Pro Tools, Vegas 5 - Differing Tempos/Track Lengths - A BIG Headache!

Rikki_B wrote on 12/9/2004, 12:21 PM
I'm having a problem with a project that has been recorded on both Pro Tools and Vegas. I've discussed this problem with a couple of engineers (including a former Chair of Berklee's Engineering and Production Dept.) and we just can't figure this out. If anyone here has any thoughts on this I would greatly appreciate it.

Here's the problem:
We recorded basic tracks on a mac w/ Pro Tools.

The tracks were copied to a mac formatted external drive, which I took home and copied the files from to my PC, using a program called Trans-mac to convert the files to Wav.

The wave files were then copied to CD as PC data files, and loaded into various band members' PCs using a few different versions of Vegas (one guy used Vegas 1 & 2 for some of the overdubs, until I made him get Vegas 4, and I used Vegas 3 for awhile, now I'm using Vegas 5) to record overdubs. Soundcards used were a Delta 1010, an M-Box and I'm not sure what the third one was. No external clock was used for OD's, unfortunately.

Then we went back to the first studio and loaded our Vegas-recorded overdubs into Pro Tools. However, on playback in Pro Tools the overdubs start out in sync, but they become progressively more out of sync as the playback continues, and yet there is no detectable pitch shift. By the end of the song the overbubs are waaaayyyyy off (like many measures off - not even slightly subtle).

I can take all of the very same project files and open them up on my PC in Vegas 5 and they are perfectly in Sync for the duration of the song.

All recording was done at 44.1kHz, some at 16 bit, most at 24 bit.

We've gone home, carefully checked for any possible problems and found none (everything lines up and locks in perfectly when opened in Vegas) and then recopied the files and brought them back to try to get them to work correctly in Pro Tools on 3 seperate occasions now,and sure enough, the same thing happens every time. The stuff overdubbbed on the PC has a different tempo than the stuff recorded in Pro Tools when all are played back in Pro Tools, and yet there is no difference in pitch (aaaarrrgggghhhh!!!!!).

And just to keep things interesting, this is only happening on about half of the songs, and despite many hours of troubleshooting I can't really find anything particularly in common with how either half of the songs' overbubs or basics were recorded.

I have a workaround that'll get me through the project, so that's not a problem. I'm presuming there's user error here somehow, but damned if I can figure out what it is.

I'm just wondering if anybody out there might have a clue as to what's going on, because I sure don't. This seems like an impossibility to me.

This is making my head spin!!! I need a coupl'a aspirin....


Comments

pwppch wrote on 12/9/2004, 1:04 PM
The only thing i can think of - and this is a stretch - that the clock of the recording device used was not exactly 44.1 kHz. Or possibly PTools is resampling when importing the files.

What audio hardware was used with Vegas to do the recording? What was the clock source?

Peter
ScoriaMM wrote on 12/9/2004, 1:33 PM
I know it would be a huge pain, but maybe you could try rendering out each track to a new assembled one on the working Vegas system, and then take those new tracks to the ProTools session (taking care to note that you render each track to the same sample and bit depth rate).

Since it plays fine in Vegas, it might be worth the attempt.

Also, just from a troubleshooting standpoint, have you noticed if its all the Vegas tracks that drift? To put it another way, do the tracks that do drift have anything in common with each other (such as "all the tracks that drift are 16bit" or "all the tracks that drift were recorded on the Vegas Pro version 1 session") It might help narrow things down for you.

- Oops, missed the part about you not seeing anything in common, sorry about that.

-MK
Rikki_B wrote on 12/9/2004, 1:42 PM
An MBox was used for some of the recording in Vegas, using it's internal clock source. I can't remember for sure at the moment what kind of card one of the other guy was using to record in Vegas - I think he might have used a SoundBlaster Live for a few of the tracks and maybe an Echo Gina or Turtlebeach Montego on other tracks, and they all would have been using the internal clock as well.

Now that I'm thinking of it - doesn't the SoundBlaster do some kind of resampling? I don't think the problem lies with the SounhdBlaster though because I know that at least some of the problematic tracks were recorded with the MBox.
Rikki_B wrote on 12/9/2004, 2:04 PM
Yeah, that's my workaround - hate to do it though because the basics were done with top of the line Apogee A/D's with a really good external wordclock and they sound so-o-o pristine....

So far I have not been able to find anything particular in common with the problematic tracks, though I'm sure they must have someting in common...
Youn wrote on 12/9/2004, 3:54 PM
Do you (or how do you) convert the WAV files before bringing them into ProTools? Do you bring in all the PC-converted files in at once or just the overdubs over the originally MAC-recorded files?

