I need a way to specify a certain frequency (eg. 30hz) and have an algorithm scan the contents of an audio wave, and only show where the ~30hz frequencies are.

I've tried averaging the data, but I've seen some strange results. I'm assuming I need sin/cos/tan to achieve this, but how?

I've searched just about everywhere in google, and no luck. If someone could please explain in layman terms how it can be done (preferebly in pesudo).

Here is an example of an audio program equalizing a wave to only show the lower bands (bass/~30hz):

http://img225.imageshack.us/img225/156/equgt0.jpg

Also, while we're here, if we assume we are dealing with 44,100hz wave files, how do we know the length of a single frequency?

Thanks.
Here is a picture for the second question relating to how to get the frequency based on length:
http://img297.imageshack.us/img297/7683/waveri9.jpg

__________________
Yahoo! Answers 4ever