Audio Timecode with protools, syncing with video

[neve2254e]

This stuff gets messy quick.
If you can send your video editor all your ISO's so that when they send you an AAF or an OMF all the audio with their edits will follow.

There also used to be a program called Titan that would conform your field recording to the EDL.

I haven't done this in years. I just try to send the editor all the audio so it follows all the way through the edit.

Regards
Eric

[grayter1]

Originally Posted by neve2254e

This stuff gets messy quick.
If you can send your video editor all your ISO's so that when they send you an AAF or an OMF all the audio with their edits will follow.

There also used to be a program called Titan that would conform your field recording to the EDL.
Eric

Editor is already working with 8 to 10 camera ISO's, plus the line cut in Premiere. They never go with just the line cut and always want to tweak and push in on shots to cover any director mistakes made live. The multi-cam edit alone is taxing their system enough without adding an additional 40 to 60+ audio tracks. At least, that's what I've been told in the past. Frankly I'm not sure why we even do a line cut. The final never resembles what actually went to disk on show day.

It's not really all that bad - IF - the TC is good through out the process. It's just time consuming. The good news is working this way forces them to lock the show before I get it. Other than the editor continuing to work on color correction while I mix, changes after I get it are almost nonexistent. We often have a tight deadline to get it to distribution by mid December for Christmas air dates.

tg

[neve2254e]

Yeah. Just was thinking of a video editor with 64 tracks of audio. I think Media composer just added that feature.

[trevora]

I've got Prem too - should work the same in principle. Pretty easy to test.

Here's 2 EDLS - v1 is the 'as recorded' v2 is a 50min cut down/rearranged. (had to copy/paste as .edl or .txt or .zip not allowed as attachment)

Paste a big load of audio against v1 to represent the three chunks of recording then get a trial of EdiLoad and try and do a change to v2 - be interested to know if it would work (it might need genuine 'hot mix' files to do the change list to - I don't know how PT works in this regard)

TITLE: RX_TEST_v1
FCM: NON-DROP FRAME
000001 RX_TEST AA C 10:02:00:00 12:00:12:00 10:02:00:00 12:00:12:00
*FROM CLIP NAME: RX1
000002 RX_TEST AA C 15:00:00:00 16:12:14:15 15:00:00:00 16:12:14:15
*FROM CLIP NAME: RX2
000003 RX_TEST AA C 17:34:12:14 18:12:15:14 17:34:12:14 18:12:15:14
*FROM CLIP NAME: RX3
*


TITLE: RX_TEST_v2
FCM: NON-DROP FRAME
000001 RX_TEST AA C 10:02:00:00 10:05:32:17 10:00:00:00 10:03:32:17
*FROM CLIP NAME: RX1
000002 RX_TEST AA C 15:18:36:00 15:22:05:06 10:03:32:17 10:07:01:23
*FROM CLIP NAME: RX2
000003 RX_TEST AA C 10:07:47:12 10:10:39:07 10:07:01:23 10:09:53:18
*FROM CLIP NAME: RX1
000004 RX_TEST AA C 15:53:33:15 15:55:30:18 10:09:53:18 10:11:50:21
*FROM CLIP NAME: RX2
000005 RX_TEST AA C 10:10:39:07 10:13:34:05 10:11:50:21 10:14:45:19
*FROM CLIP NAME: RX1
000006 RX_TEST AA C 17:34:12:14 17:37:34:10 10:14:45:19 10:18:07:15
*FROM CLIP NAME: RX3
000007 RX_TEST AA C 15:22:05:06 15:24:28:17 10:18:07:15 10:20:31:01
*FROM CLIP NAME: RX2
000008 RX_TEST AA C 17:37:34:10 17:39:30:11 10:20:31:01 10:22:27:02
*FROM CLIP NAME: RX3
000009 RX_TEST AA C 10:15:06:15 10:17:50:23 10:22:27:02 10:25:11:10
*FROM CLIP NAME: RX1
000010 RX_TEST AA C 10:34:48:03 10:38:02:23 10:25:11:10 10:28:26:05
*FROM CLIP NAME: RX1
000011 RX_TEST AA C 10:56:59:00 10:59:00:19 10:28:26:05 10:30:27:24
*FROM CLIP NAME: RX1
000012 RX_TEST AA C 11:30:12:02 11:32:20:05 10:30:27:24 10:32:36:02
*FROM CLIP NAME: RX1
000013 RX_TEST AA C 15:13:08:07 15:15:42:07 10:32:36:02 10:35:10:02
*FROM CLIP NAME: RX2
000014 RX_TEST AA C 15:36:16:14 15:38:58:20 10:35:10:02 10:37:52:08
*FROM CLIP NAME: RX2
000015 RX_TEST AA C 10:45:18:03 10:48:10:13 10:37:52:08 10:40:44:18
*FROM CLIP NAME: RX1
000016 RX_TEST AA C 15:38:58:20 15:42:37:17 10:40:44:18 10:44:23:15
*FROM CLIP NAME: RX2
000017 RX_TEST AA C 15:51:09:01 15:53:33:15 10:44:23:15 10:46:48:04
*FROM CLIP NAME: RX2
000018 RX_TEST AA C 17:52:53:21 17:56:05:17 10:46:48:04 10:50:00:00
*FROM CLIP NAME: RX3
*

