View Full Version : Advice needed for recording on board a motor launch


Sabyasachi Patra
April 15th, 2012, 01:38 AM
I have started a project filming the mangrove deltas of Sundarbans. A lot of filming is done from a motor launch. I want to record some interviews on board. However, there is the sound of the motor engine. Please advise how to go about recording the interview.

I plan to use a MKH 416 within a rycote zeppelin. Will the motor engine sound be terrible? Will change of any mike for eg shifting to a lapel mike help?

Look forward to your expert advice.

Sabyasachi

Allan Black
April 15th, 2012, 01:59 AM
Take everything and get as far away from the motor as possible. Unless you have 2 lav mics, the 416 with a good boom operator will be best.

If you have passing mangroves in your shot some engine sound will be appropriate, as long as the persons don't have to shout to be heard.

If the noise is a real problem, get up to speed, drop the revs or cut the engine and drift.

Sounds like a nice day out.

Cheers.

Brian P. Reynolds
April 15th, 2012, 02:32 AM
Here is an obscure thing you may wish to try, I have used it on several occasions with good success.

When you record the interview try and keep the back of the mic toward the noise source (in your case motor / exhaust), the second part is record on track 2 a feed from a mic close to the engine to get good clean engine sound or even the camera top mic ... this needs to be recorded on a separate track and DO NOT mix it in with the main mic...keep it separate.

In post production phase reverse the FX mic recorded on track 2 and add a small amount to the interview mic hopefully doing a phase cancellation with your main mic and reducing the problem sound. Give it a try nothing to loose.

As I said I have done this on several occasions, once on a NAVY ship that had bad radar pulses which we phase canceled 90% out. In Helicopters this method also works.
If all else fails and you can only do a single track on location, in post do a duplicate copy of that track ... BOOST the problem frequencies ...then phase reverse.... then mix a bit into the final mix, this will then cancel some of the bad sound in the main dialogue track.

Greg Miller
April 15th, 2012, 11:37 AM
Ditto the "out of phase" cancellation suggestion.

I've done something similar when recording interviews where overhead aircraft noise was a problem. (This was taught me by a real old-timer in the DC area; sorry I don't remember his name.)

The trick was to use two identical lavs, one on the interviewer, and one on the subject. Record them on separate tracks. Then in final mix, contrary to usual procedure, keep both mics (tracks) up all the time. Invert the phase of one track, and adjust the levels so the background noise cancels out as completely as possible. Doing this results in a surprisingly noise-free recording.

Of course on a boat, in theory, you'd want both people to be the same distance from the noise source, insofar as possible, so the phase cancellation will be as complete as possible.

One caveat: keep the two people far enough apart, so that each voice picks up hardly at all on the other person's mic. Otherwise, you could end up with some strange phasing effects since the two mics will always be mixed together.

Sounds like a very interesting project... good luck!

Tom Morrow
April 15th, 2012, 09:39 PM
If you have the budget you might look into a mic with active circuitry which effectively does the phase cancellation for you, like the Sanken CS-3e or the Schoeps (super) CMIT, and aim them so the motor is in the null. Line gradient shotguns like the 416 are more directional at higher frequencies but motor noise has lots of low frequency.

A crazy idea (don't quote me on this, but it might inspire some creativity:) Use a figure 8 mic with the motor in the null, and the two people being interviewed on either side of it. The simpler design of the figure 8 might eliminate phasing comb filtering artifacts that more complicated microphones are susceptible to. Again, this is just brainstorming; I've never tried this. Probably would look strange on video.

Sabyasachi Patra
April 19th, 2012, 08:18 AM
Thanks Allan, Brian, Greg and Tom!

Some good suggestions. I don't have a figure 8 mic. I will try the phase reversal method. I think this will be of help. I will record two tracks to my SD 702 and I think I will also separately record just the engine sound with a Tascam DR 100..

The rainy season is starting and I had a superb experience of a storm. Though people describe it as nature's fury, it generated in me a feeling of awe and reverence for nature. I will do the preparations and go and film in a different weather condition.

Cheers,
Sabyasachi

Greg Miller
April 19th, 2012, 03:00 PM
Recording two mics, the same distance from the engine, on two channels of the same recorder, would be your best bet. If you invert one channel and adjust the levels carefully, the engine noise should pretty nearly null out.

