Hello, and sorry for what is perhaps a silly question.

I just captured a 54-min DV-clip with "Capture now", and it comes in fine, and while the audio is fine in the beginning, at the end of the clip it's about 01:11 behind video. Seems like the audio is just a little bit slower than the video-track, and it pretty much ruins the whole thing.

Is there anything I can do to fix this, is this something caused by bad settings or something else? (Easy setup DV Pal Anamorphic and captured, like I usually do.)

FCP6, latest updates of FCS and Leopard and on a Dual 2,5ghz G5.

Any ideas would be greatly appreciated!

So, I've checked some more, and see that when I look at the clip properties, FCP reports the audio to be in 48000,5khz.. Isn't it supposed to be 48000khz sharp? And wouldn't this maybe cause it?

Tried resampling it in Soundtrack Pro, but just got an audio file that was slightly longer than the video.. When I set the speed of the video to 99.95 it's pretty spot on at the end, but I just wish there was some other way to fix this?

Do you have to capture the whole clip?

If not, try cutting it up a little and using Capture Clip. Or even use Capture Clip and use the whole thing if that is possible for you

Do you have to capture the whole clip?

If not, try cutting it up a little and using Capture Clip. Or even use Capture Clip and use the whole thing if that is possible for you

I tried capturing 15 minutes instead, but got problems with that too. I usually just log and capture, but this tape has a lot of long clips, so I figured I'd just get it all.

Is there a difference between using Capture Clip and Capture Now?

Some cameras have inaccurate crystal for sound, with long captures audio and video drift apart. The easiest way to fix is to capture in shorter segments, or make short cuts in a long clips and readjust the sync.

Yes, thank you both for your suggestions, I had to go back and log and capture as usual, and then it was fine.

I was just hoping there would be a fix to long clips too, as they are sometimes unavoidable.

I have an XHA1 too, that camera should be more precise when it comes to this right?

It's most likely an issue at the computer end of things - i.e., the software. Both Macs and Windows PCs are prone to incorrect processing of the audio data in the DV stream.

This is a common misperception that cheap, out-of-sync audio crystal oscillators are responsible. This isn't true. The DV specification is, well, very specific about this. The amount of audio data recorded *per frame* is allowed to vary slightly (within specific limits) but the average across frames *must* be correct. This is the so-called unlocked mode. For locked mode, the specification states that each frame *must* have a prescribed amount of data.

DV decoders/NLEs that fully understand the specification generate drift-free audio.

The common cause of the drift is (unnecessarily) demultiplexing the audio and video streams such that you have a separate audio track and then assuming an (incorrect) audio sampling rate. Windows suffers this through the Type-2 DV AVI file (a kludge to support NLEs that use the older and (ought to be) defunct Video for Windows). Type-1 DV AVIs don't suffer this because the audio and video are left multiplexed in their original state.