Ferment Documentation
User guide for Ferment v0.9.2 — expressive sound exploration engine
Installation
Ferment is available in three plugin formats. Download the format that matches your DAW from the download page.
| Format | Compatibility | File |
|---|---|---|
| VST3 | Reaper, Bitwig, Cubase, FL Studio, most DAWs | Ferment.vst3 |
| AU | Logic Pro, GarageBand, MainStage (macOS only) | Ferment.component |
| CLAP | Bitwig, Reaper, future DAWs | Ferment.clap |
macOS installation paths
Unzip the downloaded archive and move the plugin file to the matching folder:
| Format | Path |
|---|---|
| VST3 | ~/Library/Audio/Plug-Ins/VST3/ |
| AU | ~/Library/Audio/Plug-Ins/Components/ |
| CLAP | ~/Library/Audio/Plug-Ins/CLAP/ |
Cmd + Shift + G in Finder
and paste the path above.
Windows installation paths
Unzip the downloaded archive and copy the plugin file to the matching folder, then rescan in your DAW:
| Format | Path |
|---|---|
| VST3 | C:\Program Files\Common Files\VST3\ |
| CLAP | C:\Program Files\Common Files\CLAP\ |
Verify in your DAW
After placing the file, restart your DAW (or rescan plugins). Ferment should appear in your plugin list under Ciderka or Ciderka Labs.
If you don't have a DAW, you can use a free standalone host like Carla or Element. See the host setup guides on the home page.
Interface Overview
Ferment's interface is divided into four tabs, each accessible from the top bar of the plugin window. The window size is fixed at 1060 × 690 pixels.
Four modes
| Tab | Purpose | Ideal for |
|---|---|---|
| Machine | Direct control of all 92 DSP parameters organized into 14 modules | Precise sound design, manual tweaking, live performance |
| Magic | AI-powered preset generation via text descriptions and a spider chart | Quick inspiration, exploring new sounds, starting points |
| Material | Instrument profile management — analyze and store your instrument's characteristics | Tailoring Ferment to your specific instrument |
| Library | Preset browser with 4×4 card grid, spider previews, filters and store integration | Browsing presets, previewing sounds, purchasing preset packs |
Top bar
The top bar shows the four tab buttons (Machine / Magic / Material / Library), the current preset name, an oversampling selector (Off / 2× / 4×) and a bypass toggle. When an instrument profile is applied, a small dropdown shows the active profile name.
Knob controls
All knobs respond to vertical drag (up = increase, down = decrease). Double-click a knob to reset it to default. When an instrument profile is active, knobs show additional visual guidance: a green arc marks the recommended "sweet spot" range for your instrument, an orange arc indicates the zone outside the sweet spot (still usable, but less typical for your instrument type) and a small teal dot below a knob signals that its response curve has been remapped to suit your instrument's characteristics.
Quick Start
The fastest way to get a sound out of Ferment:
1. Generate with Magic
Open the Magic tab. Type a short description of the sound you want — for example "warm tape delay with subtle movement" or "aggressive distortion, bright and wide". Optionally drag the spider chart axes to shape the character. Click Cast Spell.
Ferment's AI will generate 2–3 preset suggestions. Browse through the carousel and pick one that sounds closest to what you need.
2. Fine-tune in Machine
Switch to the Machine tab. All 92 parameters are now set by the AI-generated preset. Tweak individual knobs to dial in exactly the sound you want — adjust the drive, change the reverb mix, shift the EQ, or modify the modulation.
3. Save your preset
Click Save in the preset browser. Your preset is stored as a .slpreset
file in ~/Documents/Ferment Presets/ and will appear in the preset list next time
you open Ferment.
Full Workflow
For the best results, use all three tabs together. This workflow takes you from your unique instrument to a fully customized preset.
Step 1 — Material: Define your instrument
Open the Material tab and create a new instrument profile. Give it a name (e.g. "My Stratocaster"), select a category (Electric Guitar) and optionally record a few audio clips at different dynamics — quiet, normal, loud and articulation samples.
Ferment will analyze your clips and compute spectral metrics: brightness, saturation, harmonics, attack time and dynamic range. These metrics help the AI understand your instrument's character.