If you have sound forge (other similar editor's might work) I would try manually converting the files.
wobblyboy wrote on 12/9/2004, 11:34 PM
Here's a novel idea, why don't you just do your mixing in Vegas since they line up in Vegas. I been pleased with Vegas as a mixing tool and Sound Forge as a mastering tool.
pwppch wrote on 12/10/2004, 6:59 AM
>>doesn't the SoundBlaster do some kind of resampling?

Yes. The newer SB cards are hardwired for 48 kHz. If you don't use the ASIO drivers, then Windows will do a Sample Rate Conversion which can lead to all kinds of problems.

Again, 44.1 khz for one card does not have to be 44.1 for another card.

I'd be curious to know exactly how far off the problem tracks are. That is, if they are "fast" how early to they end in relation to the orginal material? If slow, how far past the end of the orginal material do they end.

Peter
Rikki_B wrote on 12/12/2004, 9:22 PM
I'll have to do a little more research - the project has been on hold for a while and I don't remember right now if the overdubbed tracks were fast or slow, but I'll check it out in more detail and then I'll get back to you on that, Peter. And thanks for your help!
Rikki_B wrote on 12/12/2004, 10:57 PM
I would love to mix the project in Vegas, but for what we're doing and what I have for a PC it just hasn't been quite up to the task. We have some songs that have an insane amount of tracks - like 80 or 90, and we do some pretty complex mixing stuff with lotsa subgroups and processing and a fair amount of edits, and Vegas pretty much just has a Grand Mal seizure well before we can get anything useful out of it I'm sad to say... I really love working in Vegas, and I hate Pro Tools, but Vegas just can't keep up with a fully decked out top of the line Pro Tools HD192 setup. Plus if we mix in the studio equipped with PT, we have access to a lot of killer analog goodies we can patch in if we want... and oh yeah, much better monitors too....

But I'd really love it if Sony would get fully behind Vegas as a pro audio app and get it up to speed so it could be a serious contender to ProTools. If there's anyone who could take on Avid/Digidesign's dominance of the pro recording world, it'd be Sony...
Youn wrote on 12/13/2004, 8:09 AM
There's nothing wrong with bouncing down in a few stages of mixing - sorta forces you to commit a bit more. For me, I like this method of working alot, but, yeah, it's not for everyone.
Rednroll wrote on 12/18/2004, 11:06 AM
"Vegas just can't keep up with a fully decked out top of the line Pro Tools HD192 setup. "

Yeah, no kidding Protools is selling you an expensive piece of hardware in a HD192 setup, so the hardware drivers are controlled by the same company selling you the software with it. So in other words they are integrated together more precisely. Vegas is expected to run with anything from a Sound Blaster, built in mother board sound cards, as well as more pro level sound cards with drivers written by a 1000 different companies. So the summary is Protools is selling you a system, and Vegas is selling you flexibility.

From the discription of your workflow if I'm understanding it correctly is that you're starting with some basic tracks then sending that out to the individual musicians to lay down their own parts, then trying to combine them all together in one master project. If this is the case then Peter is correct, it's due to a difference in the 44.1Khz clock sources of the individual systems. I'm surprised the "former Chair of Berklee's Engineering and Production Dept." wasn't able to come up with this logical answer. It's as simple as this. Take the same CD and play it from 5 different CD players from different manufacturers and record the outputs into the same DAW. Once you line up all the start points, they will not be in sync, because although all the CD players are playing at the same 44.1Khz clock, they are all not exactly 44.1Khz, there are clock tolerance differences. I've seen this same thing in some work I've done. Here's what I did. I recorded audio from various satelite radio stations. I opened that recording in Vegas and then extracted the same song from CD. There was no way to get these two identical songs that had been played on different CD players to line up before some major drift would occur. The same principle is happening in your workflow. You are expecting 5 diiferent systems to have the exact same clock rate......IMPOSSIBLE!!!.

You need to change your workflow for this to work. You need to record the basic master tracks and then pass this project to the 1st musician to add their parts, then pass this composit project to the next musician to add their parts....so on and so forth.

Studios that do this type of work, where engineers are working on different sections at the same time and share files over a network have one thing in common. There is a master word clock generator that all the individual rooms sync to. Thus they are all using the same clock.
farss wrote on 12/19/2004, 11:27 PM
I've just finished my first music video, nothing complex just a guy and a piano. I recorded the audio through a Firewire 410, vision and reference audio through PD 150. Audio was recorded at 24/48 into a laptop and much to my surprise the whole thing held sync for each recording session, longest being 10 minutes. I was prepared for drift problems but none arose. Probably as I was only having to hold vision and sound sync where 1 frame accuracy would be acceptable things may have drifted a small amount and I wouldn't have noticed However on a couple of the takes I had audio markers at the start and end and in sofar as I could see there was no drift between the camera and the 410s clocks.
All that said though even between two cameras this can be anything between a small and a big issue. It seems Cannon and Sony use significantly different versions of 48KHz which by design gives major sync drift.
Bob.