FORUMS: list search recent posts

How to capture audio on-camera rejecting the operator's noise

COW Forums : Audio Professionals

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
Adriano Castaldini
How to capture audio on-camera rejecting the operator's noise
on Jun 6, 2017 at 6:32:47 pm

Hi everyone, I'm a newbie here, and I'm honored to ask for your help about a question:
As a filmaking student I havo to make a little documentary with audio 5.1.
The bad thing about audio is that I'm completely alone (I haven't a boom guy).
The good thing is that this documentary will be almost without dialogues (it's a sort of silent and poetic tale about the life of an old woman).
In this scenario I thought to realize the surround audio field completely in post, mixing different sources taken during "clean" off-camera sessions (so avoiding to capture audio during video sessions) and this for 2 reasons:
1. creating an imaginary (and perhaps more interesting) sound landscape;
2. basically avoiding to capture my noises as camera operator (my steps, my breath, steady squeaking) during video takes.
Recreating a fictitious audio in post should work for all the channels BUT one: the C-channel, that contains the most "in-frame" audio effects: infact I can't dub my talent's own steps, movements' noises, sighs, etc. (because she's not an actor, but a real woman living her life). In few words I need at least an on-camera mic solution to capture the C-channel effects.
The problem is: which mic is able to bring only the sounds on its front while completely rejecting the sounds on its rear? If you know a mic like this, please tell me! Because up to now all the polar diagrams I've seen show a certain capsule's sensibility also under the center:
a) omni is obviously unusable for my purpose;
b) cardiod has two curves under the center, and tends to behave like omni at low frequencies;
c) hypercardioid have a little ball of sensibility on its rear;
d) shotgun is narrow but creates a sort of 8 figure in its diagram, i.e. it captures also on the back.
Surely shotgun is fine as boom mic, or even as on-camera mic if the operator is moveless and silent, but in a run&gun scenario I can't think it's the right mean to capture a clean audio. And more than this, it's too directional for a moving talent in a on-camera condition.
Someone uses all-in-one solutions like the fantastic Sony PCM-D100 mounted on the camera: I've thought about it, but I've heard tests with evident camera-operator's noises (and all this recorders have internal little capsules that are unidirectioal only at 1kHz and up, but under 1kHz the capsules behave like omni).
So does exist an on-camera mic solution to capture front-only?
I have an hypothesis that I'd need to discuss with you:
Let say I'd start with an hypercardioid because it should work indoor (without too much reverb) and basically also outdoor. I'd choose a mic with a pretty constant polar diagram for all the frequency spectrum (Audio-Tachnica AT4053B, or Audix SCX1-HC, or a modified Oktava MK012) and I'd use two in X/Y position on the camera (so I should choose a mic without handling issues). In this way I'll surely capture front + side + rear sounds, it's true, but there should be only a single common cross-diagrams portion, and that portion should be only in front of the camera, not in the back nor on the side. This "clean" portion should be obtained by algebrical operation on the two channels (in a DAW), and it should be my C-channel.
Do you think it could work?
I'd use a Tascam DR-70D (screwed under the camera) modified by Busman to have very clean preamp, and a couple of AT4053B or SCX1hc mounted over the camera in X/Y position. I've read that modified Oktavas are very interesting but I've also read that they are too sensitive to operator's movements. So, between AT4053B and SCX1hc, which is in your opinion the less problematic in a run&gun scenario?

Thanks for all your help (I really need it) and your advices.


Return to posts index

Richard Crowley
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 7, 2017 at 1:47:06 pm

[Adriano Castaldini] "I can't dub my talent's own steps, movements' noises, sighs, etc. (because she's not an actor, but a real woman living her life)."
Please explain why you think you "can't" use traditional Foley techniques? That is the most incomprehensible part of your question. (At least to me.)

You have eloquently described WHY people do not use camera-mounted microphones for this kind of recording. The biggest problem is that on-camera is not a usable location for the microphone. Even after over 100 years of research, development and incremental improvement, we have never been able to devise a magic microphone that acts like a tele-photo lens. It remains a truism: you simply cannot violate the laws of acoustic physics.

This is a very valuable learning scenario. You will learn WHY professional-quality productions are not made with camera-mounted microphones.

If I were forced at gun-point to make this happen, I would devise some sort of very light-weight "mini-boom" using a thin fiberglass shaft (like a fishing pole, etc.) and a very lightweight, but technically advanced) microphone like one of the Rode "Video-Mic" models. The biggest issue is that you MUST do two things:
1) Get the microphone AWAY from the camera. For the most obvious reasons you have described.
2) Get the microphone CLOSER to the source of the sound you are trying to capture. If you are trying to capture footsteps, etc, then stick the boom off the bottom of the camera at an angle as close to lens-axis as you can get away with at the max zoom angle you will be using.