Set this profile as the default by clicking Set Default.
Step 2 — Magic: Generate with AI awareness
Switch to the Magic tab. With your instrument profile active, the AI now knows your instrument's frequency response, dynamic behavior and tonal character. Describe what you want — the generated presets will be tailored to your specific instrument.
The AI always generates a complete preset controlling all 92 parameters, tailored to your instrument's spectral profile and the text description you provide.
Step 3 — Machine: Final adjustments
Switch to Machine and refine. With the instrument profile active, knobs show green "sweet spot" arcs — ranges that work well for your instrument. This visual guidance makes it easy to stay in a musically useful range while still having full freedom to go beyond.
Step 4 — Save
Save the preset. Version 2 presets (.slpreset) automatically store the
instrument profile context, so when you recall the preset later, Ferment knows which
instrument it was designed for.
Machine Tab
Machine is the core sound design interface — 92 parameters organized into 14 modules. Each module can be individually bypassed. The signal flows through 12 processing stages:
Modulation sources (LFO 1, LFO 2, Envelope Follower) and the Modulation Matrix run in parallel, feeding into any of 17 modulation destinations across the chain.
Oversampling
The oversampling selector in the top bar lets you choose between Off, 2× and 4×. Higher oversampling reduces aliasing artifacts (especially from the Shape module) at the cost of CPU. For live performance, Off or 2× is recommended. For rendering or recording, try 4×.
Module bypass
Each module header has a bypass toggle. Bypassed modules pass audio through unchanged, saving CPU. The bypass state is saved with presets.
See the Effects Reference below for the complete list of parameters in each module.
Magic Tab
Magic uses AI to translate your intentions into Ferment presets. Describe what you want in words, shape it with the spider chart and let the AI configure all 92 parameters for you.
Spider chart
The 8-axis radar chart lets you define the perceptual character of the sound you want. Drag any axis from the center (neutral) outward (extreme):
| Axis | Center (0) | Edge (1) |
|---|---|---|
| Warmth | Cold, clinical | Warm, analog feel |
| Brightness | Dark, muffled | Bright, airy |
| Space | Dry, close | Wet, spacious |
| Texture | Smooth, clean | Gritty, distorted |
| Movement | Static, steady | Dynamic, evolving |
| Depth | Flat, 2D | Deep, layered |
| Attack | Soft onset | Sharp, percussive |
| Width | Narrow, mono | Wide stereo |
Text prompt
The text field accepts natural language descriptions. Be as specific or abstract as you like:
- "warm tape delay with subtle movement"
- "aggressive distortion, bright and wide"
- "underwater cathedral reverb"
- "lo-fi vinyl crackle with gentle compression"
Results carousel
After clicking Cast Spell, the AI generates 2–3 preset suggestions. Each appears as a card with a name (e.g. "Amber Growl", "Crystal Rain") and can be previewed instantly. Select one to apply it, then switch to Machine to fine-tune.
Offline fallback
If the server is unreachable, Ferment falls back to a deterministic algorithm (SpiderConverter) that maps spider chart values directly to parameters. The result is a reasonable approximation without AI involvement. You'll see the message "Conjured from the local spellbook" when this happens.
Licensing
Exploring Magic — generating presets, previewing and adjusting the spider chart — is completely free. Saving a Magic-generated preset to disk requires a license. You get 3 free trial saves. Machine mode has no license restrictions at all.
Material Tab
Material is Ferment's instrument profiling system. It analyzes your instrument's sonic characteristics and uses that data to improve AI preset generation and provide visual guidance in the Machine tab.
Layout
The Material tab has a three-column layout:
- Left — Profile tile cards (up to 24 profiles: built-in + user-created)
- Center — Profile form: name, category, notes and audio clip recording
- Right — Tabbed panel: HARMONICS (Harmonic Profile Editor, new in v0.6.1) and SPECTRUM (real-time spectrum analyzer)
Creating a profile
- Click New Profile in the left column.
- Enter a name (e.g. "My Jazz Bass") and select a category.
