bcox/mini-800
1
0
Fork 0
Code Issues 36 Pull requests Projects Releases Packages Wiki Activity Actions

A new instrument: Oberheim's Xpander and Matrix family #15

New issue
Closed
opened 2026-10-12 02:41:27 +00:00 by bcox · 2 comments
bcox commented 2026-10-12 02:41:27 +00:00
Owner
Copy link

Oberheim's matrix-modulation synthesizers as a new instrument: the Xpander, the Matrix-12, and the Matrix-6 and Matrix-1000. Not started.

Everything below is from recollection, with nothing read. Find and read the owner's and service manuals before designing anything, and put them in docs/references.md.

The instruments, as recalled

  • Xpander (1984). Six voices, no keyboard. Each voice: two VCOs (sawtooth, triangle, pulse), a filter with fifteen modes, two amplifiers in a row, five envelopes with a delay stage, five LFOs, four ramp generators, three tracking generators, a lag processor, and FM from VCO 2 to VCO 1 or to the filter. Modulation is a matrix: a number of free routes for each voice beside the fixed ones. Each voice can hold a different patch.
  • Matrix-12 (1985). Two Xpanders under a keyboard: twelve voices, with velocity, release velocity and pressure.
  • Matrix-6 (1986), its rack the Matrix-6R, and the Matrix-1000 (1987). Six voices of a simpler kind: two DCOs a voice, a four-pole low-pass alone, three envelopes, two LFOs, two ramps, one tracking generator, and a smaller matrix. The Matrix-1000 is the same voice with a thousand patches and almost no panel.

To settle with the owner first

  • One instrument or two. The Xpander and Matrix-12 are one voice; the Matrix-6 and Matrix-1000 are another, with DCOs where those have VCOs and a low-pass where those have fifteen modes. The Juno and the SEM family each made several machines one instrument. Whether the smaller voice is settings of the larger, as the OB-X is settings of a SEM, wants the manuals read.
  • Its name, and its programs' names. "Oberheim", "Xpander" and "Matrix-12" are trademarks.

What it needs that is not here yet

  • The modulation matrix: #7. This is the instrument, and it should be built in engine/ so that the SEM family can have it too. Do #7 first, or as the first part of this.
  • A filter with fifteen modes. They are understood to come from one four-pole filter by mixing its stages' outputs, which engine/four_pole_filter.h could offer: it has the four stages. Which mixtures, and what each mode is called, wants the service manual. Nothing may be taken from an emulator's code.
  • Ramp generators, tracking generators and a lag processor as modulation sources, which no instrument here has.
  • More LFOs and envelopes a voice than any instrument here: five of each. Their cost on the Pi 3B at twelve voices needs measuring early. They are digital on the hardware and slow; ours are read once a block.
  • Release velocity, which engine/midi.h does not pass on.

What is already here to build on

  • engine/analog_vco.h: a VCO giving sawtooth, pulse and triangle from one ramp, with sync and through-zero running for FM.
  • engine/envelope.h: now has a delay stage.
  • engine/ramp_oscillator.h, if the Matrix-6's DCOs are made as Roland's are.
  • The SEM family's sounds, where each voice plays one of several whole patches, and engine/zones.h: the Xpander's and Matrix-12's multi mode and zones are the same idea.
  • engine/variance.h, engine/portamento.h, the shared arpeggiator and effects.
  • instruments/README.md says what an instrument must provide.

Provenance

These machines have firmware and patch memory, which the SEM family did not.

