So, after having done some Fast Fourier Transforms on small scale, on small, sparse datasets, it was time to test something more substantial, an mp3 audiofile.
First I thought I’d build an mp3 parser, but then my pal Micke pointed me to Wavosaur, a very cool freeware app for audio processing, that’s also capable of exporting audio data to various formats. So I downloaded Wavosaur, opened an mp3- file containing beautiful music generated by the BRM-16 engine (if you don’t get excited when you hear that engine, you should not drive a car!), and exported it in a format my FFT-program could consume.
In the graph above, the left side is the output from my program, the right hand side shows the same mp3-file in Wavosaur. A remarkable resemblance…! 🙂 The lower part of my program display shows the transformation from the time domain to the frequency domain.