Avid, DV, Timecode, and Double System Workflows

[Steve House]

Discussions of various timecode based, double system sound DV workflows come up from time to time in several audio forums and it is suggested that Avid is able to auto-sync audio and video using the following scenario. Not being an Avid user and not aware of exactly how it handles BWF files, I thought I kick it out amongst the gurus here and see if it's practical or not. Here's the proposed situation...

1. Picture is shot with a DV or HDV camera that does NOT have any provision for external timecode I/O or genlock. Since it does not have TC I/O, its timecode cannot be slaved to an external source or vice versa.
2. Production sound is recorded as a timestamped BWF file on a timecode capable audio recorder such as a Sound Devices 7xxT series recorder, the audio recorder to serve as the timecode master.
3. Timecode is output from the audio recorder in LTC format as an audio signal which is recorded to one of the in-camera audio tracks - the other track either being unused or recorded with audio as a scratch/backup track. The end result, of course, will be a DV/HDV video file containing picture and VTC plus stereo audio, one channel of which contains an LTC signal - the video timecode being different from that encoded in the audio due to the fact that synchronising the two generators to the frame isn't possible because they aren't talking to each other.

What is proposed for post is that the video file will be captured to the computer and added to the Avid timeline along with its audio. Avid's LTC reader will read the timecode encoded in the video file's audio and generate an auxilliary timeline identical to that code (and different from the code on the main timeline due to the fact that because they come from two independent generators, the numbers encoded in the VTC and the LTC will be different for each frame). Then when the 'real' production audio in BWF is added as its own track, Avid will be able to line up the timestamp in the incoming BWF with the matching timecode on the auxilliary timeline thereby establishing sync between the camera audio and the BWF audio. Lock the BWF and the video file together; mute the video file's own audio portion, and voila! perfectly sync'ed double system audio without manual adjusting or reading of slates required.

Now the question I have for you Avid gurus is ... does it really work like that or does some variation of that workflow exist that would allow a BWF timestamp to auto-align to LTC encoded on another audio track?

[David Parks]

Steve,

Avid works at handling BWF audio + metadata container files but you might need this time stamp conversion tool. You might check over on the Avid forum for other workflows for sample rate and pull down import settings.

http://www.avid.com/exchange/forums/...howThread.aspx

http://www.24p.com/AvidInsider2.htm (conversion tool)

I'm not sure I quite follow you on the one channel of longitudinal (LTC) on one of the audio tracks of your HDV/DV when you have VITC already. I might be wrong, but your aux TC track would come from your BWF. Is the LTC from your HDV/DV audio reference/scratch? If so then that makes sense.

http://www.24p.com/BWF_recording.htm

I'm sure you already know this: a backup to all this great workflow technology, I would also slate eveything. After all, it has been working well for 80 years.

I hope this helps,

Sounds like a fun project. Good luck.

David

[Steve House]

Originally Posted by David Parks

Steve,

Avid works at handling BWF audio + metadata container files but you might need this time stamp conversion tool. You might check over on the Avid forum for other workflows for sample rate and pull down import settings.

http://www.avid.com/exchange/forums/...howThread.aspx

http://www.24p.com/AvidInsider2.htm (conversion tool)

I'm not sure I quite follow you on the one channel of longitidinal (LTC) on one of the audio tracks of your HDV/DV when you have VITC already. I might be wrong, but your aux TC track would come from your BWF. Is the LTC from your HDV/DV audio reference/scratch? If so then that makes sense.

http://www.24p.com/BWF_recording.htm

I'm sure you already know this: a backup to all this great workflow technology, I would also slate eveything. After all, it has been working well for 80 years.

I hope this helps,

Sounds like a fun project. Good luck.

David

I echo your advice to use slates, but the question and this suggested answer keeps coming up in various discussion groups. The problem is there's no way to jam the timecode generators in the audio recorder and the camera to the same value since the camera doesn't have TC input or output ability, not does it accept genlock. Thus the VITC recorded on the DV tape and the timecode associated with the BWF audio file will never match each other. Without a TC slate at that stage there's no common denominator between the video+2 audio tracks on the camera tape and the BWFs TC data in the file coming from the audio recorder. So what is suggested is to sacrifice one of the camera's audio channels and feed it LTC from the audio recorder which of course would actually match the BWF code.. then to get Avid to use that LTC track INSTEAD of the VITC to control where the BWF file lines up when it's added to the timeline. Frankly I'm skeptical it would work but since I don't know Avid I didn't want to poo-poo the idea.

[David Parks]

In your case you don't need to jam sync while shooting, Avid AutoSync takes care of the TC offset for you. So your video time code doesn't need to be the same as your audio time code. That's the cool thing about TC offset. Old time audio post houses having been using them on external TC reader/generators to lock video and 24 track analog since the 80's.

I wouldn't poo poo the idea. Once you get your BWF files imported with the correct pull down and sample rate conversion, Syncing the audio in Avid is the way to go using AutoSync. You basically highlight the clips you want to sync, in this case, after creating group clips, select the video files and corresponding imported BWF files, choose AutoSync. A dialog ask what Aux TC to sync to and hit okay. It figures out the offset for you. You now have clips that are in sync before you edit in the timeline. The slates are extra protection if you are a few frames off, then you can resync tighter using AutoSync and base it on in points or out points. (Head slates or tail slates)

The main scary part is in any sample rate or frame rate conversions because if you aren't careful your audio may be 10% faster or slower. Then it becomes a major pain.

As long as you stay 48k and 23.976, then you should be fine. I assume your shooting 24p???

[Steve House]

Originally Posted by David Parks

...
The main scary part is in any sample rate or frame rate conversions because if you aren't careful your audio may be 10% faster or slower. Then it becomes a major pain.

As long as you stay 48k and 23.976, then you should be fine. I assume your shooting 24p???

LOL - I'm not shooting it at all, at least not now. The subject comes up frequently as people wonder how to use timecode with various DV projects and my interest up to now has been primarily academic. This approach has been discussed as workflows for HD, HDV, and SD DV projects so frame rates for the video could be either 24p or standard NTSC 29.97

[David Parks]

http://www.24p.com/downloads/MPEG_LA_FINAL.pdf

Here's a ppt that lays out the frame rate and sample rate issues and conversions.

The one other thing I failed to mention is that cheaper digital audio recoders run at 60Hz so they record at 29.97. So your BWF's are 29.97. In a 24p or 23.976 project, you have to instruct Avid or even your audio editing app. to do an audio pull down and slight sample rate conversion on import to restamp them as exactly 47952 Hz.

Believe it or not, it is much easier now than it was manually syncing or mathematically coming up with an offset using a TC calculator.