Lately, I’ve been sitting at my kitchen table and working a lot. It is summer in Ann Arbor and the cherries and berries are coming into season. The birds around the house have been going nuts lately and I love to hear them chattering during the day. I wanted to see if I could build a simple bird call recognition system with Python. The idea is that I would have a script that runs and when it detects audio that might be a bird call it would record it and match it against a data set. Maybe if it got half-way decent I would have it tweet the birds and their songs when they happened. The first step in the process was to try and record chunks of live audio off of my laptop microphone. Once I have the chunks I process them to see if there is a bird sound, and if so what bird it might be coming from. To do the recording, and display the wave form I clobbered together bits and pieces from numpy, pyaudio, and matplotlib. It took all of an hour to get a working prototype. To test it I played some sounds from the Cornell Lab of Ornithology. The Cornell website is a fantastic resource and I hope to scrape the MacCaulay Library page to build my data set.