Detecting a clap in a noisy environment

B

Bumbleboom

New member
Hi there

I'm looking for some advice on a microphone that will be capable of providing a signal clear enough to identify a single hand clap or similar sound in a non-silent environment.
I've made a little preamp for a small electret microphone and this works fairly well when there is no noise but the clap is often lost in the signal when there is some noise in the room. The microphone will be within 1m of the clap.
Its really only the voltage out that I'm interested in and not any other quality to the signal other than the clap.
I've thought about directional microphones but tips set-ups and particular components would be great.

Cheers,
J
 
What you are looking for is a shotgun mic. I dont have any model numbers off ths top of my head, but that's what you need to google.
 
Thanks for that...
I've used a component electret microphone because I thought I might get the fastest signal out of the cirtcuit if there was no additional operation. The microphone signal is to be used a timing event relative to other signals generated from the so the temporal accuracy is critical to ms.
I expect this and the lack of a physical filter is one of the reasons for my noisy signal. The program I'm using is a generic data logger that takes in several signals via a path box and an A to D box.
Does this change anything - i.e. are pre-constructed shotgun or omni microphones best or will building particular components be more appropriate?

Cheer
 
A shotgun mic will do this as long as it is pointed at the clap. An omni will pick up everything from all sides, so the clap will possibly get lost in all the rest of the noise.

A shotgun mic is the kind used on a boom when people are trying to record dialog and such in a noisy environment.
 
A noise gate can help keep the ambient noise out of the detection circuit.

Also, do you have a high pass/low cut filter to eliminate the rumble that can add up, making the noise floor closer to the peak of a clap.

It's also nice to roll of some of the highs, too, to cut down on the noise.

Amplifying the signal after the noise gate will help exaggerate the peak also.

I did something like this in Max/MSP for a heart beat where I amplified the signal by three orders so I could clearly see the peaks. In this case, it wasn't really an audio signal anymore, so much as data that I could manipulate - like an amp with infinite headroom.

So I guess the closest thing here might be to use the least gain possible so the clap becomes one of the only things to register meaningfully.

Good luck.
 
You must log in or register to reply here.
Back
Top