You wouldn't say that the first picture looks and sounds like a bandlimited sawtooth?foosnark wrote:That's correct. In FM8 about as close as you can get is 42 feedback on the modulator, 33 from the modulator into the carriers:tony tony chopper wrote:is it? I think it reaches noise before it gets the full brightness/the higher harmonics of a sawtoothThere's a point where it "is" a theoretical sawtooth
Any more modulation and you start to lose the sawtooth shape:
Any more feedback and you add noise:
That "noise" you're getting is aliasing, (edit- hmmm, this is with feedback so maybe not, don't know) with a better synth for testing you get a hell of a lot more sidebands (and some funky sci-fi waveshapes on the way) before it gets aliasing "noise".
But I don't know what would happen in a theoretically ideal situation, that would actually have to be done with a spreadsheet or whatever.
Anyway the point is, yes you do have a "bandlimited saw" at a certain point, and even more importantly you can easily demonstrate and use the three basic families of subtractive synthesis waveforms and their (family) connection to C:M ratios, and go from there. The family difference between saws and pulses are the ordered and predictable presence or absence of specific harmonic partials, surely you can see how making these connections makes programming phase modulation synths easier.