---------------------------------------------------------------------------------
Recording audio without metering and monitoring is exactly like framing and focusing without looking at the viewfinder.


Return to posts index

Peter Groom
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 8, 2017 at 8:38:14 am

Hi
I agree (as usual) with Richards comments.
Often the world just doesn't work the way you'd like it to, which can be frustrating, but usually for good reason.
Peter

Post Production Dubbing Mixer


Return to posts index


Adriano Castaldini
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 9, 2017 at 8:55:59 am

Thank you all for your advices. I'm very honored of receiving your replies.
I'll take a time to think about what to buy.

(I'll do some tests with the poor means I own at the moment. If I'll obtain something interesting, I'll surely post my impressions here.)

Anyway, someone of you have experience about booming or handling AT4053B vs SCX1hc vs modified MK012? Anyone can say which is the less problematic in moving situation?

Really thanks as always.


Return to posts index

Adriano Castaldini
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 9, 2017 at 9:38:20 am

And (perhaps this could seem a bit topic-off) let try to forget the camera: what mic do you recommend for ambient noises recording? Infact, as I described the kind of sounds I'd like to capture from the old lady (sighs, steps, almost no words, etc.) these could be considered like ambient sounds, right? So, if I decide to capture all AWAY from the camera, the hyper (like AT4053B or similar) could be still a good choice? I've read a lot of people preferring something like DPA4060 (little omni) for field recording, but I have the suspect that perhaps - for the kind of sounds I'd need to capture - something like the AT4053B could be more flexible. (If possible I'd like something able to capture little specific sounds - as I said sighs, etc. - but also a field situation, is it possible? I've read for example that someone does all with the Sony PCM D100: both field and near sounds). So what is your advices? An hypercardioid could be the flexible solution for all the days? Or the Sony? Or something else?

Really thanks as always.


Return to posts index

Richard Crowley
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 9, 2017 at 6:41:33 pm

IMHO, the AudioTechnica AT4053B or the Audix SCX1-HC would be good ambient-recording microphones. And good "indoor" boom mics as well. A lower-cost option would be something from Røde.

---------------------------------------------------------------------------------
Recording audio without metering and monitoring is exactly like framing and focusing without looking at the viewfinder.


Return to posts index


Bruce Watson
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 12, 2017 at 4:38:24 pm

[Adriano Castaldini] "Infact, as I described the kind of sounds I'd like to capture from the old lady (sighs, steps, almost no words, etc.) these could be considered like ambient sounds, right?"

First, let me just say that this is an excellent assignment! However this works out, I guarantee it will be an interesting learning experience for you. Enjoy it.

The reason this stuff is done as Foley in post is because a single mic can't be in the right position to capture "sighs" and "steps" at the same time.

With a microphone, it's all about positioning -- this is your main control of signal-to-noise ratio. To capture her voice you need to have to mic above and in front of her head, pointing at her mouth at about a 45 degree angle, all while still out of the frame. And you have to track her movements. Which is why it takes a boom op. Clearly, if you then want to use this same mic to capture shoe sounds, it's way out of position.

What you'd get from shoe sounds from that far out of position won't match well with the voice sounds even if you could capture it. That's because you're getting different amounts of room sound, and timing, from the room. That is, the reflections off the walls, ceiling, floor, furnishings won't sound like it's a recording from 50cm from the source (like the mouth is recorded). Instead all those aural queues provided by the reflections will sound like a recording from 2m away from the source. Because that's exactly what it will be that you're recording.

Nearly all sound that you hear in a movie, be it narrative or documentary, is provided in post. Typically the only sound recorded "on set" is dialog. That's tough enough as it is, no need to make it harder. So all other sounds (running water at a sink, setting down a teacup on a table, flicking on a light switch, newspaper rustle, footsteps, turning a door knob, every frellin' thing, is Foley or sound effects.

My advice, such as it is, is try to capture whatever comes from her mouth and add the rest in in post. Record Foley as dry as you can, build up your dry soundscape (getting the relative levels correct), then polish it (appropriate compression, and a reverb to match the quality of the room (most reverb plugins in DAWs have a "small room" built in)).

And since you can't boom (no boom op), put a lav on her. An omni lav. This will consistently capture her voice, and be close enough to capture even soft sighs. You can hide the lav under her clothes if you need too, but discuss this first, and have a (female) assistant on hand to do the work if it makes the (female) subject more comfortable.

