ButterComp2

Compressor — Position 2

ButterComp2 offers four distinct compression models in a single module. Each model captures a different circuit philosophy — from the unique bipolar interleaved character of the original Airwindows algorithm to the fast transient snap of a FET design. All four share the same NY parallel blend approach: you always retain full control over how much compressed signal mixes with the dry input.

Select the model from the dropdown at the top of the module panel. The controls below the selector update to match the chosen model.

---

Choosing a Model

Model	Best For	Character
Classic	Mix bus glue, subtle cohesion	Rich, complex, harmonically dense — “the lushest glue compressor”
VCA	Drum buses, fast transients, punch	Transparent-to-colored; hard knee; fast and precise
Optical	Vocal bus, acoustic instrument bus, bass bus	Soft knee, program-dependent release, warm and musical
FET	Drum bus, any stem needing forward saturated character	Forward, punchy, saturated; behaves like a pushed 1176

Classic is the right starting point if you want compression you barely notice but always miss when it’s bypassed. VCA excels when you need a quick, controllable transient response. Optical smooths out expressive stems without clamping down unnaturally. FET adds the kind of forward character that makes a bus cut through the mix.

---

Classic (Airwindows ButterComp2)

Chris Johnson’s original ButterComp2 implements four independent compressors per channel in a bipolar, interleaved configuration — producing a complex, harmonically rich gain reduction character that feels more like analog hardware than conventional plugin compressors.

Airwindows FFI

The Classic model is compiled from Chris Johnson’s C++ source via Rust FFI. Audio processing is called once per buffer — O(1) FFI overhead — for real-time safety. VCA, Optical, and FET are native Rust implementations.

Controls

Control	Range	Description
Compress	0.0 – 1.0	Overall compression depth. Values above 0.6 become aggressive. Sweet spot: 0.30–0.55 for glue, 0.65–0.82 for NY parallel.
Output	0.0 – 1.0	Post-compression makeup gain. Start at 0.75–0.85 and trim to match bypass level.
Dry/Wet	0.0 – 1.0	Parallel blend. 0.0 = fully dry, 1.0 = fully wet. This is the NY parallel blend control.

Quick-Start Settings

Subtle Mix Bus Glue

Control	Value
Compress	0.42
Output	0.82
Dry/Wet	0.55

Gentle cohesion that makes elements feel related without obvious compression artifacts. A good starting point for any genre where the mix bus is already balanced and just needs to breathe together.

NY Drums (Classic)

Control	Value
Compress	0.78
Output	0.90
Dry/Wet	0.33

Aggressive compression blended sparingly beneath the dry signal. Brings up room ambience, snare rattle, and hi-hat bleed. Transients ride through untouched on the dry path.

Heavy Electronic Glue

Control	Value
Compress	0.62
Output	0.95
Dry/Wet	0.78

All elements welded into a single mass. Appropriate for electronic music where density and loudness take priority over preserved dynamics.

Tone Character by Compress Value

Compress	Character
0.20 – 0.35	Transparent — barely perceptible gain reduction, mainly adds harmonic glue
0.35 – 0.55	Musical glue, sustain enhancement, gel between elements
0.55 – 0.70	Assertive compression, audible gain reduction, dense character
0.70 – 0.85	Heavy compression — use lower Dry/Wet for NY parallel blend
0.85 – 1.00	Maximum compression — distortion character at high Dry/Wet values

---

VCA

Hard-knee voltage-controlled amplifier compression. The fastest and most controllable of the four models. Behaves like a precision gain-reduction element — what you set is what you get, with minimal program dependency. Add Character to introduce 1176-style harmonic coloration as the circuit is pushed.

Controls

Control	Range	Description
Threshold	-60 – 0 dB	Level at which gain reduction begins. Displayed as an integer dB value.
Ratio	1 – 20	Compression ratio above threshold. 2–4 for gentle control, 8+ for limiting behavior.
Attack	ms (integer)	Time for gain reduction to engage after the signal crosses threshold. Shorter values preserve less transient.
Release	ms (integer)	Time for gain reduction to recover after the signal falls below threshold.
Character %	0 – 100	Controls VCA circuit flavor and harmonic color. At 0, the model is clean and transparent. Higher values introduce 1176-style saturation and forward character.
Mix	0.0 – 1.0	Parallel blend between dry input and compressed signal. NY parallel control.

Quick-Start Settings

Drum Bus Punch

Control	Value
Threshold	-18 dB
Ratio	4
Attack	5 ms
Release	80 ms
Character %	30
Mix	0.60

Catches peaks quickly while leaving enough attack for kick and snare to cut. Character adds presence without obvious coloration.

Aggressive Parallel (1176-Style)

Control	Value
Threshold	-24 dB
Ratio	12
Attack	1 ms
Release	50 ms
Character %	75
Mix	0.35

Heavily compressed signal blended in for energy and forward presence. The dry signal preserves transient attack; the compressed signal adds density and color.

Mix Bus Control

Control	Value
Threshold	-6 dB
Ratio	2
Attack	20 ms
Release	200 ms
Character %	15
Mix	0.80

Light ratio and slow timing for transparent mix bus control. Character kept low to preserve the work done in individual tracks and buses.

