[sdiy] They aren't sawtooths, they're ramps

[Richard Wentk]

On 10 Nov 2009, at 22:13, cheater cheater wrote:

> Actually falling and rising exponential ramps are harmonically
> different

You can do this stuff easily with DSP - arbitrary trig functions,  
varying curves, whatever.

It turns out that for most variations, there are only relatively minor  
differences in the sound.

The positive thing about the original analog waveshapes is that  
they're musically useful. Musically it doesn't really matter if your  
Moog sine has x% distortion - it still sounds sine-like, and can do  
sine-like things.

Likewise pulse width modulation is a musically interesting effect -  
much more interesting than almost any other kind of shape modulation,  
short of wavefolding.

Differences between ramps are musically trivial in comparison. Even a  
saw<->tri transformation doesn't sound that interesting, because you  
just get more harmonics, and not the more interesting comb-like effect  
you get from pulse modulation.

> So I say that there *IS* a point in making a distinction in names.

Well - yes and no.

Functionally, to reach the next level of musical usefulness, you  
really need to use a digital oscillator with an arbitrary harmonic mix  
- ideally, one with a harmonic map across the pitch range, so the mix  
is different for each note.

One of the features of the 901 core is that it sounds subtly different  
at different frequencies. That makes for a more interesting VCO than  
one that produces a mathematically perfect saw/ramp/etc from 0.001Hz  
to microwave frequencies.

Then again, there aren't many modular users who take the time to do  
even the most basic waveshaping and filtering, and get beyond 'sort of  
buzzy and filtered' as a timbre.

Richard