Need help with FFT
I'm not a mathematician, DSP programmer, or scientist. I am a logic guy; a bit twiddler.
I have some accelerometer data that I've acquired from the iPhone at 10, 40, and 100 Hertz. I need to find the fundamental frequency (I think that is the correct term). I've written an app using SM2DGraphView to help me visualize the data.
Now I'm working on combining the axis of two or three lines of data at a time (normalizing vectors?). When I'm finished I need to run it through an FFT.
My question is, how best to proceed on this last step? I hear that Apple has a framework called Acceleration and this framework has an FFT integrated into it, which I can use. I've also heard talk of OpenCL. Not sure I want to go that far, I'm just trying to determine if I can get a reasonable answer using the data gathering and analysis techniques I've already come up with.
Anyway, can someone point me to a sample, example, or tutorial on using Apple's FFT function? I'm also having trouble finding the documentation for it.
I have some peripheral experience with FFTW but I really don't want to take the time to come up to speed with this package, integrate it into the Mac desktop and then have to integrate it into the iPhone. That is, after all, where all of my effort is going to end up, when my desktop experiments are completed.
I appreciate any help you can give me on this.
-Michael
Python - numpy
Your Mac also comes with the programming language Python installed which can easily give you a fft-result. Check this site for example:
http://bytes.com/topic/python/answers/557852-computing-fft-python-numpy-1-0-a
and also read the NumPy specification for the fft (google numpy fft, or something).
Just remember that you won't be able to see any frequencies above half of the sampling frequency.
// David Lindegren