Me Spectrumsta!

Again, a spectrum? Yes, why not?! This time it’s done (as to my knowledge, which just grew exponentially today) correct, using a DFT. The previous spectrum used a continous Fourier Transform which kept repeating frequencies! If you test it for frequencies around 16KHz it’d show two bars in two places. The worst thing was that I thought I could be fixed using the normal Fourier Transform(my knowledge of FT is entirely, ad-hoc!). Then I looked at DFT on Wikipedia, realizing that it couldn’t be easily fixed as I thought. And there I see the images illustrating the EXACT same thing! Yay. (Re)implementation time! I redid the spectrum using DFT, and it is still not FFT. Thus, why I didn’t use the word before, it’d be misleading if I use it to describe this project.

This new version has a good frequency response, for frequencies which can be heard 20Hz-20KHz.

So, “Ba Dum Tiss”, here’s the 20Hz to 20KHz “sweep” and some random parts of Paul Van Dyk feat. Michelle Leonard – Lost In Berlin.

And the source.

References,

http://en.wikipedia.org/wiki/Discrete_Fourier_transform

http://stackoverflow.com/questions/9645983/fft-applying-window-on-pcm-data

http://stackoverflow.com/questions/7674877/how-to-get-frequency-from-fft-result

Rabin Karp and Boyer Moore implemented for byte data

This is a another assignment from the University. I’m posting this here mostly because there’s not much examples or proofs of these methods in the internet that’s understandable(I have to agree that the Boyer Moore explaination in the report must need some work to be done by the reader, but that’s because we were limited in page count for explainations).

This implementation features string matching on any data given, its alphabet size is 256(unlike other implementations found on the internet).

Download PDF

Download Source