  • Rule 3: no firmware disassembly and nothing extracted from a ROM. Their envelope and LFO timings and their curves come from the manuals or from measuring a real instrument, or are placeholders.
  • Rule 4: no factory patches. The Matrix-1000's thousand are Oberheim's. Reading a user's own SysEx dump at run time is allowed, and these instruments' SysEx is documented, so a numbered panel.h like the Juno-106's may be the right shape.
  • Rule 2: several emulations exist. None's code may be read for more than how a problem is structured, and nothing lifted.
Oberheim's matrix-modulation synthesizers as a new instrument: the Xpander, the Matrix-12, and the Matrix-6 and Matrix-1000. Not started. Everything below is from recollection, with nothing read. **Find and read the owner's and service manuals before designing anything**, and put them in `docs/references.md`. ## The instruments, as recalled - **Xpander** (1984). Six voices, no keyboard. Each voice: two VCOs (sawtooth, triangle, pulse), a filter with fifteen modes, two amplifiers in a row, five envelopes with a delay stage, five LFOs, four ramp generators, three tracking generators, a lag processor, and FM from VCO 2 to VCO 1 or to the filter. Modulation is a matrix: a number of free routes for each voice beside the fixed ones. Each voice can hold a different patch. - **Matrix-12** (1985). Two Xpanders under a keyboard: twelve voices, with velocity, release velocity and pressure. - **Matrix-6** (1986), its rack the **Matrix-6R**, and the **Matrix-1000** (1987). Six voices of a simpler kind: two DCOs a voice, a four-pole low-pass alone, three envelopes, two LFOs, two ramps, one tracking generator, and a smaller matrix. The Matrix-1000 is the same voice with a thousand patches and almost no panel. ## To settle with the owner first - **One instrument or two.** The Xpander and Matrix-12 are one voice; the Matrix-6 and Matrix-1000 are another, with DCOs where those have VCOs and a low-pass where those have fifteen modes. The Juno and the SEM family each made several machines one instrument. Whether the smaller voice is settings of the larger, as the OB-X is settings of a SEM, wants the manuals read. - **Its name**, and its programs' names. "Oberheim", "Xpander" and "Matrix-12" are trademarks. ## What it needs that is not here yet - **The modulation matrix: #7.** This is the instrument, and it should be built in `engine/` so that the SEM family can have it too. Do #7 first, or as the first part of this. - **A filter with fifteen modes.** They are understood to come from one four-pole filter by mixing its stages' outputs, which `engine/four_pole_filter.h` could offer: it has the four stages. Which mixtures, and what each mode is called, wants the service manual. Nothing may be taken from an emulator's code. - **Ramp generators, tracking generators and a lag processor** as modulation sources, which no instrument here has. - **More LFOs and envelopes a voice** than any instrument here: five of each. Their cost on the Pi 3B at twelve voices needs measuring early. They are digital on the hardware and slow; ours are read once a block. - **Release velocity**, which `engine/midi.h` does not pass on. ## What is already here to build on - `engine/analog_vco.h`: a VCO giving sawtooth, pulse and triangle from one ramp, with sync and through-zero running for FM. - `engine/envelope.h`: now has a delay stage. - `engine/ramp_oscillator.h`, if the Matrix-6's DCOs are made as Roland's are. - The SEM family's sounds, where each voice plays one of several whole patches, and `engine/zones.h`: the Xpander's and Matrix-12's multi mode and zones are the same idea. - `engine/variance.h`, `engine/portamento.h`, the shared arpeggiator and effects. - `instruments/README.md` says what an instrument must provide. ## Provenance These machines have firmware and patch memory, which the SEM family did not. - Rule 3: no firmware disassembly and nothing extracted from a ROM. Their envelope and LFO timings and their curves come from the manuals or from measuring a real instrument, or are placeholders. - Rule 4: no factory patches. The Matrix-1000's thousand are Oberheim's. Reading a user's own SysEx dump at run time is allowed, and these instruments' SysEx is documented, so a numbered `panel.h` like the Juno-106's may be the right shape. - Rule 2: several emulations exist. None's code may be read for more than how a problem is structured, and nothing lifted.
bcox commented 2026-10-17 04:13:13 +00:00
Author
Owner
Copy link

Settled by Ben on 2026-10-17:

  • One instrument, the Xpander, Matrix-12 and Matrix-6 (and 6R), named minixp.
  • The Matrix-1000 is not named. Its voice is the Matrix-6's, so it comes as settings; its presets are not wanted (and are out by rule 4 anyway).
  • Placeholders are fine for envelope, LFO and ramp timings and curves the manuals do not give.

Already in since this issue was written: the modulation matrix (#7), release velocity, envelope delay, sounds and zones. Left: read the manuals; a 15-mode filter in engine/; ramp, tracking and lag sources; five envelopes and five LFOs a voice; the Matrix-6 as settings; the instrument itself; an early Pi bench.

Settled by Ben on 2026-10-17: - **One instrument**, the Xpander, Matrix-12 and Matrix-6 (and 6R), named `minixp`. - **The Matrix-1000 is not named.** Its voice is the Matrix-6's, so it comes as settings; its presets are not wanted (and are out by rule 4 anyway). - **Placeholders are fine** for envelope, LFO and ramp timings and curves the manuals do not give. Already in since this issue was written: the modulation matrix (#7), release velocity, envelope delay, sounds and zones. Left: read the manuals; a 15-mode filter in `engine/`; ramp, tracking and lag sources; five envelopes and five LFOs a voice; the Matrix-6 as settings; the instrument itself; an early Pi bench.
bcox commented 2026-10-17 05:43:38 +00:00
Author
Owner
Copy link

Built as mini-XP (minixp, port 8818, instruments/xpander/): the Xpander, Matrix-12 and Matrix-6/6R as one instrument, with a 15-mode filter, ramp, tracking and lag generators, a quantized route curve and a squared amount law in engine/, and six zones for every zoned instrument. Ben played it on the Pi on 2026-10-17 and found it sounds great, and confirmed every choice made for it. Not done, and not to be ticketed: the Matrix-6 SysEx and numbered panel, stored multis, a MIDI channel per zone, the rear-panel inputs (#17), separate outputs. Closing.

Built as mini-XP (`minixp`, port 8818, `instruments/xpander/`): the Xpander, Matrix-12 and Matrix-6/6R as one instrument, with a 15-mode filter, ramp, tracking and lag generators, a quantized route curve and a squared amount law in `engine/`, and six zones for every zoned instrument. Ben played it on the Pi on 2026-10-17 and found it sounds great, and confirmed every choice made for it. Not done, and not to be ticketed: the Matrix-6 SysEx and numbered panel, stored multis, a MIDI channel per zone, the rear-panel inputs (#17), separate outputs. Closing.
bcox closed this issue 2026-10-17 05:43:38 +00:00
Sign in to join this conversation.
No Branch/Tag specified
Branches Tags
main
No results found.
Labels
Clear labels
No items
No labels
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
bcox/mini-800#15
No description provided.