Home›Articles›GR!MM: dark-ambient firmware for the M-VAVE FM-1, built…
Articles

GR!MM: dark-ambient firmware for the M-VAVE FM-1, built on Dinghy, Felucca and SLOOP

GR!MM firmware turns the M-VAVE FM-1 into a four-track dark ambient machine. A look at its sound sources, tools, Apetactics collab and limits.

Bernardo de Azevedo
Bernardo de Azevedo
· Updated
GR!MM firmware: M-VAVE FM-1 pocket synthesizer beside the Hartmann Neuron

The GR!MM firmware, published on GitHub by the account GRiMMBIENT, turns the M-VAVE FM-1 into a four-track drone and atmosphere machine aimed at dark ambient, dungeon synth and what its author calls deathdream. It is built on Dinghy, Felucca and SLOOP, and, according to a single MATRIXSYNTH post, a later collaboration with the Swiss artist Apetactics produced a MIDI channel filter. This article describes what the firmware offers, what its documentation does and does not confirm, and whether a pocket FM synth with shared voices suits slow, textural work.

Table of Contents

What the GR!MM firmware is

The project’s repository describes GR!MM as unofficial, open-source firmware for the FM-1, and it states plainly that M-VAVE neither made nor endorses it. The page does not name an individual author. It is owned by the GRiMMBIENT account, and MATRIXSYNTH’s coverage credits the firmware’s own video to “GR!MM // Deathdream”.

The first public release of the GR!MM firmware, tagged v1.0, appeared on 17 October according to the release page, with no year shown. The manual and the community cheat sheet both cover version 1.0, so that is the only version this article can describe. The firmware is free and licensed under GPL-3.0, while the four bundled factory samples are released under CC0.

The central design decision is the track structure. Four tracks run side by side, each with its own sound and its own 64-step sequence, and all four draw on a pool of six voices together with shared effects. The community cheat sheet on the Firmware Hub uses the same description of a drone and atmosphere machine as the repository.

Sound sources, samples and effects

According to the repository, the firmware offers seven sound sources: a virtual analog engine with a supersaw, six-operator FM, a sample player, a formant-based choir called VOICE, a tonewheel organ called WHEEL, physical models, and phase distortion. The FM-1’s own six-operator FM is therefore one of seven options rather than the centre of the instrument, which is a notable departure from the hardware’s original purpose.

The sound library of the GR!MM firmware is modest and clearly organised. There are 40 factory sounds in eight categories and 64 user slots. The four factory samples, named URCHIN, LEAVE, NIKOLAI and M-SOFT, last up to 6.6 seconds each and sit in replaceable slots. The manual adds that empty sample slots are filled with these on first start, while any slot already in use is left alone.

Four effects shape the output. TAPE and DELAY handle the worn, repeating textures, GRAIN includes a freeze function, and SPACE is a reverb that the repository says reaches about 32 seconds, with an infinite setting beyond that. MATRIXSYNTH, relaying the maker’s video description, mentions a warm tape warble on user-loaded WAV files.

Each track also has an envelope, an LFO and a filter, and the keys can snap to a chosen scale. Those are standard parts of any synth, but on a 27-key device with a few encoders they matter because they decide how much shaping can be done without a computer.

Sequencing, song mode and the performance layer

The sequencer runs 64 steps per track and offers chance, ratchets, note lengths, step locks and chord steps. The cheat sheet adds copy, paste, randomise and mutate tools. For dark ambient these are useful mainly for probability: a pattern that drops and returns steps behaves less like a loop and more like a slowly shifting texture.

There is also an arpeggiator and a one-key chord mode. A latch mode keeps a note sounding until its key is pressed again, which is what makes a drone practical on a small keyboard, since the hands are free for the knobs.

Song mode has four slots, labelled A to D in the repository. The manual describes chaining up to 16 sections and offers mute and solo controls. The manual lists four project slots, and says the firmware autosaves once the sequencer has been stopped and idle for 2.5 seconds, so a sketch survives a power cycle.

The PERFORM page collects momentary tricks such as freeze, throws, tape stop, reverse and stutter. The manual also covers MIDI learn for up to 32 controller links, assignment of MIDI channels and sending sequencer and arpeggiator notes out over USB. The release notes list MIDI and USB audio support.

Foundations and credits

GR!MM credits its sources carefully. Felucca supplies the platform and several synthesis engines, and our Felucca article covers that foundation in detail. Dinghy, by zvenson, is a deliberately small starting point: its README describes about 300 lines of sine-wave code on top of finished plumbing for updates, USB, panel and screen. GR!MM grows from that template.