- Choose a play style (or leave on Auto to let Ferment detect it from the category).
- Record audio clips at different dynamics — Ferment supports up to 8 clips per profile: Quiet, Normal, Loud, Articulation, Sustain, Staccato, Harmonic and Special.
- Ferment analyzes the recordings and extracts spectral metrics.
- Click Set Default to use this profile across Machine and Magic tabs.
Categories
Ferment supports 14 instrument categories plus Generic: Electric Guitar, Acoustic Guitar, Bass, Piano, Voice, Synth, Drum, Strings, Brass, Woodwind, Percussion, Organ, Pad and Sample.
Play style
Each profile includes a play style that describes how you interact with the instrument. This affects how Ferment maps dynamics, drive response and transient handling to the knob sweet spots — a bowed instrument gets smooth onset and sustained-tone-optimized mappings, while a struck instrument gets fast transient response and minimal sustain shaping.
| Play style | Typical instruments | Effect on mappings |
|---|---|---|
| Auto | — | Automatically detected from instrument category |
| Bowed | Violin, cello, viola | Smooth onset, sustained tone, long dynamics |
| Picked | Electric guitar with plectrum | Sharp bright transient, medium sustain |
| Fingered | Fingerstyle guitar, bass, pizzicato | Softer transient than pick, warm attack |
| Struck | Piano, marimba, hammered dulcimer | Hammer transient with sustain, fast attack |
| Sustained | Organ, synth pad | Continuous tone, no transient emphasis |
| Vocal | Voice, wind instruments | Natural texture protection, de-ess sensitive |
| Percussion | Drums, hand percussion | Fastest transient, minimal sustain |
Spectral metrics
After recording clips, Ferment computes:
| Metric | What it measures |
|---|---|
| Brightness | High-frequency energy relative to total (0–100%) |
| Saturation | Harmonic distortion content (0–100%) |
| Harmonics | Harmonic richness and overtone density (0–100%) |
| Spectral Centroid | Perceived "center of mass" of the frequency spectrum (Hz) |
| Attack Time | Time from silence to peak (ms) |
| Dynamic Range | Difference between quietest and loudest passages (dB) |
These metrics are sent to the AI when generating Magic presets, allowing it to tailor results to your instrument. They also drive the "sweet spot" arcs on Machine knobs.
Knob remapping and sweet spot visualization
Profiles don't just inform the AI — they physically remap how knobs behave in Machine. Each affected knob gets a custom response curve (log or exponential) so that the same physical movement covers a musically useful range for your instrument. A bright electric guitar and a dark upright bass get different drive curves, different EQ response ranges, and different reverb dynamics — even when the same preset is loaded.
The visual indicators in Machine when a profile is active:
| Indicator | Meaning |
|---|---|
| Green arc on knob | Recommended sweet spot — parameter values that work well for your instrument |
| Orange arc on knob | Outside the sweet spot — still usable, but may not suit your instrument |
| Teal dot below knob | This knob's response curve has been remapped for your instrument profile |
Harmonic Shaping — new in v0.6.1
Harmonic Shaping is a pitch-relative correction engine embedded in the Material profile. It targets the individual harmonics of your instrument — the 2nd, 3rd, 4th… up to the 16th — and applies individually adjustable boosts or cuts that follow the pitch of your playing.
This is fundamentally different from EQ. A static EQ boost at 880 Hz helps the 2nd harmonic of A4 (440 Hz), but when you play E5 (659 Hz), the 2nd harmonic shifts to 1318 Hz — and your EQ is now boosting the wrong frequency. Harmonic Shaping tracks the fundamental with a YIN pitch detector (shared from the Octaver module) and repositions all 16 filters continuously as you move across the register.
Harmonic Profile Editor (Material tab → HARMONICS panel)
The HARMONICS panel in the right column of the Material tab shows a 16-row bar graph. Each row represents one harmonic (1 = fundamental, 2 = octave, 3 = fifth above octave, …). The horizontal bar indicates the correction applied to that harmonic:
- Bars extending right = boost (positive %)
- Bars extending left = cut (negative %)
- No bar = neutral (0%)
Editing
| Action | Result |
|---|---|
| Click + drag horizontally | Set correction for that harmonic (±100%, 5% steps) |
| Shift + drag | Fine mode — 4× more precise |
| Double-click | Reset that harmonic to 0% |
| Scroll wheel | ±5% adjustment |
Preset dropdown
Select a factory profile to start from, then adjust manually. Click Save to store your modifications as part of the current Material profile. Import and export harmonic profiles as standalone files to share with other Ferment users.
