recognise sound sample

cv_

Hey everyone!

I have yet a new project in hands that I was wondering on how to approach and thought some of you could be of help. I am building a simple communication system between machines, the idea is that each device will communicate with other over sound using samples I provided. I do not want to use FM or AM as the idea is to create a romantic network of communication.

the tricky bit has been on recognising this samples. I thought I would create a "fingerprint" for each sample, a simple file stating main attributes that identify the sound. This is proving to be trickier than I initially thought, as the samples change over time (I should have seen that coming).

So my problems really would be on how is the receiver timing each sample individually (knowing when it starts and finishes), having that I could make an average of attributes. Or maybe I am approaching the whole thing wrong,

Any ideas?? Thank you for your help

(PS: at the moment I am using Minim to handle sound input and AudioSample to play the files)