SLOOP shares platform work with both Felucca and Dinghy, and the GR!MM repository says SLOOP’s editor inspired its own. The tape delay design and the approach to browser builds come from X0X by Charles Vestal, whose firmwares have their own page on this site. The repository also credits msfa, DaisySP and Mutable Instruments Rings, among others.

Seen against these relatives, the GR!MM firmware is narrower. SLOOP is a live groovebox, and Felucca is a general multi-engine platform, whereas GR!MM commits to one mood and builds the interface, the sample set and the effects around it. That focus is the main reason to choose it over a broader firmware, and also the main limit.

The Apetactics collaboration

The collaboration between Apetactics and the GR!MM firmware’s author rests on a single source, the MATRIXSYNTH post of 21 October, which relays the descriptions of three videos. In that account, Apetactics, identified as Damian, an ambient IDM artist from Switzerland, had posted a video of the firmware running in the web emulator, and GR!MM reached out to him afterwards. He needed a MIDI channel filter that limits the device to one channel. The maker is said to have had the change ready the same evening, and the official collaboration followed a day later.

One of the videos carries a title naming a collaboration with a firmware maker and is credited to both Apetactics and GR!MM // Deathdream. The repository, the manual and the cheat sheet do not mention Apetactics, and none of them describes a MIDI filter, although the manual documents MIDI channel assignment. The change cannot be tied to a specific release, because the release page lists only v1.0. Readers should treat the story as the makers’ own account.

Installing and using the browser tools

The repository recommends installing through the FM-1 Firmware Hub in Chrome or Edge. As a manual alternative, the release provides a firmware file that a Python script installs. The page states plainly that installation is at the user’s own risk, and the hub offers a Rescue page for a stuck device. Sounds, slots, settings and samples are kept across updates, according to the release notes.

Anyone planning to flash it should read the project’s warnings first and follow the site’s safe install and recovery guide instead of improvising. This article deliberately gives no flashing steps.

Two browser tools let a prospective user try the GR!MM firmware before committing. A live demo runs the same firmware source compiled to WebAssembly, so the interface and behaviour can be tried without hardware. A separate editor connects to the FM-1 over USB for live sound editing, sample loading and backup and restore. The manual says the editor needs Chrome or Edge, and that the FM-1 writes to memory only while the sequencer is stopped and silent.

Strengths, limits and open questions

The strengths of the GR!MM firmware are practical. The documentation is thorough for a first release, with a manual, a cheat sheet and licensing files. Persistence is well handled, the demo is free to try, and the credit trail is clear. A user who dislikes the result can return to official firmware, which the hub links to.

The limits are also visible. The repository was small at the time of writing, with four commits, three open issues and 16 stars, and only one release. Those numbers will change and say little about quality, but they do mean little community experience is on record. The shared pool of six voices across four tracks means that dense chords on several tracks compete for voices: according to the manual, a new note beyond six takes the oldest one, which is faded out. Samples are capped at 6.6 seconds.

The documentation has one inconsistency. The repository, manual, release notes and cheat sheet all say seven sound sources, while the first video description in the MATRIXSYNTH post says six and the third, from the channel substatica drones, says seven. Seven is the better-supported figure.

This article is based on the project’s documentation and the makers’ own descriptions, and the firmware has not been tested on hardware. No independent review had been published at the time of writing. The manual also lists many pages and hold modes, which is normal for a one-screen device but means a learning curve for a first-time user.

Final perspective

GR!MM firmware is a clear example of the FM-1 community moving from general platforms towards single-purpose instruments. Its seven sound sources, long reverb, freeze function and latched chords address the needs of drone and dungeon-synth work directly, and the browser demo makes evaluation cheap.

The caveats are those of any new project: one release, a small repository, and a collaboration story known only through the makers’ videos. For listeners who want atmosphere rather than beats, it is worth trying in the browser first. Those who need a flexible groovebox will be better served by SLOOP or Felucca.

Sources

  1. GRiMMBIENT/grimm: GR!MM firmware for the M-VAVE FM-1, GitHub
  2. GR!MM releases (v1.0), GitHub
  3. GR!MM manual, GitHub
  4. GR!MM cheat sheet, The Firmware Hub
  5. The M-VAVE FM-1 firmware for dark ambient (post of 21 October 2026), MATRIXSYNTH
  6. GR!MM browser demo, GRiMMBIENT
  7. Dinghy, GitHub
  8. Felucca, GitHub
  9. SLOOP for the FM-1, GitHub
  10. FM-1 Firmware Hub, The Firmware Hub
Bernardo de Azevedo
Written by
Bernardo de AzevedoFounder & Editor

His connection with music began at age 6, in the 1980s, when his father introduced him to Jean-Michel Jarre's Rendez-Vous on vinyl. He works professionally in the legal field, while synthesizers became his space for abstraction and creative exploration. Founder of The Synth Source.