Factory profiles
| Profile | Character |
|---|---|
| Stradivarius Even | Even harmonics boosted (+8…+25%), odd harmonics cut — warm, singing tone |
| Guarneri Balanced | Even harmonics gently lifted (+5…+12%), odd harmonics unchanged — natural warmth |
| Bright Projection | Harmonics 3–8 progressively boosted (+8…+28%) — more projection, presence |
| Dark & Warm | Harmonics above 6th progressively cut (-12…-30%) — mellow, focused tone |
| Hollow | Even harmonics attenuated (-50%) — hollow, reed-like, klarinet character |
Amount knob
The Amount knob (shown as "Harm" in the Character module) scales all harmonic corrections simultaneously. At 0% the correction is completely bypassed; at 100% it applies the full profile. Default is 65%. This knob is shared between the Material tab preview and the Character module in Machine — moving one moves the other.
Harm knob (Machine → Character module)
The Character module in Machine gains a Harm knob (0–100%) for live blending. The harmonic profile itself comes from the active Material profile — you set the shape in Material, blend it in during performance using Harm. The AI preset system deliberately excludes this knob: harmonic correction is treated as instrument identity, not sound character.
Technical notes
- Pitch detection — YIN algorithm shared from the Octaver module (period cache, ~42 ms update, 10 ms pitch smoother for click-free transitions)
- Filters — 16 RBJ peaking biquads; Q = 8–10; corrections mapped internally to ±6 dB maximum
- CPU — ~1.5% of a single core at 48 kHz; 0% extra for pitch detection (shared)
- Latency — zero latency (biquad filters are sample-accurate); pitch tracking lag ~50 ms
- Monophonic input — harmonic correction is designed for one note at a time; when YIN pitch confidence drops below 80% (chords, noisy signals), the correction automatically fades to bypass
- Nyquist guard — filters above 0.45 × sample rate are skipped; behavior is safe at all sample rates
Library Tab
The Library tab is Ferment's preset browser — a full-screen view for discovering, previewing, and managing presets. Click the LIBRARY button in the top navigation bar to open it.
Card Grid
Presets are displayed as cards in a 4-column, 4-row paginated grid (16 per page). Each card shows the preset name, its source (Factory, User or Store) and a small spider chart preview visualizing the sonic character across 8 axes: Warmth, Brightness, Space, Texture, Movement, Depth, Attack and Width.
Filters
The filter bar at the top lets you narrow the view. Available filters include Factory (built-in presets), User (your saved presets), Store (preset packs available for purchase) and Favorites. Active filters are shown as highlighted chips; click a chip to toggle it.
Preview
Click the play button on any card to preview that preset without committing. The preset is loaded temporarily so you can hear it in context with your audio. Click a different card to switch, or press the main card area to apply the preset permanently.
Store Integration
Preset packs from the store appear alongside your local presets. Packs that require purchase show a buy button. Purchasing is handled securely via Paddle — after payment, the pack is delivered instantly and the presets appear in your Library.
Pagination
When you have more than 16 presets matching the current filter, use the pagination controls at the bottom of the grid to navigate between pages. The current page and total count are displayed.
Effects Reference
Complete parameter reference for all 14 modules. Parameters marked with * are modulation destinations (can be controlled by LFO or Envelope Follower via the Mod Matrix).
