HDV audio sync

[Ron Evans]

With DV there was a difference between consumer DV and DVCam or DVCPro in that the audio was in sync with video and the consumer DV had the possiblility of being out by a third of a frame. What is the equivalent with HDV? Because of the encoding to MPEG2 is ALL HDV in sync to the same extent? In other words if I have two HDV cameras will the audio waveform be in the same relationship to frames for both when they record the same scene? I ask this because in an NLE my TRV50 and FX1 recording DV show a definate difference of maybe 1/4 frame out of sync with the TRV50 not always staying at this fixed relationship ( as expected) it makes mixing audio from these two cameras difficult. I wondered if this issue would be solved with two HDV cameras recording HDV. The same question would apply to AVCHD too. I am trying to avoid complicated multi-track audio recording to get the quality I am after.
The only quality solution I can think of is using a multitrack recorder ( small desktop workstation from Fostex or Yamaha) and feeding the stereo output to the cameras. This would allow me to remix the audio but still have an audio sync track to work with from the cameras on the timeline.

Ron Evans

[Heath McKnight]

Only time I had some issues with sync was when using a workflow with iMovie and the QuickTime MPEG2 reader (I can't remember the official name) back in early 2005. But that was the QT MPEG2 app that had a problem. Capturing, editing and outputting in iMovie HD didn't give me problems with sync, just the workflow to get into FCP HD (4.5).

With FCP 5, no problems. Not sure about other NLEs. And, of course, nothing with sync issues in raw video, either, playing back from a deck or camera (this includes Sony, Canon and JVC HDV cameras I've used, in all frame rates that are offered).

heath

[Ron Evans]

Hi Heath
The sync problem is not the NLE it is the problem of using consumer cameras that are not gen locked AND have non locked audio!!!! The difference can be seen by expanding the timeline to single frame and looking at the waveforms. Fundamently unless the cameras are gen locked and have locked audio the waveforms will not line up and can't be made to either with frame clock. Most of the time I can get mine within 1/4 a frame but that is still obvious in some circumstances. Most of the time the mix is very heavily biased to one camera and the other just provides some "air". But some time I need to use both and that gives me a problem of "reberb / echo" that takes a lot of fiddling to get rid of. I am thinking of going back to recording from to a multi track audio recorder and feed a stereo to the cameras for sync later. These are theatre production so I like to get the sound as good as I can. USe Sony cams fo my two cam shoots and start with remote to try and reduce the fact they are not gen locked!!!!
Ron Evans

[Douglas Spotted Eagle]

Your problem has nothing to do with locked vs unlocked audio; it has everything to do with relative distance from audio source to recording source.
That said, if you can get it within 1/8 of a second, it doesn't really matter anyway. You'll find that using a master audio source with ref from your other cam(s) will provide you the best workflow, but YMMV.
The drift of non-genlocked cams is at worst a couple frames over an hour. Certainly not enough for anyone to get uptight about given the price class any of these cams are assigned to.

[Heath McKnight]

Spot,

Thanks for the input!

heath

[Ron Evans]

The production I did two weeks ago and have been editing since( my wife thinks I am a little obsesive!!! that's what happens when your retired!!!) used my two cams. A TRV50 and FX1 beside each other at the back of the theatre, about 60 feet from stage. FX1 used a ROde Video mic and the TRV50 used an AT825 through a Studio 1 XLR converter. Part one was 1hour and 23 mins and part 2 1 hour and 5 mins. Recorded DV LP on both cams. I had to adjust sync in editing twice for the TRV50 in part 1 and once in part two to minimize echo from the mixed sound noone would have known from the picture. Both cases moved just one frame. I expect that in the hour and 23 mins the actual slip was less than two frames it was just that i had no ability to move less. So it was a bit ahead one time and a bit behind the other( less than quarter of frame as indicated by waveform on the NLE ( Vegas for audio mixing though used Edius for main edit). I am just trying to improve and clearly sound is the next in line as I am happy with my video stuff now within the limits of what I have ( will change the TRV50 with something after Christmas. Waiting for a high quality cam with internal HDD.)
Thanks for your input. Recording all to a multitrack and feeding a guide track back to cameras is likely the way I will go in the future.

Ron Evans