or such, which is another ball game.

The only reason I would see to do it the olden way is if somehow the audio encoding results in better audio than the alternate method.

I would welcome thoughts and insights on this matter.

Comments

bStro wrote on 6/8/2008, 7:46 PM
If you want your audio to go through two encodes (to MPEG layer 2 in Vegas and then to AC3 in DVDA), go ahead. You could have done that "once upon a time," and you could do that today. Nothing's changed in that respect.

The main "technical reason" is the one you've already cited -- better audio with only the one encode.

Rob
baysidebas wrote on 6/8/2008, 8:18 PM
Thanks, that's what I was looking for. But do you really think that the two encodes result in significant loss of audio quality? Will the single encode result in noticeably better audio? Even if the audio is just speech? Sometimes in the quest for perfection we can lose sight of practicality.
MPM wrote on 6/8/2008, 8:38 PM
"I would welcome thoughts and insights on this matter."

Don't know about any insights, but my own 2 cents if it's worth even that. ;-)

Well... FWIW you’d probably do well to avoid the layer2 audio, since you will lose a little bit, but of course that’s up to you. Other than that DVDs do use separate, or non-muxed audio tracks. Just the nature of the beast where however many audio tracks are stuck in the VOBs along with the m2v & sub-pics. Feed it muxed video & DVDA will spit out the audio, (re?) encode it to AC3, and shove that in the VOB just as happily as if you imported the audio separately.

Potential advantages of separate AC3 encoding: opportunity to test, re-encode, &/or edit as necessary, using Vegas or whatever - possibly a slight speedup since DVDA doesn’t have to demux (less work & less disc thrashing) - & if Vegas/DVDA aren’t the only tools in your toolbox, m2v & AC3 are the standards (DVDA was the oddball there, & still prefers non-std mpg2). Last one’s admittedly a stretch, but I use it so often I stuck it in as an afterthought: I check the DVD size before prep in Windows Explorer - I can substitute AC3 tracks all day long, but it’s harder to guesstimate when there’s un-used audio in the mpg file.

As far as being more of a hassle to encode separately... I *like* habits. They're a failsafe of sorts when you get tired & bored (i.e. sloppy). Many times I could get away with encoding a muxed file, but for the times I can't, when I need to do it separately, the odds of my screwing up increase.
TheHappyFriar wrote on 6/9/2008, 4:31 AM
two encodes adds encoding time. If it takes 5 minutes to render the audio, you just doubled to 10.

if you want DVDA to do it's own encode I'd say make a mpeg template that uses uncompressed WAV audio. won't loose anything there (unless the audio is a higher quality then what you specify in the template, IE project is 96khz & the template is 48khz).
bStro wrote on 6/9/2008, 10:41 AM
if you want DVDA to do it's own encode I'd say make a mpeg template that uses uncompressed WAV audio.

Vegas won't create such a template for MPEG2. You get the option of MPEG audio or nothing at all.

Rob
bStro wrote on 6/9/2008, 10:50 AM
But do you really think that the two encodes result in significant loss of audio quality?

All depends on the listener and his equipment. Some people can tell the difference between a CD and a high-bitrate MP3; some cannot. Some can tell the difference between MPEG, AC3, and PCM audio; some cannot. Some can tell the difference between a once-encoded audio stream and a twice-encoded one; some cannot. I'm one of those that cannot. But I choose to give the viewer the best possible quality in case he can.

It's a matter of opinion / perception, as is the choice between workflows.

Personally, I don't see anything "impractical" about the two-file approach. You're going to spend just as much, if not more, time encoding if you give DVDA a single file -- you're just spending it there instead of Vegas.

At any rate, do what you want. You asked why we do what we do, and I told you. Feel free to choose the workflow that works for you.

Rob
TheHappyFriar wrote on 6/9/2008, 11:11 AM
Vegas won't create such a template for MPEG2. You get the option of MPEG audio or nothing at all.

ohhh.... you're right. :) It must of been a 3rd party mpeg encoder I used once then.
Zulqar-Cheema wrote on 6/9/2008, 11:31 AM
Just in time this question, I was just looking at making a template fro DVD_A that has PCM sound in as standard so there is no second render, I need to do this pass on 3rd parties sometimes.

Some times I get files from 3rd parties and I have to re-render the
audio, wanted to save on time in the field.

While I am here the client uses FCP or Adobe Premiere does anyone of a common setting so I get the file ready to burn with no re-rendering?

Can Vegas do a MPEG and a PCM file or are they as I suspect two different files...

Thanks
TheHappyFriar wrote on 6/9/2008, 7:08 PM
you can just use a batch render to do as many types of files you want.