another audio sync problem

mark0514 wrote on 9/1/2010, 6:45 PM
I am recording with a HD canon vixia hfs 100. I am recordning audio on a PC with audacity (I know I could be using sound forge). When I export a wave file and pull it into movie studio, the sync is off. I can sync the beginning, but 20 minutes in it (the audio) is lagging behind. I dont have the computer to capture real time. In fact, I have been converting the .mts file to a 720p wmv prior to pulling it into vegas. Any ideas what's slowing down my audi?

Comments

musicvid10 wrote on 9/1/2010, 7:45 PM
As has been discussed, any two clocks that are not synced by a wire will drift apart. The amount of drift can range from insignificant to major. The magnitude of net drift is increased by the length of the program.

Try chopping one of the tracks into 9-10 minute chunks, being careful to make your cuts exactly at zero crossing points in relatively quiet areas, and syncing the waveforms at the approximate centers of each clip.

The other method, which I do not prefer, is to time squeeze/stretch one clip so that the beginning and ending waveforms match each other.

Other than that, the clock differences between your computer and camera cannot be compensated for in the recording stage by normal means.

Hope this helps.
Chienworks wrote on 9/1/2010, 9:05 PM
Not sure why you don't prefer the time squeeze/stretch method. In effect this corrects the difference between the clocks. Given the problems of the sources, it's really about the most perfect solution possible.
musicvid10 wrote on 9/1/2010, 9:34 PM
For me Kelly, the difference is two words; quantization noise.

The other reason I have discovered, is on long recordings the intermediate drift can be quite noticeable. Just having sync points at the beginning and end is not enough to ensure sync at any given time in a 2-3 hr. program, for instance.

By splitting the dlips judiciously, I can make the gaps or overlaps inaudible. And Pluraleyes makes the syncs perfectly.

I should have added that I would only do this in a mix or surround project, and would never suggest it if I was replacing the only audio track with a separate recording. I would be more prone to time squeeze or stretch the video in this instance if the net difference was only a few frames per hour.

With my particular type of hearing loss, from which many old rockers suffer, the slightest bit of q-noise is enough to drive me up a wall. Just a little worse than fingernails on the chalkboard. It's also the reason I gave up making midi recordings a long time ago.
mark0514 wrote on 9/2/2010, 6:01 AM
Thank you both for your input. I definitely have some things to try. I am in fact replacing the only audio track with a seperate recording. My videos are never more than an hour. Can you go a little deeper into the procedure for time squeeze or stretch? I must admit I am not familiar with it.

Thanks again.