Eight operators wired freely — no fixed algorithms
Any operator can modulate any other. Every connection carries its own depth and phase, and an operator can feed back into itself.
The price is the only small thing about it. Wavetable and FM both. Eight operators, three filter stages and the output all sit on one screen. You connect them by dragging an arrow from one to the next. An operator can be a sine, noise or a wavetable, and your DX7 banks load as they are. The UI runs in nine languages.
One-time purchase · 3 devices · offline activation
Every clip is a preset the author built, played exactly as it is saved. Only the loudness is matched between them.
The phrase is chosen to suit the patch. The rhythm clip plays two drum presets together.
The synth below is not a recording and not a re-implementation. It is the same C++ DSP that ships in the VST3, compiled to WebAssembly and running as you play it.
Start the audio engine and the keyboard below becomes playable.
Nothing loads until you start it — the module is about 400 KB.
Whatever the demo left you wondering about, the person who built it explains — on video and in writing. Both are first-hand, not copy written by somebody else.
A guided tour through the operators, the routing, the filters, the effects and the output stage, driving the real plugin.
The narration is in Japanese. Turn on YouTube's auto-generated subtitles and auto-translate them (Settings → Subtitles) to follow it in English.
The written introduction: why it was built and what each part does, with screenshots. Easier to skim than the video.
Written in Japanese; note's own translation feature renders it in English.
Any operator can modulate any other. Every connection carries its own depth and phase, and an operator can feed back into itself.
An operator can play a wavetable instead of a sine. The scan position is modulatable, and Serum- or Vital-style WAV tables load directly.
Beyond the ADSR knobs, you can draw the envelope itself: as many points as you like, a curve per segment, and your own sustain and loop points.
Six LFO shapes including sample-and-hold, with tempo sync, note divisions, fade-in and note-on phase reset. An ADSR can drive a slot too.
Each stage has its own type, cutoff, resonance and slope, plus key tracking, drive and character.
Operators, filters and the output share one screen, and you connect them by dragging an arrow from an output port to an input port. FM modulation, sends into the filters, and a filter's output going back into an operator's phase input are all drawn in the same place.
Drive, chorus, delay and reverb, running at the oversampled rate ahead of the output protection stage.
Hand it a .syx and it takes in all 32 voices — ratios, fixed frequencies, key scaling, rate scaling, LFO and the sensitivities — as ordinary editable patches.
Presets are in the works. They will arrive in the next version.
You can start from Init and go from there. Your own patches save and export, and DX7 banks (.syx) import.
The plugin's interface runs in Japanese, Chinese (Simplified and Traditional), Korean, English, German, French, Spanish and Russian. The demo above is already running in yours.
Descriptive labels — parameter names, waveform and noise names — are translated. Conventional technical notation such as LPF or dB/oct is left alone. You can switch at any time from the wheel inside the UI.
VST3, CLAP · Windows 10 or later, 64-bit
VST3, AU, CLAP · macOS 10.13 or later, Intel and Apple silicon
The Windows installer is currently unsigned. SmartScreen may warn the first time you run it — choose More info, then Run anyway.
This is a limited-time beta campaign price. Feedback and bug reports are very welcome — write to satodeyannsu@gmail.com. Abuse just goes in the spam folder unread, so feel free to be blunt.