Input
| Parameter | Range | Description |
|---|---|---|
| Input Gain | -70 to +12 dB | Main input level |
| Input Trim | -3 to +3 dB | Fine gain adjustment |
| Phase Invert | On / Off | Flip signal polarity (180°) |
| HP Filter | 20 – 2000 Hz | High-pass filter cutoff |
| LP Filter | 200 – 20000 Hz | Low-pass filter cutoff |
Pre-EQ
Focus-based 3-band EQ. The Focus knob sets the center frequency — Low shelf sits at Focus÷4, Mid peak at Focus, High shelf at Focus×4.
| Parameter | Range | Description |
|---|---|---|
| Focus * | 40 – 8000 Hz | Center frequency for all 3 bands |
| Low Gain | -18 to +18 dB | Low shelf gain (at Focus÷4) |
| Mid Gain | -18 to +18 dB | Mid peak gain (at Focus) |
| Mid Q | 0.1 – 18 | Mid band width (higher = narrower) |
| High Gain | -18 to +18 dB | High shelf gain (at Focus×4) |
Shape (Waveshaper)
Distortion / saturation module with tube-amp modeling. Four continuously morphable curves (Soft Clip → Hard Clip → Fold → Wrap) are shaped by input-dependent drive, power supply sag and output transformer emulation for authentic analog character.
The drive responds to your playing dynamics — an internal envelope follower blends 30% static gain with 70% dynamic tracking, so the distortion "breathes" with your performance. A power supply sag model (fast attack ~2ms, slow release ~30ms) compresses the effective drive under heavy load, adding the natural compression feel of a pushed tube amp. An output transformer low-pass filter with drive-dependent cutoff (20 kHz clean → 3.5 kHz driven) and subtle pink noise modulation round off the analog simulation.
| Parameter | Range | Description |
|---|---|---|
| Curve Type | 0 – 1 | Morph between distortion curves (soft → hard → fold → wrap) |
| Drive * | 0 – 1 | Saturation intensity (input-dependent with envelope follower) |
| Asymmetry | -1 to +1 | Harmonic coloration (even vs. odd harmonics) |
| Knee | 0 – 1 | Transition smoothness (0 = sharp, 1 = gentle) |
| Bias | -0.7 to +0.7 | DC offset for asymmetric clipping |
| Shape Mix | 0 – 1 | Dry/wet blend for this module |
Dynamics
Professional lookahead compressor with analog-inspired behavior. A ~1.5ms lookahead buffer catches transients before they hit, while adaptive gain smoothing scales with compression depth to eliminate artifacts. A sidechain high-pass filter at 80 Hz prevents bass content from causing pumping. Gain-reduction-dependent saturation adds subtle analog warmth that scales naturally with compression depth — transparent on clean signals, warmly colored under heavy compression. A soft peak limiter at 0.95 catches residual overshoots.
Release behavior is program-dependent, adapting smoothly to signal dynamics. The compressor reports its latency for DAW plugin delay compensation.
| Parameter | Range | Description |
|---|---|---|
| Threshold | -60 to 0 dB | Level above which compression starts |
| Ratio | 1 : 1 – 20 : 1 | Compression ratio (1 = no compression) |
| Attack | 0.1 – 100 ms | How fast the compressor reacts |
| Release | 10 – 1000 ms | How fast the compressor lets go |
Multiply
8-voice SOLA chorus / doubler for thickening and widening.
| Parameter | Range | Description |
|---|---|---|
| Voices | 1 – 8 | Number of doubled voices |
| Detune | 0 – 50 cents | Pitch offset between voices |
| Spread | 0 – 1 | Stereo placement of voices (L–R) |
| Delay | 0 – 50 ms | Time offset between voices |
| Depth | 0 – 1 | LFO modulation depth |
| Rate | 0.1 – 5 Hz | Internal LFO speed |
| Mix * | 0 – 1 | Dry/wet blend |
| Humanize | 0 – 1 | Random drift in timing and level (±2ms, ±1.5dB) |
Octaver
Signal-adaptive SOLA pitch shifter, ±12 semitones. Real-time AMDF pitch detection with parabolic interpolation ensures ±2.5 cent accuracy. Pitch-up onset under 5 ms.
| Parameter | Range | Description |
|---|---|---|
| Oct Up | 0 – 12 ST | Pitch shift up (semitones) |
| Oct Down | 0 – 12 ST | Pitch shift down (semitones) |
| Mix * | 0 – 1 | Dry/wet blend |
Temporal
Delay / echo with optional tempo sync to DAW.
