0

Hey guys,

I've ran into a bit of trouble. I have 20-minute clips of the film I'm working on, exported by the director (who is also the picture editor). I've also got audio separated on 3 tracks, again directly exported from Vegas.

Now, what happens is that they don't match up perfectly. I've tried looking at it in both Pro Tools and Reaper, the result is baffling.

In Reaper, the audio track is shorter than the video+audio track (Reaper combines video + audio on 1 track).

In Pro Tools, which puts the original's audio on a separate track, the original audio is far longer than the video, while the exported audio is shorter than the video. What's happening here? It's an MP4 file, 25 frames per sec.

Here's a screenshot in PT: https://dl.dropbox.com/u/48858350/len/video-len.png

Here's a screenshot in REAPER: https://dl.dropbox.com/u/48858350/len/video-length.JPG

This is what GSpot has to say about the video file: https://dl.dropbox.com/u/48858350/len/gspot.PNG

My sample rate matches up, all files are 48 kHz. The sessions I'm importing in are 48 kHz. I've tried playing with Pro Tools Pull up/down setting, seems like it's not that.

I'm stumped! Could it be some kind of encoding problem on the director's side?

Thanks, I appreciate your time and help a lot!

3 Answers 3

2

I hate idiosyncrasies like this. First things first:

  1. How many frames is it off?

  2. Have you asked the director how he rendered?

  3. Are you working from his rendered files or something like an OMF?

I've had similar issues where the audio will be about 1 frame longer than the video (when going from a Pro Tools audio render to a Premiere video render). Sometimes it would happen, sometimes it wouldn't, very erratic. If it's only a frame or two I would say just mark it to a software problem and don't worry about it, but you say "far longer" so I'm assuming it's off by a few seconds.

You've done a lot of troubleshooting yourself making sure the sample rates and frame rates are the same. I'm sure you'll weed out the problem; any other Vegas users experiencing this?

2

The most important thing is to find out is do you have the correct video?

if you import the audio from the video and compare it to the exported audio you were given Does the audio drift out of sync (ie there is a ratio problem) or does it suddenly become out of sync (ie the audio & video are from different pic cuts)

If its a ratio/drift issue

Put the timeline in ProTools to display samples (not timecode or minutes/seconds or bars/beats etc)

  • Measure the length of the video, note the number of samples (it will be a large number)
  • Measure the length of the audio, note the number of samples (it will be a large number)

Now get a calculator and work out what the percentage difference is ie divide short length by long length, does this ratio look familiar?

Now do a bunch of calculations to see if that result is same as 44100/48000? 23.976/24? 23.976/25? 24/25? 25/29.97? 23.976/29.97? 24/29.97?

I once had this problem and it turned out they had somehow speed changed the audio twice, so in my case it was the 24/25 ratio, but doubled!?!

0

Here's a workaround that's fast and gets rid of calculation and other nuisances...
(Although I used Apple programs, I'm sure the principle will work as well with Win apps.)
I had a video file ("file.mp4") that in Avidemux reported as having a video length of about 29 min, an audio length of over 36 min. No "delay" or preferences value helped me, as the delay lengthened towards the videos end. Real SH.. !

That's the way I worked around it:
- I used Quicktime player 7 to export sound as WAV >> "file.wav".
- I loaded "file.mp4" into Avidemux and ...
- added "file.wav" using this menu item:
. . . . . . Audio > main track > Audio source (External WAV) > Open "file.wav"
- ... for the Audio track I used MP2(lav), 160 bps, to reduce WAV's size

All this took much less time than all my previous tries with delaying etc.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.