video sync issues

Daniel L Miller · September 5th, 2005, 11:55 PM

just to clarify AJ, its not that the audio is getting out of sync with its own video, its that i have two cameras getting out of sync with each other. both are at 29.97, the sample rates are the same for both of them. i dont get messages of dropped frames or anything, its just that as the two tracks play side by side, the get out of sync with one another. if you knew that all along then i apologize, if you perhaps did not, i hope this clarifys a bit.

-daniel out

A. J. deLange · September 6th, 2005, 06:24 AM

Daniel,

Yes, I did understand that. The thread started to drift when it was suggested that what you were seeing was a clocking problem, which indeed Canon has been accused of having.

To sumarize:

1. You cannot expect two cameras to run at the same rate unless they are genlocked (and that's what genlock is for) because their crystals are sensitive to temperature, voltage, age and vibration. This can be ameliorated to some extent by, for example, putting the crystal in an oven inside an oven for extreme temperature stability but we wouldn't want this done in our prosumer cameras because of size, weight, power and especially cost. Given this I think you must live with the situation and the only thing I can suggest is that you try to keep the cameras at the same temperature and resync in post.

2. If Canon does or did have an unusual audio sampling rate FCP is apparently capable of handling it such that audio and video stay in sync WITH A SINGLE CAMERA.

Cheers.

Daniel L Miller · September 6th, 2005, 09:05 AM

awesome, thanks man. i appreciate all the help. i learned allot from this thread.

-daniel out

K. Forman · September 6th, 2005, 02:42 PM

From what I have found, MiniDV in general tends to drift more with longer captures. Do shorter batch captures, and it should help you out. It did the trick for me.

A. J. deLange · September 8th, 2005, 06:46 AM

Maybe this should be in a new thread but it is relevant here as well...

Last night I made a 22 minute XL2 DV of me clapping my hands in front of the lens every 5 minutes or so. Simultaneously I recorded sound with a separate stereo microphone through a digital interface using sampling clock derived from the XL2's composite video out. I captured the DV tape using final cut pro, turned the digital recording into an AIFF, imported it into Final Cut Pro and lined it up with the XL2's audio tracks. Both audio tracks stayed in precise alignment with the video and with each other for the full 22 minutes but both seemed to be 2 frames early with respect to the video. In playing back (through FCP) the best alignment of sound and video was obtained when the sounds were delayed 2 frames. Has anyone else seen this? Is there some offset parameter in some preference menu of which I am unaware? Am I finally losing my marbles?

Steve House · September 10th, 2005, 06:09 AM

Quote:

Originally Posted by A. J. deLange

Maybe this should be in a new thread but it is relevant here as well...

Last night I made a 22 minute XL2 DV of me clapping my hands in front of the lens every 5 minutes or so. Simultaneously I recorded sound with a separate stereo microphone through a digital interface using sampling clock derived from the XL2's composite video out. I captured the DV tape using final cut pro, turned the digital recording into an AIFF, imported it into Final Cut Pro and lined it up with the XL2's audio tracks. Both audio tracks stayed in precise alignment with the video and with each other for the full 22 minutes but both seemed to be 2 frames early with respect to the video. In playing back (through FCP) the best alignment of sound and video was obtained when the sounds were delayed 2 frames. Has anyone else seen this? Is there some offset parameter in some preference menu of which I am unaware? Am I finally losing my marbles?

Were you far enough away from the camera and mic that you're seeing the delay caused by the time for sound to travel from your clap to the mic?

Hang on to that tape - sounds like just the thing for an Avant-Garde art museum installation "performance piece" - "One Man - Slow Clapping" LOL

A. J. deLange · September 10th, 2005, 06:45 AM

About 5 feet ~ 5 mS. At 33.4 mS per frame that's more like 1/6th of a frame than 2 frames.

I did think about offering a DVD of this performance but....

September 6th, 2005, 02:42 PM	#19
K. Forman Capt. Quirk Join Date: Apr 2002 Location: Middle of the woods in Georgia Posts: 3,596	From what I have found, MiniDV in general tends to drift more with longer captures. Do shorter batch captures, and it should help you out. It did the trick for me. __________________ www.SmokeWagonLeather.us

September 5th, 2005, 11:55 PM	#16
Daniel L Miller New Boot Join Date: Aug 2005 Posts: 21	just to clarify AJ, its not that the audio is getting out of sync with its own video, its that i have two cameras getting out of sync with each other. both are at 29.97, the sample rates are the same for both of them. i dont get messages of dropped frames or anything, its just that as the two tracks play side by side, the get out of sync with one another. if you knew that all along then i apologize, if you perhaps did not, i hope this clarifys a bit. -daniel out

September 6th, 2005, 06:24 AM	#17
A. J. deLange Major Player Join Date: Jun 2004 Location: McLean, VA United States Posts: 749	Daniel, Yes, I did understand that. The thread started to drift when it was suggested that what you were seeing was a clocking problem, which indeed Canon has been accused of having. To sumarize: 1. You cannot expect two cameras to run at the same rate unless they are genlocked (and that's what genlock is for) because their crystals are sensitive to temperature, voltage, age and vibration. This can be ameliorated to some extent by, for example, putting the crystal in an oven inside an oven for extreme temperature stability but we wouldn't want this done in our prosumer cameras because of size, weight, power and especially cost. Given this I think you must live with the situation and the only thing I can suggest is that you try to keep the cameras at the same temperature and resync in post. 2. If Canon does or did have an unusual audio sampling rate FCP is apparently capable of handling it such that audio and video stay in sync WITH A SINGLE CAMERA. Cheers.

September 6th, 2005, 09:05 AM	#18
Daniel L Miller New Boot Join Date: Aug 2005 Posts: 21	awesome, thanks man. i appreciate all the help. i learned allot from this thread. -daniel out

September 8th, 2005, 06:46 AM	#20
A. J. deLange Major Player Join Date: Jun 2004 Location: McLean, VA United States Posts: 749	Maybe this should be in a new thread but it is relevant here as well... Last night I made a 22 minute XL2 DV of me clapping my hands in front of the lens every 5 minutes or so. Simultaneously I recorded sound with a separate stereo microphone through a digital interface using sampling clock derived from the XL2's composite video out. I captured the DV tape using final cut pro, turned the digital recording into an AIFF, imported it into Final Cut Pro and lined it up with the XL2's audio tracks. Both audio tracks stayed in precise alignment with the video and with each other for the full 22 minutes but both seemed to be 2 frames early with respect to the video. In playing back (through FCP) the best alignment of sound and video was obtained when the sounds were delayed 2 frames. Has anyone else seen this? Is there some offset parameter in some preference menu of which I am unaware? Am I finally losing my marbles?

September 10th, 2005, 06:45 AM	#22
A. J. deLange Major Player Join Date: Jun 2004 Location: McLean, VA United States Posts: 749	About 5 feet ~ 5 mS. At 33.4 mS per frame that's more like 1/6th of a frame than 2 frames. I did think about offering a DVD of this performance but....