MultiCam Sync and Drift - A built in fix?

HI All-

Most of my work is done within DRS MultiCam (15.1 currently). I use 3 Cams and a separate audio feed. In the past I thought a MultiCam going into a drift with audio sync could be remedied within an opened MultiCam timeline. I do that and the IN points don't match.

Firstly, am I missing something? Perhaps a tool within DRS? Is there a workflow workaround?

Am I looking at breaking my timelines into smaller pieces? (Cut-point at 5 minutes off of where I notice the first sign of drift?) I don't have the timecode tools yet, but surely there must be something I can manually do with waveforms?

Thank you for your time.

In multicam, the cameras each have to have identical timecode and genlock reference, precisely so that they will not drift from each other and a separate audio track. Were the cameras handled that way? If not, they're going to drift and you will manually have to correct it. Sometimes it's very subtle (like 1-2 frames every 5 minutes), and sometimes it's worse.

I have done big rock concert projects where we had upwards of a dozen cameras and absolutely nothing drifted anywhere, because every camera had a Denecke SB-T timecode/sync generator jammed about an hour prior to show start, using the timecode supplied by the sound department. Without that, I might still be syncing and struggling somewhere in a a dark room.

In multicam, the cameras each have to have identical timecode and genlock reference, precisely so that they will not drift from each other and a separate audio track. Were the cameras handled that way? If not, they're going to drift and you will manually have to correct it. Sometimes it's very subtle (like 1-2 frames every 5 minutes), and sometimes it's worse.

I have done big rock concert projects where we had upwards of a dozen cameras and absolutely nothing drifted anywhere, because every camera had a Denecke SB-T timecode/sync generator jammed about an hour prior to show start, using the timecode supplied by the sound department. Without that, I might still be syncing and struggling somewhere in a a dark room.

Thank you, Mark. I am going to have to budget for the TC generator.

The issue is I don't currently, and I have used the 3-clap method, I mark those as IN points (3 Cams and Separate audio), 15-20 minutes into the multicam and the drift becomes noticeable. Even if I cut the multicam timeline, how am I correcting manually? - There are no claps, perhaps some discernable peaks/valleys.

The issue is I don't currently, and I have used the 3-clap method, I mark those as IN points (3 Cams and Separate audio), 15-20 minutes into the multicam and the drift becomes noticeable.

Take the 3 cameras in a room and have them shoot a clock for 15 minutes. Clap at the head and clap at the end. Now take the files, lay them in a timeline, and figure out what kind of drift each camera has. If you're lucky, the drift is consistent and A cam will always be +1 frame every minute, B cam will be +2 frames every 5 minutes, and C cam will be dead on. Once you have a number for each camera, you can figure out how to compensate for the offset.

If you don't have access to the cameras anymore, you'll just have to eyeball it and fake it. I did a 22-camera concert film (Rolling Stones: Shine a Light), and I think half the picture is all my eyeball sync. We had no drift on that project, but we did have to manually sync every single camera because in most cases, there were no timecode slates and this was film with no timecode on the picture. It can be done if you're organized and figure out how to read lips.

The issue is I don't currently, and I have used the 3-clap method, I mark those as IN points (3 Cams and Separate audio), 15-20 minutes into the multicam and the drift becomes noticeable.

Take the 3 cameras in a room and have them shoot a clock for 15 minutes. Clap at the head and clap at the end. Now take the files, lay them in a timeline, and figure out what kind of drift each camera has. If you're lucky, the drift is consistent and A cam will always be +1 frame every minute, B cam will be +2 frames every 5 minutes, and C cam will be dead on. Once you have a number for each camera, you can figure out how to compensate for the offset.

If you don't have access to the cameras anymore, you'll just have to eyeball it and fake it. I did a 22-camera concert film (Rolling Stones: Shine a Light), and I think half the picture is all my eyeball sync. We had no drift on that project, but we did have to manually sync every single camera because in most cases, there were no timecode slates and this was film with no timecode on the picture. It can be done if you're organized and figure out how to read lips.

My oh my, that is a plan!!!! Marc, thank you for providing that detailed description.