Timecode
 

I can't seem to find a clear answer on this one... Why wouldn't I want to record both picture and sound at 23.98? I'm trying to figure this out for a shoot I'm working on and if someone doesn't mind giving me some direction or clues as to the reasoning I'd appreciate it! Right now my understanding is just that not all equipment is capable of running 23.98 timecode (ie. audio recorder/slate). So, if it's the frequency of the signal that matters most (48/96kHz syncs with 23.98 and 29.97) then I'm not sure exactly what the issues are beyond equipment though I have a hunch they might involve NTSC deliverables. Thanks, Ryan

You're leaving out some major pieces of information that would be necessary to give a meaningful answer. Such as ... camera used, audio recorder used, shooting format (SD or HD), on-set and postproduction workflows, final deliverables required. For example, most of the cameras discussed in these forums (forae?) lack timecode I/'O and so most of the time any frame rate set on the audio recorder becomes moot because the camera and audio aren't talking or jammed to each other anyway, since audio sample positions in the audio file and video frame positions in the video file can't be matched to each other based on identical timecode labels anyway. The frame-rate setting in the audio recorder has no effect whatsoever on recording or playback speedsl - 1 second's worth of sound is recorded and replayed in 1 second, regardless of the frame rate the recorder is set to or the frame rate used for the project in the NLE. Sample rates determine speed, timecode frame rates don't.

Steve...don't mean to hijack this thread, and if this is too off topic, I will repost in a new thread... I have a PD 170 and want to record additional audio with a Zoom H2 and try to sync the H2 audio in post...is this difficult to do seeing that there is no timecode linking between the two.

All timecode does in a DV workflow is establish a single line-up point for the audio and video files, serving the same function as an old-fashioned clapperboard slate. It's a convenience when syncing a large number of takes but that's all, just a convenience. Get a slate or use a handclap or something with each take and there's no need for timecode.

A more serious issue is maintaining sync over the length of a long take. If you get the audio and video perfectly aligned at the start of the clip, you may well find they have drifted apart by the end if the shot lasts for some appreciable length of time, say 5 or 10 minutes. To prevent this you have to have some way of getting the camera's sample clock and the audio recorder's to lock together so they share a common timebase. Timecode doesn't do that. The problem with the combination of gear you mentioned is that manufacturers of inexpensive consumer gear like the Zoom have to save money somewhere and highly accurate clocks cost money (a Lockit Box external sync clock costs 3 or 4 times the entire Zoom recorder, for example). A slight variation from dead-on 44.1 or 48kHz makes no difference at all when strictly recording sound but when it has to sync to video it makes a huge difference. How bad it will be, how long the shot can run before the audio and video have drifted apart to an unacceptable degree, depends on many factors. To prevent it, you need an audio recorder that will accept sync from the camera such as the Tascam HD-P2 or a pro-level recorder with better clocks such as the Sound Devices recorders.

With limited funds and without purchasing more expensive equipment...can the audio and video recorded on something like an H2...that becomes out of sync after 10 or 15 minutes be corrected in post by stretching the video to match the audio. If so, will the "stretching" over an hour long video, be noticable?

Yes it can be time adjusted. Most editors and audio workstation programs these days are able to do to a reasonable degree without changing the pitch.

But let's make sure we are stretching/shrinking the audio, not the video!

Agreeing with Steve here, I've done this many times with event video of musical performances. I use Vegas, which has probably the best abilities to work with audio of any NLE, as it started out as an audio-only application.

One thing that can be helpful is a tail-clap in addition to the clap at the head. In post, slide the audio track to line up the head-clap, then stretch/shrink to line up the tail-clap.

A bit more discreet is to use a clicker. You'll find it for under $2 at a large toy store. It's a kid's noisemaker that has a small piece of thin steel under tension, when you squeeze it it makes a loud click, then another when you let it go.

Here's a wide ranging discussion on numerous issues relating to shooting and post with prosumer gear:
http://www.dvinfo.net/conf/all-thing...-synching.html

And here's a summary of a typical post-production workflow:
http://www.dvinfo.net/conf/all-thing...questions.html

Steve and Seth...Greatly appreciate all the advice and info.