Recording engine noise on a different recorder may not help much. There could be enough difference in sample rates and sample positions (in time), so that you don't get a good cancellation. Also, if your "engine noise" mic is a different type of mic, or in a significantly different position, then the noise waveform you record will be significantly different from the waveform recorded on the "talent" mic, and that will just make cancellation that much more difficult.

And if you use the first technique (reversed phase mics on the talent) then you shouldn't need or want any further noise source... those two identical mics should give you the maximum possible cancellation.

Sabyasachi Patra
April 20th, 2012, 09:42 PM
Thanks Greg!

I will look for two identical lav mics, as I don't have two at the moment. Also, I will try the same track reversed and added to do the phase cancellation.

May be before going there, I should try recording something in a room with a fan on and try reversing to reduce the fan noise.

Thanks for sharing your knowledge.

Greg Miller
April 21st, 2012, 02:46 PM
If you record only one track, then add "the same track reversed" you will completely cancel the engine noise, but you will also completely cancel the voice on the track.

The point is that the noise varies from instant to instant, the waveform never exactly repeats. However, if you pick it up on two [nearly] identical mics, at the same distance, then the noise waveform on the two recorded tracks will be [nearly] identical and if you reverse one of those tracks and add them together you will get [nearly] complete cancellation.

However, track #1 will have voice #1, and NOT voice #2. Track #2 will have voice #2 and NOT voice #1.* So when you reverse the phase of one of those tracks and add them together, the noise will cancel out, but the voice will not. (You will need to carefully adjust levels for maximum cancellation, then leave those relative levels set throughout the mix of that particular take.)

Mathematically:

Tk 1 = V1 + Noise
Tk 2 = V2 + Noise

Tk 1 - Tk 2 = (V1 + Noise) - (V2 + Noise) = V1 + Noise - V2 - Noise = V1 - V2.

The noise will be gone, both voices will be there. The phase of one of the voices will be reversed from the original, but that won't have any effect on intelligibility.


* Note what I said in an earlier post. Each talent's mic must be as close as possible to the talent's mouth. And you need sufficient distance between the two talents -- I'd say 3 feet minimum, more if possible -- so that each mic picks up only one voice, not the voice from the other talent. (Yes, the mic will pick up the "other" voice, but if it's -20 dB or so, relative to the "desired" voice, you should be OK.)

If each mic picks up both voices at nearly the same level, and you subtract them (or add them, for that matter) you will get strange phasing and comb filtering effects which will make the voices sound very un-natural.

So there are three keys here:

1. identical mics on both talents
2. both mics the same distance and location relative to the noise source (as nearly as possible)
3. each mic as close as possible to the talent's mouth, but several feet away from the other talent.

If you don't meet all three conditions, you will have some sort of problems.

Colin McDonald
April 21st, 2012, 03:38 PM
I like Allan's suggestion of switching the engine off. Record a bit of engine sound to add in in post.
Or can you have the boat towed (on a long rope)?

Steve House
April 22nd, 2012, 05:30 AM
I like Allan's suggestion of switching the engine off. Record a bit of engine sound to add in in post.
Or can you have the boat towed (on a long rope)?
Ditto ... or perhaps green screen

Bruce Watson
April 22nd, 2012, 09:14 AM
I plan to use a MKH 416 within a rycote zeppelin. Will the motor engine sound be terrible? Will change of any mike for eg shifting to a lapel mike help?

It's an interview -- does the mic have to be out of the picture? If you can not show the mic you've got good suggestions already.

If you can show the mic, use a good ENG mic (an EV RE-50 omni might not work as well in this case as, say, a Sennheiser MD46 cardiod which will give you a little directionality but won't make it hard for the talent to move around a bit while they talk), and get it as close to the speaker's mouth as you can -- say 10cm or closer. The closer you can get it, the better your signal to noise ratio will be of course. Which is the point, yes?

Another alternative might be something like a Countryman E6 earset mic (http://www.countryman.com/store/product.asp?id=92&catid=10). This will get you really close to the mouth and the speaker can move about without effecting the sound much at all.