Warmth Engine User Manual

WARMTH

Controls the overall intensity of the WarmthCompressor module. At zero, the module is bypassed. As the value increases, the dry/wet blend uses a square-root curve, meaning you hear a noticeable effect even at low settings. At 0.3, you get subtle harmonic enrichment. At 0.7 and above, the saturation and compression become clearly audible.

CHARACTER (TYPE)

Morphs the entire compression and saturation character along a single axis. This control simultaneously adjusts the compression threshold, attack/release times, ratio weights, saturation amount, and EQ coloring.

Soft (-1): Higher threshold (-18 dBFS), slower attack (30 ms), longer release (200 ms), emphasis on gentle 2:1 compression. EQ coloring adds +6 dB low-mid warmth at 150 Hz with minimal high-frequency rolloff. Tape saturation is minimal (0.15). The result is transparent, warm, and non-aggressive.

Neutral (0): Balanced settings. -24 dBFS threshold, 10 ms attack, 100 ms release. Even blend of compression ratios. Moderate tape saturation (0.5).

Punch (+1): Lower threshold (-30 dBFS), very fast attack (3 ms), short release (50 ms), heavier emphasis on 5:1-20:1 ratios. EQ coloring adds modest low-mid (+2 dB) but significant high-shelf cut (-4 dB at 2.5 kHz). Full tape saturation (1.0). The result is aggressive, dark, and heavily compressed.

TRANSIENT

Positive (Soften): Detected transient peaks are attenuated. At maximum, up to 30 dB of gain reduction is applied to transient peaks. A gentle high-shelf EQ at 2.5 kHz is also applied (up to -5 dB) to further soften the perceived attack. The result is rounder, more relaxed transients.

Negative (Enhance): Detected transient peaks are boosted (up to 12 dB). A corresponding high-shelf boost at 2.5 kHz (+3 dB) adds presence to the attack. The result is snappier, more aggressive transients.

The transient detection is stereo-linked (both channels share the same gain envelope) to prevent stereo image shifts.

ATTACK

Controls how quickly the gain recovers after a transient has been detected and shaped. With fast recovery (low values), the transient shaping is focused only on the immediate attack and the signal returns to normal almost instantly. With slow recovery (high values), the gain change lingers longer, affecting the early sustain portion as well and creating a more compressor-like effect.

SPECTRAL (Tilt)

Positive (Bright): Applies a low-shelf cut (-6.9 dB at 200 Hz) and a high-shelf boost (+6.9 dB at 4 kHz). The result tilts the spectrum toward brightness, reducing bass weight and adding top-end energy.

Negative (Dark): Applies a low-shelf boost (+6.9 dB at 200 Hz) and a high-shelf cut (-6.9 dB at 4 kHz). This tilts the spectrum toward warmth and darkness, adding low-end weight while softening the top.

Both the bright and dark EQ chains run simultaneously with fixed coefficients. The SPECTRAL knob crossfades between them using per-sample smoothing (100 ms time constant) for click-free operation.

PRESENCE

Provides both static and dynamic control of the 2-5 kHz presence band.

Static component: A peaking EQ at 3 kHz is crossfaded in based on the PRESENCE value. Positive values boost presence (up to +4.8 dB), negative values cut it (up to -4.8 dB).

Dynamic component: The processor monitors the presence region and adjusts movement according to the PRESENCE value. Positive settings make presence feel more controlled and forward; negative settings let it recede more during quieter moments.

Air band: When PRESENCE is positive, a high-shelf boost at 10 kHz (up to +3 dB) is progressively added, providing subtle shimmer above the presence band.

COHESION

A single control that scales the entire multiband compression behavior:

• Compression ratio: Scales from 1:1 (no compression) at 0% to 6:1 at 100%.
• Threshold: Moves from -20 dBFS at 0% to -40 dBFS at 100%, compressing more of the signal.
• Inter-band linking: Activates at 40% and increases to full linking at 100%. When linked, the average level across all three bands influences each band's compression, preventing one band from compressing independently while others remain uncompressed.
• Release time: Increases logarithmically from 200 ms (at 50%) to approximately 600 ms (at 100%) for smoother gain riding at higher settings.

The low band uses a sidechain high-pass filter at 90 Hz to prevent sub-bass content from triggering excessive compression. Automatic RMS-based makeup gain compensates for level loss, with a gentle +1.2 dB boost at full cohesion to make the effect feel additive.

STEREO

Negative (Narrow): The side channel level is reduced (from 1.0 down to 0.4 at -1.0), collapsing the stereo image toward mono. The mid channel receives a slight boost (+0.08 at maximum) to compensate for the narrowing. Additionally, bass frequencies below 180 Hz are progressively mono-ized (up to 40% mono blend at -1.0) to tighten the low end.

Positive (Wide): The side channel level is boosted (from 1.0 up to 1.4 at +1.0), widening the stereo image. The mid channel is very slightly reduced (-0.04) to maintain a natural balance. No bass mono-izing is applied in widening mode.

At center: No processing. The module bypasses entirely when the value is near zero.

INPUT GAIN (IN)

Adjusts the level entering the processing chain. Driving the input harder will push the WarmthCompressor saturation more aggressively. Reducing the input gives the modules less material to work with, resulting in a subtler effect.

OUTPUT TRIM (OUT)

Final level adjustment after all processing and dry/wet mixing. Use to match the output level with the bypassed signal for accurate A/B comparisons.

DRY/WET MIX

Parallel blend between the processed and original signal. At 50%, you get equal parts processed and dry signal. This is a powerful tool for retaining the natural dynamics and transient character of the source while still adding warmth and cohesion from the processing chain.