Wednesday, March 04, 2009
I have been experimenting with an under clocked SN76489 sound chip. This particular sound chip is the same as is found in a number of video game consoles, most relevantly the Sega Master System. In many ways it is quite a limited chip - for example it has three pitched pulse wave channels, whose duty cycle can't be set. Another limitation is the frequency range, which is actually relatively small and only has 1024 steps.
Normally (on the Sega, for example) there isn't much bass at all (ie. the chip can't go very deep in pitch), which is why I thought I would underclock a discrete SN76489 and see what it sounds like. I underclocked it by just under half - therefore extending the range by about one octave downwards (but of course sacrificing range and resolution in the upper octaves).
I am sure that anyone who listens to chiptune is aware of the phasing sound produced by two square waves that are slightly out of tune with each other. By (very slightly) detuning two of the channels on the SN76489, I was able to make some interesting timbres. The thing to listen to in this example are the changes in the harmonics as well as the changes in timbre. I have set the detune amount by only one data point between the two waves, and as such the phasing effect that you might hear with a more detuned pair of waveforms resolves into discrete timbres with abrupt changes instead of a continuously changing effect - although the difference is "only" a difference in temporal perception, it's actually a large subjective difference, (or that's what I think at least).
Listen to an example here: