noise reduction help

[efpef]

Hi,
I'm a research student using Audacity for sound files (speech) editing.
I recorded a set of sound files and now I want to get rid of some background noise.
Here are some questions:

1) The noise reduction option allows treatment of individual files. I have many files to treat and I was wondering whether there's any way to automate this process?
2) Can I pick and choose features, like the frequency range I'd like to get rid of?
3) I have 2 mics - a digital headset and a high-quality digital mic with a stand. The mic with the stand picks a wider range of frequencies and sounds much better, but it also picks up the echo from the room I record in (although the room is designed for recording). How do you cancel the echo?

Thanks!
Efrat

[steve]

I have many files to treat and I was wondering whether there's any way to automate this process?

You can do basic batch processing using "Chains" (there are many limitations to what batch processes are currently supported)
http://manual.audacityteam.org/index.ph ... dit_Chains

When using Noise Removal in a chain the noise profile that is used is the last noise profile that was captured prior to using the chain. This means that you would need to manually capture a noise profile that is as typical as possible for the files that you are going to process.

2) Can I pick and choose features, like the frequency range I'd like to get rid of?

Not directly, though you can process the noise sample before you create the profile. For example, if you apply a high-pass filter to the noise before you create the noise profile, then Noise Removal will (mostly) only affect high frequencies.

How do you cancel the echo?

That's extremely difficult to do because (by definition) the echo is very similar to the sound that you want to keep (just quieter).
The only program that I know of that can do echo cancellation is the "Deverberator" effect in "PostFish" http://www.hydrogenaudio.org/forums/lof ... 22051.html (though this is an unfinished, but usable, experimental program).

The better solution is to avoid recording the echo. "Cardioid" (directional) microphones will tend to pick up less "echo" than "Omni-directional" microphones.
Speaking very close to the microphone (with a "Pop Shield" to avoid blowing on the microphone) will also reduce the proportion of "reverb" (echo).

"Microphone Shields (Mic Screens)" can also substantially reduce the amount of reverb picked up. These are quite expensive but cheaper than applying acoustic treatment to the entire room - a cheaper alternative is to make a sound absorbing barrier with coat stands with thick winter coats or heavy blankets hanging on them behind and around the microphone.

[efpef]

thanks for your prompt and detailed reply!
i'll try applying what you suggested.

cheers,
efrat