UB DSP Grit Blender [WiN]

• Product: Grit Blender
• Developer: UB DSP
• Version: 1.3.0
• Format: VST3, AAX, CLAP
• Requirements: Windows 10 or later
• Source: https://ubdsp.com/grit-blender.html

Grit Blender is a saturation plugin built around a barycentric blending architecture that combines up to twelve discrete nonlinear algorithms simultaneously within a single processing stage. The DSP encompasses tape, tube, transistor, digital, diode, fuzz, desk, ribbon, pentode, rectifier, wavefolder, and asymmetric types — each a separate nonlinearity model — whose output weights are set continuously in real time via a triangle cursor or per-type sliders. It occupies any insert position where a saturator would normally sit, with M/S and L/R split modes extending its role to bus and mastering contexts. The primary differentiator is that harmonic character is navigable as a continuous spatial parameter rather than a switched selection.

Key Takeaway

Sessions where the target saturation character sits between two or more distinct types — tape warmth with transistor edge, desk sheen with measurable fuzz content — are where the blending architecture resolves a real routing problem rather than adding one. It displaces stacked saturator chains when the workflow cost of gain-staging multiple instances outweighs the benefit of discrete units. The pitch-tracking KEY TRACK crossover requires a MIDI feed to function; bass-register fundamentals stay clean only when note data arrives. Engineers mixing to a fixed saturation type — always tape, always the same console desk character — gain nothing from the triangle interface and may find Dual Stage the only compelling section.

Barycentric Triangle and the 12-Way Grid

The Triangle mode assigns three saturation types to the vertices of a triangle and maps the cursor’s position to a barycentric blend of all three — moving toward Tape adds tape character while proportionally reducing Tube and Transistor contributions. The cursor position is automatable, so blend shifts can be written as a continuous parameter move rather than programmed as a sudden type switch. Any three of the twelve types can occupy the vertices, covering the full range from very soft types (Ribbon, Asymmetric) to very hard (Fuzz, Rectifier, Wavefolder) in a single triangle.

The 12-Way Advanced mode exposes independent sliders for all twelve types simultaneously. No vertex assignment is required — each type’s weight runs from zero to full independently, and the plugin normalizes the output mix. This mode is slower to navigate by cursor than the triangle but allows precise harmonic recipes: 60% Tape, 20% Desk, 10% Asymmetric, 10% Tube is a configuration the triangle cannot represent directly because it has more than three active types at non-zero weight.

Both modes share the same saturation engine downstream. A fourth-power drive curve (x⁴) is applied before the type-specific gain stages, which means the lower half of the Drive knob is graduated very gently and the upper half accelerates rapidly. Engineers accustomed to linear drive knobs on other saturators will find the useful range for subtle coloring concentrated below 40% and the aggressive territory beginning around 60%. The Drive default of 50% is deliberately calibrated to land each type at its musical saturation sweet spot rather than at half of a linear range.

The triangle interface is not a morph that interpolates one algorithm into another — it is a weighted sum of the outputs of multiple simultaneously-running nonlinear stages. At a 50/50 split between Tape and Fuzz, both algorithms are processing in parallel and their outputs are summed at those weights. The resulting harmonic content is not a midpoint character; it is an additive combination of what each type produces separately. Moving toward Fuzz does not subtract the Tape content — it reduces Tape’s contribution weight while increasing Fuzz’s.

Pre/Post Emphasis and Spectral Behavior Under Drive

Before the saturation stage, a -4 dB high shelf at 3 kHz attenuates high-frequency content. Nonlinear processing generates harmonics proportional to the input energy at each frequency band, so an unattenuated high-frequency input would produce proportionally larger upper-harmonic content during saturation. The pre-emphasis shelf reduces this input before the algorithms process it; a +4 dB shelf at 3 kHz after the saturation stage restores the original spectral balance. The audible result is that raising Drive thickens the midrange and adds harmonic density without the top-end harshness that typically accompanies direct high-frequency saturation.

The delta FFT display in the center of the interface shows the spectral change the plugin is producing — not the raw output spectrum, but the difference between input and output at each frequency band. Orange above the center line indicates frequencies gaining relative energy; blue below indicates frequencies losing it. The display is normalized to remove overall loudness change, so the graph represents tonal redistribution rather than volume. At low Drive on Tape, the display typically shows a broad midrange gain and a slight high-frequency reduction — the tape compression character — while Fuzz at moderate Drive shows dominant odd-harmonic peaks.