Also, capture a fair amount of room tone (a few minutes of the room with nothing else) to use when you edit. Once you place a sound effect like setting down a glass on a table, you'll need to cut out what the lav picked up, and replace that with room tone. So you end up with the same room tone, plus the sound effect stacked -- so it will sound like it happened in the real room. You may need to add reverb to the individual effect to match with the room tone, which is yet another way to do things.

Have fun with this -- it really is an excellent learning opportunity.


Return to posts index

Richard Crowley
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 9, 2017 at 6:21:51 pm

[Adriano Castaldini] "Anyway, someone of you have experience about booming or handling AT4053B vs SCX1hc vs modified MK012? Anyone can say which is the less problematic in moving situation?"

What does "modified MK012" mean? The Oktava MK012 has a reputation of having above-average handling noise AND above-average self-noise. Its main feature is its budget price. But if you can properly shock-mount it AND you have ABOVE-average, very quiet microphone inputs, it is useful.

The AT4053B is a known quantity and favored by many. But many people also like the Audix SCX1hc

---------------------------------------------------------------------------------
Recording audio without metering and monitoring is exactly like framing and focusing without looking at the viewfinder.


Return to posts index

Adriano Castaldini
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 12, 2017 at 12:43:13 am

Mr. Crowley, thanks for your advice. After a lot of thinking I realized that you are absolutely right: the best solution for me is capturing sounds during an off-camera sessions (like a field/foley recordist) to dub the video takes in post.
Now, as I described, the sounds I'll capture are "little" sounds (sighs, footsteps, rubbing clothes, etc.) and also a bit of outdoor field sounds.
My doubt is between two options:
1. the all-in-one pocket Sony D100 with its built-in mics;
2. MixPre3 + AT4053b.
Which is the lower-noise-floor winner? (Or anyway, which is the winner for my purpose, in your opinion?)

Thanks a lot.


Return to posts index


Richard Crowley
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 12, 2017 at 5:01:23 am

I think everyone would agree that the Audio Technica AT4053B and SoundDevices MixPre3 would have a much more reliably better signal-to-noise ratio than any single-piece recorder with built-in mics. Not even a horse-race.

---------------------------------------------------------------------------------
Recording audio without metering and monitoring is exactly like framing and focusing without looking at the viewfinder.


Return to posts index

Adriano Castaldini
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 12, 2017 at 3:15:55 pm

Well, my doubt came from specs:
- Sony D100 has a noise-floor of -127dBu (even if only at rec-level 5) and a S/N ratio of 100dB (that's impressive);
- MixPre3 has a noise-floor of -130dBu (great), but AT4053b has a S/N ratio of only 78dB (and it has its own self-noise of 16dB).
So I have 2 main questions (as a newbie naturally):
1. When does the difference between S/N 100dB (D100) vs 78dB (AT) become crucial? (Is it a secondary detail, or is it important in "little" sounds recordings?)
2. Considering that I'd use the MixPre3 with an AT4053b (not with a super-silent-schoeps! but a mic with its own 16dB noise-floor), so the question: is the difference between the two noise-floor levels (D100 vs MixPre3+AT4053b) so crucial for "tiny" sounds recording?

Going further, D100 is stereo, while AT is not, obviously. I'm interested into obtaining a stereo image only outdoor for field recordings. Remaining in the price range of the AT4053b, the only "cheap" options that fit a blimp are:
- M/S option: AT4053b + AKG CK94 (a figure-8 with a self-noise of 22dB ☹ and a S/N ratio of only 72);
- X/Y single mic option: BP4025 stereo mic (self-noise 14dB, S/N ratio 80dB);
or...
- again Sony D100.
On the sheet it seems that the BP4025 option beats the AT4053b+CK94, but again BP4025 has its own noise-floor that is added to the noise-floor of the MixPre3. So here the last 2 questions:
1. Are you sure that the sum of BP4025+MixPre3 noise-floors beats the noise floor of the Sony D100?
2. If yes, is this a "day vs night" difference for field recording?

I know that many of my questions are silly for an expert, but your answers are fundamentals for me.
So thanks a lot.