| Parameter | Range | Description |
|---|---|---|
| Time | 1 – 2000 ms | Delay time (or note division when synced) |
| Feedback | 0 – 0.95 | Repeat amount (0 = single echo) |
| Tone | 0 – 1 | Low-pass filter on feedback loop (0 = dark, 1 = bright) |
| Stereo | 0 – 1 | L/R delay time difference (0 = mono delay) |
| Sync | On / Off | Lock delay time to DAW tempo |
| Mix * | 0 – 1 | Dry/wet blend |
Spatial
Reverb engine built on an 8-line Feedback Delay Network (FDN) with Hadamard coupling matrix and RT60-based feedback. Combined with mid/side stereo processing. Clip-free output via soft limiter.
| Parameter | Range | Description |
|---|---|---|
| Width * | 0 – 1 | Stereo width expansion |
| Pan * | -1 to +1 | Stereo pan position |
| Reverb Size | 0 – 1 | Room size (0 = small, 1 = hall) |
| Decay | 0.1 – 10 s | Reverb tail length |
| Damp | 0 – 1 | High-frequency absorption (0 = bright, 1 = dark) |
| Predelay | 0 – 100 ms | Gap before reverb onset |
| Reverb Mix | 0 – 1 | Dry/wet blend for reverb |
| Mid/Side | 0 – 1 | Balance between mid (center) and side (stereo) energy |
Character
Tone shaping with harmonic enhancement and pitch-relative harmonic correction. Body adds even-harmonic content (2nd, 4th) for tube-like warmth. Tone controls a presence peak at 3.5 kHz for definition and air. Harm applies the Harmonic Profile from the active Material instrument profile (see Harmonic Shaping).
| Parameter | Range | Description |
|---|---|---|
| Body * | -1 to +1 | Thin ↔ Fat — even-harmonic saturation on low-frequency energy |
| Tone * | -1 to +1 | Dark ↔ Bright — presence peak at 3.5 kHz (±3 dB) |
| Harm new | 0 – 100% | Pitch-relative harmonic correction amount — blends the active Harmonic Profile in. Default 65%. Shared with Material tab Amount knob. |
Tube Blend
Tube saturation preamp placed before the waveshaper. Adds asymmetric soft-clipping with even harmonics, blendable from fully clean to driven.
| Parameter | Range | Description |
|---|---|---|
| Tube Blend | 0 – 1 | Dry/saturated blend (0 = clean, 1 = full tube preamp) |
Crackle
Vinyl and tape noise generator. Random impulse bursts through a low-pass filter for authentic surface noise — from subtle atmosphere to pronounced lo-fi character.
| Parameter | Range | Description |
|---|---|---|
| Crackle | 0 – 1 | Noise intensity (0 = silent, 1 = heavy vinyl crackle) |
Analog
Console emulation with subtle stereo decorrelation, noise floor and soft saturation. Adds the imperfections that make analog gear feel alive.
| Parameter | Range | Description |
|---|---|---|
| Analog | 0 – 1 | Console emulation amount (0 = digital clean, 1 = full analog character) |
LFO 1 & LFO 2
Two independent modulation sources. Use the Modulation Matrix to route them to any of 17 destinations.
| Parameter | Range | Description |
|---|---|---|
| Rate | 0.1 – 20 Hz | Oscillation speed |
| Curve | 0 – 1 | Waveform shape (sine → triangle → square) |
| Skew | -1 to +1 | Waveform asymmetry |
| Smooth | 0 – 1 | Smoothing / randomness filter |
| Sync | On / Off | Lock to DAW tempo |
| Phase | 0 – 360° | Starting phase offset |
| Rand | 0 – 1 | Random variation amount |
Envelope Follower
Tracks the input signal's amplitude and converts it into a modulation signal. Use via the Modulation Matrix.
| Parameter | Range | Description |
|---|---|---|
| Attack | 0.1 – 100 ms | How fast the follower reacts to rising level |
| Release | 10 – 1000 ms | How fast it follows falling level |
| Sensitivity | 0 – 1 | Input sensitivity (0 = low, 1 = high) |
Modulation Matrix
4 routing slots. Each slot connects a modulation source to a destination parameter with a controllable amount.
