veils

veils · utilities · schema v1

Quad VCA + soft-clip summing mix (Mutable Instruments Veils archetype — analog hardware, clean-room from-spec impl). Four independent VCAs, each with audio in, CV in (summed with knob), gain knob spanning [0, 2], and a per-channel response toggle: LIN for CV / control signals, EXP (squared) for audio / smooth fades. Per-channel direct outs are pre-mix, pre-clip. A separate MIX out sums all four channels and applies a tanh soft-clip — gain is NOT clamped at 1.0 per channel, so knob + CV can push above unity for warm overdrive on the mix bus.

A quad VCA and mixer modelled on the Mutable Instruments Veils: four independent voltage-controlled amplifiers that each have a gain knob, a CV input, a direct out, and a linear/exponential response toggle — feeding one summed MIX out. Per channel out = in · shape(knob + cv), where the gain knob spans 0..2: at unity-position knob a ±1V CV sweeps gain from 0 (CV at -1) to 2 (CV at +1). The four channels sum and pass through a tanh soft-clip (mix = tanh(sum)) — gain is NOT clamped at unity, so pushing knob+CV high overdrives into warm saturation rather than digital clipping. Use it as four utility VCAs, a CV-controlled mixer, or four cross-fading/amplitude-modulated voices. There is a DSP worklet for the per-sample VCA + soft-clip math.

the faceplate

8 inputs · 5 outputs · 8 params

about this module

Hover a control, or open the patch panel and hover a jack.

inputs

id	cable	what it does
`in1`	audio	Channel 1 audio input — multiplied by its VCA gain shape(gain1 + cv1) into the channel-1 direct out and the mix. audio signal
`in2`	audio	Channel 2 audio input, multiplied by shape(gain2 + cv2). audio signal
`in3`	audio	Channel 3 audio input, multiplied by shape(gain3 + cv3). audio signal
`in4`	audio	Channel 4 audio input, multiplied by shape(gain4 + cv4). audio signal
`cv1`	cv	Channel 1 gain CV. Summed RAW with the gain knob (no scaling): a ±1V LFO at a unity-position knob sweeps gain 0..2, the natural full-range VCA modulation, so this is the per-channel tremolo / ducking / cross-fade control. control voltage (CV)
`cv2`	cv	Channel 2 gain CV (raw, summed with gain2). control voltage (CV)
`cv3`	cv	Channel 3 gain CV (raw, summed with gain3). control voltage (CV)
`cv4`	cv	Channel 4 gain CV (raw, summed with gain4). control voltage (CV)

outputs

id	cable	what it does
`out1`	audio	Channel 1 direct VCA out — in1 · shape(gain1 + cv1), taken BEFORE the summing bus and soft-clip, so downstream sees the raw post-VCA channel. audio signal
`out2`	audio	Channel 2 direct VCA out (pre-mix, pre-clip). audio signal
`out3`	audio	Channel 3 direct VCA out (pre-mix, pre-clip). audio signal
`out4`	audio	Channel 4 direct VCA out (pre-mix, pre-clip). audio signal
`mix`	audio	The summed bus, soft-clipped: tanh(out1 + out2 + out3 + out4). Because the sum is not clamped before the tanh, driving channels past unity produces warm saturation — Veils' signature overdrive — instead of a hard clip. audio signal

params

id	label	range	default	curve
`gain1`	Ch1	0..2	0	linear
`gain2`	Ch2	0..2	0	linear
`gain3`	Ch3	0..2	0	linear
`gain4`	Ch4	0..2	0	linear
`resp1`	Resp1	0..1	0	discrete
`resp2`	Resp2	0..1	0	discrete
`resp3`	Resp3	0..1	1	discrete
`resp4`	Resp4	0..1	1	discrete

controls

control	what it does
Ch1	Channel 1 VCA gain knob, linear 0..2 (default 0 = silent). Sums with CV1; the past-unity range exists so knob + CV can push the channel into the mix soft-clip.
Ch2	Channel 2 VCA gain, linear 0..2 (default 0). Sums with CV2.
Ch3	Channel 3 VCA gain, linear 0..2 (default 0). Sums with CV3.
Ch4	Channel 4 VCA gain, linear 0..2 (default 0). Sums with CV4.
Resp1	Channel 1 response toggle: 0 = LINEAR (gain follows knob+CV directly — best for control signals), 1 = EXPONENTIAL (the signal is squared, giving smoother perceptual fades — best for audio). Default LINEAR.
Resp2	Channel 2 response toggle (0 = linear, 1 = exponential). Default LINEAR.
Resp3	Channel 3 response toggle (0 = linear, 1 = exponential). Default EXPONENTIAL (audio-friendly out of the box).
Resp4	Channel 4 response toggle (0 = linear, 1 = exponential). Default EXPONENTIAL.

source

veils.ts on GitHub.