Return to posts index

Richard Crowley
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 12, 2017 at 5:03:19 pm

[Adriano Castaldini] "1. When does the difference between S/N 100dB (D100) vs 78dB (AT) become crucial? (Is it a secondary detail, or is it important in "little" sounds recordings?)"
I'm afraid that what you are seeing is marketing types playing with the specifications for marketing purposes. Even Sony is apparently not immune from this (IMHO) dishonorable practice.
[Adriano Castaldini] "2. Considering that I'd use the MixPre3 with an AT4053b (not with a super-silent-schoeps! but a mic with its own 16dB noise-floor), so the question: is the difference between the two noise-floor levels (D100 vs MixPre3+AT4053b) so crucial for "tiny" sounds recording?"
You are correct that all those factors are important for quality recording of quiet sounds. However a proper "SDC" (small-diameter condenser) microphone, and especially a premium product like the AT4053B, will have a inherently lower noise-floor than any tiny electret condenser mic capsule as used in the Sony recorder. Also, no matter how good the microphones in the Sony recorder are, a separate microphone is ALWAYS more versatile than any built-in microphone(s). Not to mention that you have you choice of specific microphones for specific purposes vs. using the one built-in to the recorder.
[Adriano Castaldini] "1. Are you sure that the sum of BP4025+MixPre3 noise-floors beats the noise floor of the Sony D100?
2. If yes, is this a "day vs night" difference for field recording?"

I would wager that a truly professional microphone like the AT4053B along with a premium mixer like anything from Sound Devices will beat the socks off any little consumer handy-recorder. Despite what the advertising specifications seem to say. If I had my choice between an AT4053B/Sound Devices mixer and a little consumer voice recorder, I could make that decision is a heartbeat. Of course the AT/SD combination costs considerably more $$$$, so that should come as no great surprise.

---------------------------------------------------------------------------------
Recording audio without metering and monitoring is exactly like framing and focusing without looking at the viewfinder.


Return to posts index


Adriano Castaldini
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 12, 2017 at 5:41:49 pm

More $$$$, that's unfortunally right...
AT4053b + BP4025 + MixPre3 + Rycote Blimp (for BP4025) = 3x Sony D100 + Rycote kit. More or less 2400€ vs 800€...
Everyone cheering for the skinny David against Goliath (at least for the money) but in real life size matters.

Thank all of you for your precious advices! This forum is my school ☺ Really!


Return to posts index

Richard Crowley
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 12, 2017 at 5:53:42 pm

Few recordings of sound effects, ambient "nats", or Foley are actually done in stereo.
The stereo (or 5.x or whatever) "soundstage" is created in the post-production editing process.
The final sound track is a synthesis of a bunch of different sounds.
If each of them carried their own stereo "ambiance" you would end up with a jumbled mess of sound when you mix it together. That is why we try to capture each sound as isolated and pristine as possible, so that any ambiance, stereo-staging, etc is produced in a controlled environment of the post-production mix-down.

To be sure, SOME sounds like the general background ambiance of a location are appropriate to record in stereo so that it sounds like you are actually there. But individual sounds like footsteps, teacups, newspaper, door knob, and even any dialog, are usually recorded on monaural. Once you start mixing down the sound track, you will understand why.

---------------------------------------------------------------------------------
Recording audio without metering and monitoring is exactly like framing and focusing without looking at the viewfinder.


Return to posts index

Peter Groom
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 14, 2017 at 1:09:12 pm

Been following this thread interestedly, but I find it interesting how the discussion has become soooooo high end detailed in the finer points of mic choices etc etc, when the starting point was such a low one about gun and run onboard mics?
Come a long way.
Peter

Post Production Dubbing Mixer


Return to posts index


Adriano Castaldini
Re: How to capture audio on-camera rejecting the operator's noise
on Jun 15, 2017 at 12:18:47 am

Well, your merit guys! As I said, this forum is a precious school for me :)


Return to posts index

Adriano Castaldini
Re: How to capture audio on-camera rejecting the operator's noise
on Aug 11, 2017 at 1:26:31 am

Gentlemen, I'm afraid of having damaged a mic...
I've bought the Audio-Techina BP4025 mic that's a stereo X/Y mic with a 5pin-to-2x3pin cable. I connected the mic to the stereo-recorder with the two 3pin-XLR (Left and Right channels) and I gave 48v phantom to BOTH the channels (so the mic itself received 2x 48v). Later I discovered that giving the 48v just to one single channel (i.e. left) was enough to power both the channels as well. So arrived my doubt: Is it possible that I have damaged the mic giving 48v to BOTH channel? Or not?
Thanks a lot.


Return to posts index

Adriano Castaldini
Re: How to capture audio on-camera rejecting the operator's noise
on Aug 11, 2017 at 2:41:37 am

Perhaps it's better to move this last question into a new dedicated topic...


Return to posts index

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
© 2017 CreativeCOW.net All Rights Reserved
[TOP]