Sound Recordings

Matlab software for making sound recordings is on the middle PC along the far wall in Clapp 420. You will need a microphone (available from Sue Barry or Mark Peterson). Plug it into the pink microphone jack on the back of the computer (be VERY careful if you have to move the computer at all). Start Matlab from the icon on the desktop. Then give the command

soundcard

Within a minute or so you should see a graphical interface with buttons allowing you to record sound, play it back, graph the signal, find its power spectrum, and save it. You can also print from the figure window.

Fourier analysis of EEG signals

Record EEG signals by appointment with Sue Barry. The graphical interface connecting the computer to the electronics doesn't have built in Fourier analysis routines, so you will have to do this in Matlab yourself. Say you have a recorded signal 'data1'. You Fourier analyze it (i.e., find the complex Fourier coefficients) with

ftdata1=fft(data1);

The routine 'fft' is the Fast Fourier Transform. The resulting complex coefficients are stored in 'ftdata1'. I chose that name just to make it easy to remember what those numbers are. Now to see the power spectrum use

plot(real(ftdata1.*conj(ftdata1)))

This shows the modulus squared of each Fourier coefficient. To interpret it, you have to know what the fundamental frequency is (the one called frequency '1' here). Since Fourier analysis assumes the signal is periodic, the time represented by the whole signal is the period T. The fundamental frequency is the reciprocal of this. What that is in Hz depends on the settings you used when you recorded the trace, i.e., how much total time the signal represents.