Comb+ Filter Simulation



Author: Snyxol

On the MW2/XT you can emulate a comb+ filter without resonance (mono chorus) modulated by envelope/LFO/whatever:
The mod source (in this example: free envelope) shifts the relative phase between both osc.s.
It behaves the same as a 1-osc-synth with comb+ filter.
If the stages of the envelope have different slopes, the fatness changes between them.
When you apply the normal chorus in addition, you can get mind-blowingly fat phasing. Sounds cool with saw, sqr and chorus2.
The case of sqr is particularly interesting: the sum of 2 phase shifting square waves is a symmetric PWM pulse wave, containing only odd harmonics; the free envelope controls the pulse width.

Oscillators: same wave, exactly (!) same pitch, no detune
Phase 1: 3°
Phase 2: try out!
Mixer: both waves at same level (63)
Trigger2: mode: poly; assign: normal
Modmatrix:
Modifier 2: src 1: modify #4; type: filter; param: 120
Modifier 3: src 1: modify #2; src 2: modwheel; type: *
Modifier 4: src 1: free env; type: diff
Mod 1: src: modify #3; amnt: -36; dest: Osc1 pitch
Mod 2: src: modify #3; amnt: +36; dest: Osc2 pitch
Mod 3: src: keyfollow; amnt: -45; dest: M2 Amount
Mod 4: src: keyfollow; amnt: +45; dest: M1 Amount


Set the mod depth of phase with modwheel, and set startingpoint of phase modulation with the phases of both oscs.

Here's how it works:
Mod 1..4: see "Pitch Independent Phasing" and "Frequency as Mod Source".
To move the relative phase the oscs must be detuned.
The phasing speed is proportional to the frequency difference (not detune!) of the oscs.
Mod 3/4 make the frequency difference pitch independent.
Modifier 4: the slope of the free envelope must be proportional to the phasing speed => the env must first be differentiated and then it must modulate the frequency difference.
Modifier 2: before the differentiated envelope modulates the frequency difference, it is smoothed to reduce abrupt fatness changes between the stages.

Happy synthesizing,
Snyxol