Off camera audio solutions for 5D shooting
 

I'm looking for solutions to take my audio acquisition off-camera (shooting with 5Ds). The Zoom H4n has two XLR inputs correct- that means it can be used instread of a full fledged audio board on a shoot. In other words if I'm running two wireless mics I can simply plug them into the Zoom and capture that audio without having to mount them to cameras.

Also is there an audio format that will keep in sync with the 5D's 30fps?

TIA everyone!

You can shoot at any rate that you'd like. Typically, it's 48kHz for film. 44.1 will work equally well. If you're mixing the dialog with music from a CD, which is 44.1kHz, recording at 44.1 and putting it on a 44.1 timeline makes the most sense - no resampling needed in that case.

If you slow the 30p down to 29.97p, then slow the audio by 0.1% as well. You might still have some drift from the start to the end of long clips, but that's due to running the systems open loop, not due to 30p.

Best of luck with the production...

The Sony PCM-d50 is also a good solution for external audio....

How long before drift an issue?
 

I did a test with the Zoom H4 and the 5D. It seemed to me that with out doing anything to the original audio or video it would sync ( close enough ) up until about 7 minutes. That is if you manually sync the audio from the H4 at the beginning of the clip, you wouldn't see a noticeable drift until you hit the 6-7 min mark. That is a rather long one shot clip for most projects. In my line I guess toasting could go on longer but for that I use a UWP receiver and got right to the cam via mini plug to mini plug. the transmitter is coming out of the DJ board. Clean enough. Back up via H4.

I am assuming that most of you who are having the hardest time with this are shooting seminars or long one clip stuff? I think if you have some cut aways to throw in the sync becomes less of an issue, just somewhat time consuming?

Is there a setting on the Zoom you can use to avoid it drifting? What exactly causes it to drift? It's my understanding the the 44.1 vs 48 is simply the sample rate and doesn't affect the "timing" of how it plays out- in other words it's not analogous to frame rate.

I don't quite grasp it all- trying to wrap my head around it.

No. It's like building two grandfather clocks. No matter how closely you try to match them, over time, one will run slower than the other.

Electronic clocks tend to change their pace slightly due to temperature. If you were to heat your 5D and cool your Zoom, you'd get one result. Cool the 5D and heat the Zoom and you'd get another.

Back when I worked at Grass Valley Group, we built a Master Sync Pulse Generator that was used as the master sync for TV stations and studios. It used a very expensive oven controlled crystal as its reference. The crystal was in a metal case in a temperature controlled environment.

Minimizing drift with independent clocks is expensive. Eliminating it requires that you connect the two clock and slave one to the other.

Of course, if you shoot short takes, the drift issue goes away. :)

If using a Zoom H4N with the 5DM2, for e.g. static interviews, is it feasible to get a good quality backup by simultaneously taking the H4N audio out, padding it down, and connecting to mic in socket on the 5DM2?

Anyone done this successfully? Any advice?

Thanks.

IF i'm understanding the current trend (I've been doing my best to struggle and keep up here in the forums re: audio), one solution that might work for you, Glen, is to hang tight for a few more weeks (if possible) to see how Tram's solution works out.
Then you'll take your wireless mics and run them into a preamp and then straight into your camera. This should get us uncompressed hi-quality audio.
Barring that, for a duel system, I believe you have it right: run your mics straight into the H4n (or go through a pre-amp first if you need the boost)(for example: Camcorder XLR Audio Adapter/Preamp: Buy Direct and Save from juicedLink) and just hit Record on each device (camera and audio) for each take (and remember hit stop - which is sometimes a problem for a thin-brained guy like me).

I've heard the Sony PCM-D50 has better sound, so I almost bought it instead of the H4n. I don't think has XLR inputs, so it isn't a solution for me, though.

For me, I'm not hurting for audio for another month or two, so I'm waiting to see how Tram's solution works out before I buy anything external.

It's not the level that's the problem. It's that most inexpensive audio circuits use "auto gain control." AGC is very helpful to people doing non-critical work like videotaping birthday parties and social situations. Essentially, if the circuit sees a quiet audio signal, it boosts it. If it sees a loud signal, it reduces it. The point is to keep all sounds in the same good range.

However. In professional use, AGC is a huge problem. When things go very quiet, the AGC pumps up the gain until you hear the distinctive hiss of the circuitry. An annoying sound that dives back down as soon as the ambient audio level goes back up. Only to start building up again. Nobody cares about that on a home TV watching the kids play in the pool. For professional work, however AGC is unacceptable. The 5d Mk ii has AGC that can't be switched off. And it will affect ALL audio fed into the camera.

So companies and people are trying to find ways to "trick" the AGC into staying off. On the new Beachtek units, I believe they use an inaudible "pilot tone" to make the AGC think there's ALWAYS a robust sound input hitting the mic plug. This may work great. Or it may make audio levels harder to set and ride - nobody knows yet.

Other companies may try different approaches. But the point is that until there's a reliable way to defeat the camera's inherent AGC, Pros will consider on-camera audio on the 5D to be at best a "scratch track" to be replaced in post.

Whatver TRAM or any other manufacturers attempt, it MUST deal with this or the system simply won't produce audio to professional standards.

Just the way it is.

The PCM-D50 has an optional XLR input module... but it isn't cheap...
also as an added benefit, the PCM-D50 has a remote control too...

I just mount the PCM-D50 on the camera hot shoe... works fine and you can
monitor the sound with head phones.

Thanks Bill. I understand about the AGC issue. My question was about getting a 'good quality backup' using the method I referred to. I realise that any audio going into the 5DM2 in its current state will suffer the AGC problem. Just wondered if anyone was using the H4N for quality audio with the 'belt and braces' of a 5DM2 backup.

I've shot a couple of projects where we run through a small mixer first, then from there to both the 5D and H4. It works, but the quality difference between the two is pretty significant so I don't know that it would work well as anything other than an emergency backup.

Just got the H4n and shot my first interviews with it last week (48/16) and find it still drifts pretty quickly - but retiming to 99.9% speed seems to work to keep it synced. I've tried various different combinations of editing timebases and conforming the 5D clips to 29.97, etc, but nothing works as easily or consistently as just retiming the H4n audio.

Thanks Evan.

I was really wondering how users dealt with feeding the headphone output of the H4N to the mic input of the 5DM2. As for comparative quality, maybe with Tramm Hudson's latest firmware upgrade the 5DM2 will approach the quality of the H4N.

Are you saying that the H4N always runs ahead of the 5DM2 or is this the 30 fps/29.97 fps editing issue that people talk about?

Jon Fairhurst is going to get ahold of an H4n to record to compare to the 5D capabilities using Magic Lantern. If you haven't tried it yourself, you are missing something. I am going to go out on a limb and guess that the 5D fed with a Juiced Link or a mixer like my ENG44 will give you better recordings than straight into the H4n. So why would you bother, unless separate sound is an absolute necessity for you.

Thanks for that vote of confidence Chris. I do hope to give Magic Lantern a whirl soon and will report back.