The five Tone Curve Types (Default tilt EQ, Warm low shelf, Air high shelf, Smile co-shelf, Presence bell at 3 kHz) are post-saturation EQs whose shape the TONE knob controls. The same TONE position at -30 produces a different spectral result under each type: Default darkens with a rotating tilt, Warm reduces a 250 Hz shelf, Air cuts a 8 kHz shelf. Types are saved with presets. The transfer curve overlays the active tone shape in blue when TONE is off-center, so the EQ curve is visible alongside the saturation clipping shape simultaneously.

Dual Stage: Inter-Stage Filtering and Cascade Architecture

Dual Stage mode runs two independent blend panels in series — Stage 1 output feeds Stage 2 input with an automatic inter-stage filter between them. The inter-stage filter applies a high-pass at 80 Hz to remove DC offset and low-frequency rumble accumulated through Stage 1’s nonlinear processing, and an adaptive low-pass whose frequency adjusts between 5 kHz and 16 kHz depending on the hardness of the Stage 2 blend. Harder Stage 2 types (Rectifier, Fuzz, Wavefolder) pull the adaptive LP lower; softer types (Ribbon, Asymmetric, Desk) allow it to open toward 16 kHz. The filter runs automatically — no user control exposes the inter-stage cutoff.

The practical consequence is that feeding an aggressive Stage 2 with Stage 1’s output is more controlled than stacking two separate saturator instances on a chain. The adaptive LP prevents the accumulated upper-harmonic content from Stage 1 from being amplified again by Stage 2’s nonlinearity before entering Stage 2’s drive stage. Running Tape into Rectifier, for example, produces transistor-style hard clipping on material that already has gentle even-harmonic coloring, rather than an additive harmonic pile-up.

Stage 1 and Stage 2 have independent Drive controls (Drive defaults at 50%, Drive 2 at 30%) but share the Tone and Mix parameters. There is no independent tone shaping or output mix between the stages — the Tone EQ follows Stage 2, not Stage 1. Engineers who want different tonal shaping between stages need a separate EQ between two Grit Blender instances. The inter-stage filter is not a substitute for per-stage EQ.

Tube and Pentode types include RC power supply sag simulation in both stages. A 25ms charge rate and 80ms release rate model capacitor discharge under sustained loud input and recovery when the signal drops. Maximum gain reduction from sag is approximately 1.4 dB at full saturation with 100% Tube or Pentode weighting. In Dual Stage mode, each stage tracks its own sag state independently, so a Pentode Stage 1 feeding a Tube Stage 2 can be in different sag states simultaneously. Other saturation types are unaffected by the sag model regardless of blend position.

M/S Split, KEY TRACK Crossover, and Frequency Band Isolation

M/S mode decodes the stereo input into Mid (sum) and Side (difference) channels and processes each independently through the full saturation engine, including separate Drive, Drive 2, and Mix controls per channel. The Side channel at zero Drive passes through without saturation while the Mid channel runs hard. The transfer curve displays Mid in orange and Side in blue when M/S is active. L/R mode processes left and right channels with the same architecture without the M/S matrix step.

The Advanced panel exposes two independent Linkwitz-Riley 4th-order (LR4) crossovers that isolate which frequency range enters the saturation stage. The low crossover (default 80 Hz, range 40 Hz – 16 kHz) routes content below the threshold to a clean delay-compensated path that bypasses saturation; the high crossover (default 16 kHz) routes content above its threshold to a separate bypass path. Only the band between the two crossover points enters the saturation algorithms. In M/S or L/R mode, each channel carries its own pair of crossover handles with independent frequency settings — the Mid and Side channels can saturate different frequency ranges simultaneously.

When KEY TRACK is enabled via MIDI, the low crossover frequency follows the detected pitch of incoming MIDI notes in real time. A steep crossover tracks the played note, routing the fundamental frequency and everything below it to the clean bypass path while the harmonics above enter the saturation stage. RES AMT additionally applies a pre-saturation peak EQ boost (Q=3.0, up to 12 dB) at the detected pitch before saturation — the saturator then amplifies the harmonics generated around that peak. KEY TRACK requires a MIDI feed routed to the plugin. Without MIDI input, the crossover sits at the 80 Hz fixed default and pitch tracking is inert.

