SV: Re: SV: Re: [sdiy] Nonlinearities in IR3109 filters

[Sean Costello]

The typical sine oscillator in computer music languages uses a sawtooth
generator (non-bandlimited) as an index into a sine table, which is an
example of extreme waveshaping.

I often create cheap sine waves for LFOs, by generating a sawtooth, using
abs() to convert this into a triangle wave, and then running this through a
low-order polynomial to approximate sine. All of these operations are
conventionally thought of as nonlinear operations which add harmonics, but
in this case they take away harmonics.

Sean Costello

----- Original Message ----- 
From: "JH." <jhaible at debitel.net>
To: "Jeff Farr" <moogah at gmail.com>; "Synth DIY"
<synth-diy at dropmix.xs4all.nl>
Sent: Thursday, August 03, 2006 3:40 PM
Subject: Re: SV: Re: SV: Re: [sdiy] Nonlinearities in IR3109 filters


> >I started to recognize that the
> >VCA really did color the sound and effect the dynamics quite a bit!
>
>
> Sure. And the real miracle is that the distortion of slightly overdriven
> differential pairs (think OTA-VCAs, or VCAs built from diff pairs)
> works "in reverse" for many sounds:
>
> While it would add harmonics to a sine wave ( ... where THD measurements
> are done), it actally *reduces* the level of harmonics on typical synth
> waveforms with sharp corners (tri, saw, ...)
>
> Just think for a moment, the very same overdriven ota structure is often
> used
> to form a crude sine wave from a triangle in synthesizer waveshapers .. .
>
> And there is more. The compression effect from soft limiting / signal
> rounding,
> be it in VCFs or VCAs, helps to reduce the amplitude modulation that
> occurs with two saw wave VCOs beating against each other, consequently
> allowing a higher amount of detuning without sounding chopped.
>
> JH.
>