posted 23 January 2002 19:14            

voice_mod_fm_drone.kym

This sound is totally bizarre but interesting in its own right. The keyboard is used to play the LPF at the output of the FM synth. The Keynumber and LFO and filter envelope is used to open up or close down the LPF. But the keynumber has no bearing on the tone produced by the oscillators. Playing high keys opens up the LPF and playing low keys closes it down.

The frequency of oscillators is set by computing the arctangent of the ratio of energy in the FFrq bandpass filter, to the total energy through 3 parallel filters including this one. The CMRatio (carrier to modulation frequency ratio) also adds spice. When this value is greater than 1 it produces metallic timbres. When less than 1 it sounds a bit more behaved.

The modulation index (MI) of the FM synth is controlled by the arctangent of the ratio of the FMI bandpass filter energy to the total energy through 3 parallel filters including this one and FFrq BPF.

The third filter is simply used to establish a baseline for comparison with the other two. Energy making its way through this FBase BPF contributes only to the total energy measurement.

The output of the synth goes into a stereo spreading delay line set to 3/16th BPM. It is stereo spreading by virtue of the fact that the right delay has and adjustable +/- additional delay. I typically set it to about 15-20 ms. Setting it to zero makes both left and right synchronous and sounds good with the 3/16th beat delay. But setting the additional delay to around 15 ms makes the whole space open up dramatically.

I have arbitrarily set the analysis filter Q's to 1, 2.5, and 10, because I typically run them in my analysis work at low, middle, and high frequencies. Feel free to play around with them...

The DSP use is relatively light at around 3.

These filters are fed an audio source -- in this case Kurt counting. But you'd never know it... (You would also never believe where these kinds of filters are used in my other life...)

- DM