audio/video sync issue

[dpalomaki]

It goes back to the accuracy and stability of the clock (crystal osciltator) in the device. Consider how accurate a typcial, economical crystal wristwatch is. Typically within a half second (12-15 frames) or so per day. That is about 1/2 frame per hour. Crystals have accuracy on the order of 10 PPM and can be trimmed and temperature compensated to be much more accurate. 10 PPM is about 1 frame per hour.

Assuming that recorder manufactureres buy similar quality crystals, one could expect similar accuracy.

Sound travels about 30 feet during one frame. A one frame sound sync issue would be roughly the same as moving 30' further (or closer) from the talent.

As to distorting the pitch, I'm not convinced that anyone can hear a difference of a couple frames per hour in an absolute sense (e.g., 1000 Hz vs 1000.02 Hz. They may be able to detect it in a relative sense if doing an A-B comparison or a mix where they can hear the a beat with a one minute period.

Consider the speed accuracy of turntables as another reference point.

[AJL14]

Just shot a live play - 2 hours 45 minutes - with four cameras. 3 were Canons (HF-S21 and XA-10's) and a Sony PJ-650. The HF-S21 was hooked up to the audio mixer from the on-stage mikes, and that track was the "master audio" used in the edit. Believe it or not, the audio had to be slipped only 2 frames by the end of the multi-cam edit during the 2nd Act (1 hr. 20 min, into the play). I was pleasantly surprised by the crystal accuracy of the mixed-breed cameras. Sometimes it's just good luck!
Alan

[Fred Dwyer]

Unlike tape the AVCHD cameras do stay within the frame though so if you are not that critical it would be OK. With timeline scale at 1 sec you would not see the difference but would hear it as a reverb or echo. DV for instance only needs to be within 1/3 of a frame to meet spec. So syncing 2 DV cameras could be 1 2/3 frames out between each other.

Ron Evans

Well, these are AVCHD cameras, and I judge the accuracy of the sound by observing lip sync. Our eyes/ears can perceive very slight variations in spoken words and the lips speaking them. As long as I'm satisfied with the lip sync I observe, it's spot on in my book. Yes there can be slight variations that show up as echo, but I seldom use the audio from more than 1 camera at a time. The Marantz is plugged to the house PA and provides the primary sound track. My HFG10 carries a mixer with a radio mic receiver and a Rode VideoMic. The radio mic is the primary supplement to the Marantz. The on-camera VideoMics on the other two cameras seldom provide any of the audio which we use.
When it comes to the audio recorder, as long as the same clock that was used to control recording, is used to control playback, the result should be accurate in duration to the live event.

[Ron Evans]

..
When it comes to the audio recorder, as long as the same clock that was used to control recording, is used to control playback, the result should be accurate in duration to the live event...

That is the case but if the file is taken from the card in the Marantz and played back in the NLE then the clock is not the same. It now becomes the clock in the PC.

In the past non file based systems the master clock would control everything and they would all stay in sync during recording and linear editing. Now individual pieces create files that say they were recorded at a particular clock rate embedded in the file but that clock rate is different for all the pieces of equipment used. Unlike linear editing where there was a master clock , file based system use the clock in the PC as the master clock for editing and playback. Very slight differences but they are not the same as being genlocked.

I do much the same, use feeds to either cameras or recorders and only use the camera audio as a guide. Audio track normally comes from one camera or the Zoom recorders with a little mixing from the camera for audience presence.

Ron Evans

[Fred Dwyer]

T file based system use the clock in the PC as the master clock for editing and playback. Very slight differences but they are not the same as being genlocked.



Ron Evans

Good point Ron.