Audio Fed To Camera...Q.C..?

[Jesse Korosi]

I have a question about audio that is fed to camera.

If I have had audio fed to camera, how can I quality control that the sound is good enough to be used as a master, versus spending time syncing my master audio i was given on disc from set?

Or can/should the fed audio ever be used as your master in terms of its quality?

Thanks in advance!

Jesse Korosi

 

[Alexander Ibrahim]

RED Audio certainly can be used as your master... but there are limitations.

RED records 24 bit 48KHz sound.

Most on set audio engineers record 24 bit 96 or 192 KHz. (I think 192KHz is overkill, but there is some value there if there is to be extensive post.)

For a clean recording there really isn't much difference.

For any recordings with problems, the higher frequency recordings allow more headroom for the audio guys to work their magic. Even a cheap audio recorder can provide much more data with which to massage things in post than a RED can.

Here is the thing- almost every shoot is going to have problematic audio. On some shoots most of the audio will be problematic. Even with all the best people and gear on set you may still have to do extensive ADR.

On Star Trek Phase 2 we often record audio only in prosumer cameras. Of the five episodes I've worked on, three were recorded on an HVX200, one was recorded on a mix of HVX200 and EX1. On Kitumba we recorded right into the RED, though I haven't seen any of that posted with production sound yet.

I won't say the audio is award winning, but no audience member ever complains about our finished audio quality.

So, the point is that you can definitely record usably clean audio right into the RED, and even lesser cameras.

Back to your original question, how to control audio quality if you are recording just into the RED.

First off what gear do you need?

Well, first off get a mixer. You need to be able to output a 1Khz reference tone, and you should have a compressor/limiter on the mixer. Headphone outputs, XLR outputs and phantom power. You need 3 inputs, but I always demand 4 inputs. All of this is the minimum standard.

The cheapest unit I'd point you at is the Azden FMX-42, which I own, for about $500.

http://www.bhphotovideo.com/c/product/493791-REG/Azden_FMX_42_FMX_42_Microphone_Field_Mixer.html

The Azden is the cheapest 4 mic mixer I can find that meets my minimum requirements stated above. The audio processing in this unit is already better than that in most cameras. The build quality and other features leave something to be desired.

My favorite mixers are Sound Devices units. My intention is to move all my productions to a Sound DEvices 788T mixer/digital recorder.

Here is their 5 input mixer, which now records the stereo mix on SD cards:
http://www.bhphotovideo.com/c/produ...es_552_552_Portable_5_Channel_Production.html

and here is the 788T

http://www.bhphotovideo.com/c/product/555771-REG/Sound_Devices_788T_788T_8_Channel_Portable.html

The 788T gives you the ability to record up to 8 separate mic tracks

Then you need someone who is as good at audio as your camera operator is at video, then hand them the stuff and some help to run boom poles, then let them do their job. Getting your money's worth out of a mixer is a full time and attention task.

Good Luck!

 

[Jesse Korosi]

Wow,
Thanks for the detailed response Alexandre!

I am not going to be doing any sound work on set any time soon, but I am very often given 24 bit, 48 KHZ audio that was recorded for a commercial. (I am converting files for dailies.) Now if the Rocket can indeed provide the specs I need for the editor, which is about 75% of the time, this is really quick compared to re-exporting FCP QTs in FCP if I am not running the same version of software the editor is. Or even syncing in Avid can add a lot of time to my nights.

I guess I have always been hesitant to keep the fed audio as my new master not understanding how compressed it may be. I will take a look into better understanding the gear you mentioned above so that my technical questions to the sound guys on set can go up a notch if there is a problem with what i am hearing through my proxies.

Thanks again!

 

[Tom Visser]

99.9% of production audio is recorded at 24/48 KHz, unless someone is using an older recorder that can on only do 16/48. If it's not 24/48, then it's probably 24/48.048 KHz.

The audio on board the RED can be excellent if calibrated correctly. The audio from something like a 788t will be a little bit better, but for simple dialogue, I probably wouldn't equate much of a qualitative difference when done correctly.

The main issue is that if your using an experienced sound mixer, that mixer may or may not be that familiar with the camera and things such as different revisions of audio boards, different adapter cables, different firmware versions, and other unexplained phenomena, have not exactly give RED the best reputation amongst the sound community. I probably work with RED cameras about 90% of the time, and usually the same couple of bodies, so I've become a little bit more comfortable with them and am pretty confident of the sound quality and its limitations.

Here are the differences...

1) Sound mixer is more familiar with his equipment than yours, he is likely to get better results with proven tools. If you prescribe to the notion that it is the experience of the operator and not the equipment that gets results, this makes logical sense.

2) Camera operator, AC, DP, Director, or whoever is monitoring audio, may either have no clue what to listen for, don't have an ear, or have some other things, such as picture, to worry about and will not do as thorough of a job monitoring audio as a sound mixer would. Let the guy with the audio responsibilities monitor the audio and make the call. If you are going to record on camera, then the mixer will need to have some facility to get a return audio feed from the camera back to his kit so he can get a confidence monitor. I tend to like to go wireless these days, so that means some IFB capable transmitter or transceiver. Allocate some $$$ for this is this is the way you want to do it.

3) ISO tracks. Editorial generally doesn't like to deal with ISO tracks, and for dailies, absolutely never. Need either a mono or stereo mix track for the camera. If there are ISO tracks on there too, they are generally ignored, and track count is of course more limited on the camera than the audio recorder. Nagra VI can do 6 tracks, Sound Devices 788t 12 tracks (only 8 inputs), Zaxcom Fusion / Deva can also do 12+ depending on the exact model / configuration. My recommendation for most small budget narratives that at least have some post assets, ISO+MIX delivered via DVD from the audio recorder. Mono or stereo mix only to the camera for dailies and to assist editorial in synching, in the event there is a timecode anomaly or emergency. Quality of the scratch track is not so important, so if you can't afford the best quality wireless hops, it's not such a big deal, simple Sennheiser or Comtek set will work fine.

Obviously if you don't have the budget to hire a mixer, the sound equipment, or even Gucci microphones, don't worry, this stuff doesn't apply to you. Buy the NTG-2 or NTG-3 and be done with it. Hope the PA you have holding the "stick" doesn't get in your way too much. Go direct to camera and just focus on getting the best sound you can from your 1 boom mic. No ISO or mix to worry about, it's all the same. A little bit of technique from the boom operator or hopeful will give you a million % difference in performance compared to obsessing about external mixers / recorders, or sample rates.

 

[jimhare]

That was excellent advice Alexander, great post.

I'm tired of reading "the only way to be taken serious is to record separately."

You put it all out and clearly described in what scenarios higher bit-rate recording can get you out of a bind, while also outlining that if you can't afford the post anyway, the additional frequencies aren't likely to help you anyway!

Thanks for posting.