---

Optical (Opt)

Optical compression uses a light-dependent resistor or equivalent photocell circuit to control gain reduction, producing a soft-knee response that follows the program material. Release time is program-dependent: it shortens during sustained loud passages and relaxes during quiet sections. This makes it forgiving on expressive, dynamic performances that would cause pumping or distortion with a VCA.

The Character control adjusts the optical element’s response curve — how much the release behavior tracks program dynamics and how much tonal color is introduced.

Controls

Control	Range	Description
Threshold	-60 – 0 dB	Level at which gain reduction begins. Displayed as an integer dB value.
Character %	0 – 100	Controls program-dependent release behavior and optical tube coloration. Higher values increase both the program dependency of release and the warmth of the tonal character.
Attack	ms (integer)	Initial gain reduction engagement time. Optical circuits are inherently slower than VCA — very short attack values may produce click artifacts on transient-heavy material.
Release	ms (integer)	Base release time. Actual release varies with program content when Character is above 0.
Mix	0.0 – 1.0	Parallel blend. NY parallel control.

Quick-Start Settings

Lead Vocal Bus

Control	Value
Threshold	-16 dB
Character %	50
Attack	15 ms
Release	250 ms
Mix	0.85

Soft knee keeps the compression invisible on most phrases. Program-dependent release lets the compressor breathe naturally between words and lines.

Acoustic Guitar / Piano Bus

Control	Value
Threshold	-20 dB
Character %	40
Attack	30 ms
Release	350 ms
Mix	0.90

Very musical response on instruments with complex transient structures. Long attack preserves pick attack and hammer-on character.

Bass Bus

Control	Value
Threshold	-14 dB
Character %	60
Attack	10 ms
Release	180 ms
Mix	1.0

Glues the bass stem together — DI, amp mic, room mic, and any sub or sample layers summed into a coherent low-end foundation. Character adds warmth that helps the different sources blend into a single instrument.

---

FET

Field-effect transistor compression is associated with fast, aggressive, forward-sounding character. The FET model is driven by an input gain stage rather than a threshold control, which means you push the compressor by increasing the input level — the same approach used by the original hardware units this style emulates. A high ratio crushes transients and creates the characteristic FET density. Auto Release enables program-dependent release behavior even in a model that is fundamentally fast and assertive.

Controls

Control	Range	Description
Input	dB (integer)	Drive level into the compressor stage. Higher values increase gain reduction and harmonic character.
Output	dB (integer)	Makeup gain after compression. Compensate for the level reduction caused by high Input settings.
Ratio	1 – 20	Compression ratio. At maximum (20:1), behaves as a limiter.
Attack	ms (integer)	Gain reduction engagement time. Very fast attacks (1–3 ms) produce the signature FET “all buttons in” density.
Release	ms (integer)	Base release time. Overridden by program content when Auto Release is enabled.
Auto Release	Toggle	Enables program-dependent release. The release time shortens on sustained loud signals and relaxes on quieter passages, reducing pumping at aggressive settings.
Mix	0.0 – 1.0	Parallel blend. NY parallel control.

Quick-Start Settings

Snare Bus

Control	Value
Input	+6 dB
Output	-4 dB
Ratio	8
Attack	2 ms
Release	45 ms
Auto Release	Off
Mix	0.70

Fast attack unifies the transient across top mic, bottom mic, room mics, and overhead bleed — all the sources on the snare bus hit as one. Moderate release allows the combined sustain to decay naturally. Compensate output to match bypass level, then increase Mix until the snare bus sits forward in the mix.

Room Mic Crush

Control	Value
Input	+12 dB
Output	-8 dB
Ratio	20
Attack	5 ms
Release	120 ms
Auto Release	On
Mix	0.45

Extreme compression on room mics blended in parallel for the exaggerated ambient character common in alternative and rock production.

Bus Density

Control	Value
Input	+4 dB
Output	-2 dB
Ratio	4
Attack	8 ms
Release	150 ms
Auto Release	On
Mix	0.80

Moderate drive with Auto Release keeps the mix breathing while adding the forward, present character that separates FET from cleaner compressor styles.

---

NY-Style Parallel Compression

The Mix control on every model (and Dry/Wet on Classic) implements parallel compression directly in the module — no aux routing required. The compressed signal is mixed beneath the dry signal, bringing up room ambience, sustain, and low-level detail without affecting transient attack, which comes through on the dry path.

Finding the Blend Point

Start Mix (or Dry/Wet) at 0.25 and slowly increase while listening to the subtle space between main elements — room reverb tails, string noise, breathing. Stop when you hear those details “fill in.” For drum buses this is typically 0.30–0.50. For mix buses, 0.50–0.75. For sources you want fully processed (e.g., a vocal bus), run Mix at 1.0 and adjust the compressor parameters directly.

The technique works with all four models because the underlying principle is circuit-agnostic: a compressed copy adds density and sustain without forcing those qualities onto the transients. The model choice determines the character of that compressed copy — from the harmonically rich Classic to the forward, colored FET.

---

See Also

• Techniques and Presets — NY parallel compression presets
• Genre Signal Chains — genre-specific ButterComp2 settings
• Instrument Buses — bus-specific compression starting points