Sound Dampening Idea!

[Sillyname]

How about a medical suction cup microphone like the ones they attach to measure your heartrate, except you attach it to the side of your camera and run the signal out of phase into the mic jack using a coupler so you can plug in your shotgun mic into it also. Then you can use a small inline volume control to vary the amount of camera noise you extract while listening over headphones. This would cutout not only motor noises but handling noise as well!

Chances are, you're going to be using an EQ to extract the noise anyway. Might as well do it in post!

------------------
Deep is not the root word of depression.

[This message has been edited by Sillyname (edited 26 September 2000).]

[CrazyBee]

This process (although I have never heard of it used with a camera) is used for industrail noise control. Like with big-ass fans. It was also looked into for reducing car interior noise.

Be aware though, that, there is a lot of mechanical noise from within the camera that will vibrate a microphone's element. Someting like this can't help that. But, if you disable teh onboard mix then that problem is solved.

You basically have the right idea going. One can use op-amps or other simple devices to adjust gain and phase of a signal. It won't be perfect due to the time difference of the wave being inverted and apllied but it will work somewhat effectively.

Hmmm.. so.. 3 circuits, one for boom mic, one for noise cancellation mic, and last one for mixing the two. I guess you could get away with two, one for noise cancellation and mixing, but, it'd probably be better to have seperate gain controls for each item and then a final mixing circuit fot gain and impedance control to whatever device is recording audio.

Radio Shack..

[Sillyname]

Way ahead of ya'. Got a Telephone Pickup at lunch break. $3.99 + $1.99 for 1/8" mono In-line Phone Jacks. Gonna get an in-line headphone volume adapter for $4.99 tomorrow. It wasn't at the Radio Shack I went to.

Probably use some gaffer tape to tie it down to the side of the camera. Or velcro.

[Patrick]

I don't believe that this Radio Shack hardware is going to be shielded very well. You're liable to pick up all sorts of interesting RF interference. However, for the minimal cost involved, it's certainly worth a shot.

[wkulecz]

Sillyname,

Your idea is sound, but far more complex than you imagine to actually impliment. It is a variant of a class of filters called Kalman Filters which only started to have practical applications when they could be implimented digitally.

The military continues to spend big bucks researching such things to solve the problems of vehicle noise interfering with crew communications.

The same idea has been used in medical settings in an attempt to remove EMG "noise" (the signals from muscular activity) from EEG (brain-wave) recordings.

This ain't simple stuff!

As a quick look at an introductory text like: "Introduction to Random Signal Analysis and Kalman Filtering" by Robert Grover Brown, John Wiley & Sons. 1983, will show.

--wally.


[This message has been edited by wkulecz (edited 26 September 2000).]

[Flying dutchman]

Moreover, do you simply want to invert the signal and add it to the microphone signal?
That won't work (at least not for the higher frequencies).

If you use an external microphone that's, say, 20 cm in front of the camera, the camera noise it picks up is delayed by plusminus 0.6 ms (because of the 20 cm distance...)
You'll have to delay the "disturbing" signal as well before you can cancel it out...

[CrazyBee]

Hmmm.. true... for a real world effective application.. you will only get so-so results.

Maybe easier to have a mic on the camera and have it as a second audio channel and teh boom mic as the first. Then, invert the camera mic sound and sum with the boom mic?
You'll have to adjust for the recording levels.

OR, have a mic on the camera, invert the camera sound, and output it via a small speaker towards the general direction of the boom mic. Won't have to worry to much about the delay.

[Flying dutchman]

You might consider the following experiment:
Make the right-channel mic on your camcorder "deaf" with a drop of wax. It'll only record the internal camera noise. Then record your footage.

Afterwards, using a digital audio editing program, invert the right sound channel and add it to the left channel. You should get a pretty good (albeit mono) signal with less noise than before...

[Phil Pugh]

You may want to leave it to post processing. As mentioned in other posts - the theory is perfect - on practise all sorts of problems will appear - you will knock one the taped on mike - or get a different frequency response.

I have had reasonable success using the 'Remove Noise' process with CoolEdit (Goldwave has this also I believe). I just separate the audio, open Cooledit and choose a 10-15 sec piece of 'quiet' sound as the sample of noise I want to remove. This can be quite effective as most of the noise is well defined (e.g. motor noise) that is suited to this type of processing. I find, however, that I sometimes have to boost the top-end of the resulting audio a little. It doesn't take too long to do this sort of processing and it can help you out in those moments where the background noise is very annoying. Choose a high number of samples in the FFT routine for best results,