Sources
None, LFO 1, LFO 2, Envelope Follower
Destinations (17)
Output Gain, Drive, Pre-EQ Focus, Post-EQ Sub (60 Hz), Post-EQ Low (170 Hz), Post-EQ Low-Mid (400 Hz), Post-EQ Mid (1 kHz), Post-EQ Hi-Mid (2.5 kHz), Post-EQ Presence (6 kHz), Post-EQ Air (14 kHz), Multiply Mix, Octaver Mix, Temporal Mix, Spatial Width, Spatial Pan, Character Body, Character Tone
Amount
Each slot has an Amount control from -1 to +1. Negative values invert the modulation.
Post-EQ
7-band graphic equalizer with fixed frequency bands for final tone sculpting.
| Band | Frequency | Range |
|---|---|---|
| Sub * | 60 Hz | -18 to +18 dB |
| Low * | 170 Hz | -18 to +18 dB |
| Low-Mid * | 400 Hz | -18 to +18 dB |
| Mid * | 1 kHz | -18 to +18 dB |
| Hi-Mid * | 2.5 kHz | -18 to +18 dB |
| Presence * | 6 kHz | -18 to +18 dB |
| Air * | 14 kHz | -18 to +18 dB |
Output
| Parameter | Range | Description |
|---|---|---|
| Output Gain * | -∞ to +12 dB | Final output level |
| Mix | 0 – 1 | Global dry/wet blend |
| Bypass | On / Off | Bypass all processing |
Presets
File format
Presets are stored as .slpreset files (JSON). There are two versions:
- v1 — Machine-created presets. Contains 90 parameter values.
- v2 — Magic/Material-aware presets. Additionally stores the instrument profile context (name, category, spectral analysis, knob mappings).
Both versions are fully backwards-compatible — v2 presets work in all contexts.
Storage location
~/Documents/Ferment Presets/
Ferment scans this folder at startup and displays all found presets in the preset browser. You can organize presets into subfolders — they will appear as categories.
Saving and loading
Use the preset browser in the Machine tab header. Click Save to store the current state, or browse and click a preset name to load it. Presets created via Magic automatically include the instrument profile context if a profile is active.
Sharing presets
.slpreset files are portable JSON. Share them with other Ferment users by
sending the file — they can place it in their presets folder and it will appear on next scan.
Instrument Profiles
An instrument profile captures your instrument's unique sonic fingerprint. Once created, it enhances both AI generation (Magic) and manual sound design (Machine).
What a profile contains
| Field | Description |
|---|---|
| Name | User-defined (e.g. "Adam's Voice", "1962 P-Bass") |
| Category | One of 15 types (Electric Guitar, Bass, Voice, Synth, etc.) |
| Play style | How you interact with the instrument — Auto, Bowed, Picked, Fingered, Struck, Sustained, Vocal or Percussion |
| Audio clips | Up to 8 recordings at different dynamics and playing styles |
| Spectral analysis | Brightness, Saturation, Harmonics, Centroid, Attack, Dynamic Range |
| Knob mappings | Per-parameter curve exponents and sweet spot ranges (adapted to play style) |
| Notes | Free-text field for personal notes |
How profiles affect the experience
In Machine tab
Each knob gains a visual "sweet spot" arc (green) showing the parameter range that works well for your instrument. An orange LED appears if you set a value outside the sweet spot. Parameter curves are remapped (log/exponential) so that the same knob movement produces musically appropriate results for your instrument.
In Magic tab
The AI receives your profile's spectral analysis alongside your text prompt and spider chart. This means a "warm delay" preset for an electric guitar will be different from a "warm delay" for a voice — the AI tailors the results to the instrument.
In presets
Version 2 presets store the profile context. When recalling a preset, Ferment shows which instrument it was designed for.
Profile limits
Ferment supports up to 24 profiles: 3 built-in demo profiles, up to 14 reference profiles for common instrument types and up to 7 user-created profiles. Profiles can be exported and imported for backup or sharing.