[trevora]

Originally Posted by grayter1

Frankly I'm not sure why we even do a line cut. The final never resembles what actually went to disk on show day.

Otherwise the director / cameras have no idea what the coverage is - at least there should be one camera steady and in focus ;-)

[grayter1]

Originally Posted by trevora

I've got Prem too - should work the same in principle. Pretty easy to test.

Here's 2 EDLS - v1 is the 'as recorded' v2 is a 50min cut down/rearranged.
.
.
.

Thanks. I'll look into this.

tg

[trevora]

Originally Posted by neve2254e

Yeah. Just was thinking of a video editor with 64 tracks of audio. I think Media composer just added that feature.

Wrangling multiple tracks with pictures is not trivial, even if there will be a sound dub sorting it out after the fact. Overlapping elements, trimming, maintaining sync with picture and sync cues as the whole edit changes...

I just did a major sporting event where the recording & delivery was 16 channels. Attached screen shot gives an idea of how it looks simple, but there's a lot of manipulation required & potential gotchas. And that is just a simple line record turnaround. The empty tracks are for overlaps - and would've got used a lot more if there had been any actual crowd!

[trevora]

Originally Posted by grayter1

The multi-cam edit alone is taxing their system enough without adding an additional 40 to 60+ audio tracks.

Actually, in PP, you could use 2x adaptive tracks (they go up to 32 per track I think) for the isos - would be pretty easy to keep those along for the ride, they wouldn't even need to listen to them.

Wrangling 64 individual extra tracks, on the other hand... (well in MC I suppose you could put them in 8x 7.1 containers)

[Bigbottomend]

I used to do a lot of re-conforming work for making international versions of TV shows using Virtual Katy: https://virtualkaty.com. It's not cheap but if you do a lot of this kind of work it's a lifesaver.

There's a bit of a learning curve for the workflow, but basically once you learn how to correctly filter the EDLs provided from the NLE, the job takes about 30 seconds.

If the editor is working with a stereo hot mix, they simply provide an EDL of audio only - from memory, CMX3600 was the format that worked the best.

[trevora]

Originally Posted by [email protected]

What is happening is a tv production there there are multiple cameras and they just send an audio ltc to me. I record into PT straight from the console with AVB.

Great, so thats the master timecode reference.

No idea how PT records but surely it can reference that TC in the wav files (the audio LTC is just the means of TC distribution, used to have to be recorded on a track now it's mostly stored as metadata in the file).

This is the most important thing you can do to simplify your conform.

They edit the footage and i have to mix the songs performed as well as the lav mics etc.

Ok so you mean mix live here and they are editing from seperate video recordings (with guide mix).

during recording i cut very often( mainly to make sure nothing goes wrong) so in the end i dont have a single recording to just add the unedited footage to and then manually conform the files which( for a 2 hour program out of a 6 hour shoot) is really hard work!!!

Breaks in record are fine

My goal is to be able to automatically sync the audio to the edited video cuts i get! I hope its not impossible!!

Field recorder workflow is your friend here, I believe.

The picture editor should send you an AAF, which will contain the guide mix.

You should be able to import that to a session, press the field recorder workflow button (I jest - I don't know PT, you'll have to read the manual) and that will, I think, let you conform based on various criteria.

Normally this is complicated by sources recorded by multiple recorders on multiple days and you need to conform by soundroll/tape/ other metadata + timecode but yours should just be by timecode.

Result should be expanded tracks where your multitrack records match the video guide mix (might need slipping by milliseconds to match perfectly as the video should be genlocked + audio probably wasn't - your breaks in record should help in this regard)