SMOOTH is a three-band dynamic EQ operating at 2.5 kHz, 6 kHz, and 12 kHz. Each band has an independent envelope follower that applies gain reduction only when excess energy is detected at that frequency. At gentle Drive settings or with soft saturation types, SMOOTH produces no audible effect. PUNCH uses a dual-envelope transient detector — a 0.1ms attack envelope for instantaneous peaks and a 20ms attack envelope for sustained level — and boosts signal when the fast envelope exceeds the slow one, restoring the peak-to-sustain ratio that saturation compresses. Neither PUNCH nor SMOOTH has user-accessible threshold or ratio controls; intensity scales via the 0–200% knobs in the Advanced panel.

Loudness Match, A/B Architecture, and the Comparison Problem

Saturation compresses dynamic range and increases perceived loudness. Without gain compensation, the processed signal almost always sounds louder than the bypass signal at the same output level, which creates a perceptual bias toward the saturated version regardless of its actual sonic contribution. Grit Blender’s LOUDNESS MATCH applies asymmetric K-weighted RMS measurement to counteract this: the input is measured with a standard +4 dB K-weighting shelf at 1.5 kHz while the output uses a reduced +2 dB shelf, correcting for the fact that standard K-weighting overestimates the loudness of harmonic-rich saturated content. Input measurement blends 70% K-weighted RMS with 30% peak level; output blends 60% K-weighted RMS with 40% peak.

The A/B system maintains two independent parameter snapshots accessible via toggle. Switching from A to B saves the current state before the switch, so neither slot is lost during comparison. The snapshots persist within the session but do not survive plugin reload or preset change. For comparing two entirely different saturation approaches — Tape-heavy on A, Desk-heavy with Dual Stage on B — the A/B toggle is faster than opening a second instance. The system does not support more than two comparison slots simultaneously.

Bypass uses a 5ms crossfade at the output to eliminate clicks during on/off toggling. The bypass path routes the pre-input-gain signal directly to the output — the Level control and ANALOG stage do not process the bypassed signal. ANALOG is an output stage that runs separately from the main saturation path and remains active even when saturation Drive is at zero, so the transformer roll-off, low-frequency head bump, and noise floor it adds persist through the signal chain as long as the toggle is on.

There is no per-type output metering that shows the harmonic contribution of each saturation algorithm at the current blend position. The harmonic spectrum display shows the aggregate effect of the blend, not a breakdown by type. Engineers who want to understand which type is contributing which harmonic content have to isolate types manually by moving the cursor to single-type positions and comparing spectrum states.

Where Blending Ends

The triangle interface navigates a continuous space between three types — it does not produce a type that does not already exist in the twelve. Blending Tape and Fuzz at equal weight produces an additive combination of tape even-harmonic saturation and rectified odd-harmonic distortion; it does not produce a new nonlinearity that neither type generates. Genuinely novel distortion characters — specific transformer saturation behavior, vintage germanium circuit topologies with thermal components, frequency-modulated distortion — are outside the scope of the existing twelve models regardless of blend position.

The plugin does not have a global output limiter beyond a soft clipper active above -0.9 dBFS. At high Drive with hard saturation types in Dual Stage, the output level before the soft clipper can reach levels that require significant Level attenuation to avoid clipping the DAW channel. Monitoring the output meter during aggressive Drive settings is a necessary step that the LOUDNESS MATCH system, which only operates in the gain-compensation path, does not substitute for.

Linux support is described as community-tested rather than officially validated across all distributions and DAWs. CLAP and VST3 formats are available on Linux; AU and AAX are not. Engineers running exclusively Linux sessions with Pro Tools or Logic will find the AAX and AU formats absent. The Windows and macOS builds carry full format coverage including AAX for Pro Tools integration.

The 39 factory presets span seven categories and cover the primary instrument types and blend strategies. The preset library represents a small starting-point collection relative to the parameter space the twelve-type blending system opens. Producers who work from presets as a primary workflow will find themselves designing from scratch more often than other saturators at this price point require.

FAQs