v0.3.0

Added

• Linux: the Standalone app, VST3 and LV2 for x86-64, in a .tar.gz with an install script that installs for your account or system-wide and lets you leave parts out. LV2 hosts list the 128 slots as presets; picking one loads that slot from your library, though the names in the host's list are placeholders. Syncing with an FM-1 on Linux is untested with the hardware.
• Hardware character (Settings tab, saved with the project): makes the plugin sound like the FM-1's USB audio, with 16-bit output at the FM-1's level and its slightly softer high end, both measured from the hardware. An FM-1 volume control sets how far below full the synth's volume is, for the grain its output has at lower volumes. Off by default.
• MIDI files for the sequencer: export this pattern or all 16 (one track each) as a standard MIDI file with the notes as they play, including ratchets, accents, ties and transpose; import a MIDI file's notes into the selected pattern on its note-value grid, with held notes tied.
• Settings tab: pitch-bend range, the MIDI channel the plugin plays from the host, and a fixed velocity for the on-screen keyboard (and step recording with it). Changes take effect at once and are saved with the project; Save as default makes them what new instances start with, and Revert to default goes back to them.
• Reads the FM-1's GLOBE settings when it connects (FM-1_093): its MIDI and FX channels, pitch-bend range, key velocity, glide, drive, CC7 volume and overdub setting, shown on the Settings tab. The plugin talks to the FM-1 on its channels, so the FX channel no longer has to be set by hand. Copy from FM-1 takes its bend range and key velocity into the plugin's settings.

Changed

• The pitch-bend range defaults to ±12 semitones, the FM-1's default; it was ±2.

Fixed

• On FM-1_093, Show on FM-1, Store to FM-1 and Send to FM-1 select the preset on the synth's own MIDI channel, so they work when the FM-1 listens on a channel other than 1.

v0.2.0

Added

• JSON presets and patterns. A readable, editable format for backups and for writing sounds and patterns by hand or with a script or AI. A file can hold one preset, a group, a bank, all 128, one pattern, all 16, or everything. Each preset carries its raw bytes too, so a file read back unchanged gives exactly the same sound and a hand edit changes only what was edited. Described in docs/JSON-FORMAT.md, with a JSON Schema that editors can check files against and an example file.
• Export menu on the Library & Sync tab: the selected presets, the current bank, all 128, or everything (presets and patterns) as JSON; the selected presets or all 128 as .syx.
• Select several presets in the list with Shift-click or Cmd-click (Ctrl-click on Windows) to export them together. A plain click still picks the preset to play.
• Import and Export on the Sequencer tab: one pattern or all 16 as JSON, with every step setting, including ratchet, chance, accent, ties, transpose and Chain.
• Import asks where presets go when a file says which slots they came from: back into their own slots, or in from the selected slot.

Changed

• Import... on the Library & Sync tab now takes .json files as well as .syx.
• Exporting the preset being edited saves it with its unsaved changes, in both formats.
• Virtual Analog presets are exported without FM settings, which the VA engine does not use; their raw bytes carry the whole sound.
• macOS installer: no "How do you want to install this software?" step; everything installs on the startup disk.
• macOS installer: the license reads as normal paragraphs instead of breaking lines in odd places.

v0.1.0

Full Changelog: https://github.com/jbschooley/Virtual-FM-1/commits/v0.1.0