Lost audio sync during long DV Capture

[DigitalDave]

I would always suggest you capture with "when audio sampling changes" ticked as a minimum in Deck Settings. Any dropout disturbance can freak the sync on a continuous file. That said 8 mins....thats 0.5% and given the subject being recorded, I suppose it is difficult to identify where it went wrong or whether it was a gradual drift. Another factor is whether DVCam has been selected in the I/O Hardware settings. DV does not have locked sync.

DD

[Jim N]

Thanks Anton,

I checked the file and it has 48k audio and the Edius preset includes 48k audio so it still remains a mystery.

Your assertion that the wave audio limit of Windows is three hours would explain why "Storm Video" stopped recording but why didn't "DV capture" do the same?

It's all very curious.

[antonsvideo]

1. wav audio length limit of windows is 3 hours, so if you capture longer than 3 hours, the file will break every 3 hours

2. was the camera set to 48k audio, if 32k then audio drift will also happen unless you open the file in OHCI 32k preset

[Jim N]

Hi guys,

Had an interesting experience recently where I had to record a specimen in a test rig under cyclic load for a continous period until it failed. I set up a system feeding DV into my old Storm card using Storm Video as the record application but it only caputured the feed from my camcorder for three hours then quit. I eventually used Canopus DV Capture which seemed to capture happily to the external hard drive. Nearly 27 hours later the specimem broke. Believe it or not that 338 GB file loaded to an Edius 3.62 timeline but the audio was way out of sync towards the end of the recording.

I was surprised that it was out of sync. The audio is running slower than the vision so the specimem broke before any audio of it breaking appears. It is hard to tell, but the audio may be around 8 minutes behind the video, perhaps more!

Has anyone any idea why audio sync would be lost in this instance. I was under the impression that audio would be perfectly locked to the vision during acquisition. Would it be the camera file or Edius where the delay is occuring and why?