AU: voice picker in the header #7

Merged

peterswimm merged 1 commit from feat/au-voice-picker into main

2026-10-14 23:33:24 +00:00

peterswimm commented

2026-10-14 23:31:38 +00:00

Owner

Summary

The AU's header gets a menu for choosing a voice, so you can pick a HEXAWE voice without going through the host's preset menu.

The menu offers Init (the wavetable engine) first. Below that is one submenu per release, or per song for albums, listing that release's instruments.
Placement: the menu sits in the header beside the UDG pages rather than on them, so udg_parity is unchanged.
Staying in step with the host: voicePath is now KVO-observable, so the header updates when the host changes presets or restores state. A restore also sets currentPreset to the factory preset it matches, with a KVO notification.
Voice data: O1NKVoicePreset now stores the release group and the instrument separately, and the preset name is built from them. The names are unchanged.

Testing

tests/lgpt/run.sh: all checks pass. au_check now also covers three things:
- every voice has a release group and an instrument
- a restore reports the matching factory preset
- choosing a preset notifies voicePath observers
udg_parity: passes.
Build: xcodebuild for the iOS Simulator succeeds.
Not tested visually: the o1nk host app doesn't show the AU's view, so I haven't seen the menu rendered. It needs a look in an AU host such as AUM or GarageBand.

## Summary The AU's header gets a menu for choosing a voice, so you can pick a HEXAWE voice without going through the host's preset menu. - **The menu** offers Init (the wavetable engine) first. Below that is one submenu per release, or per song for albums, listing that release's instruments. - **Placement:** the menu sits in the header beside the UDG pages rather than on them, so `udg_parity` is unchanged. - **Staying in step with the host:** `voicePath` is now KVO-observable, so the header updates when the host changes presets or restores state. A restore also sets `currentPreset` to the factory preset it matches, with a KVO notification. - **Voice data:** `O1NKVoicePreset` now stores the release group and the instrument separately, and the preset name is built from them. The names are unchanged. ## Testing - **`tests/lgpt/run.sh`:** all checks pass. `au_check` now also covers three things: - every voice has a release group and an instrument - a restore reports the matching factory preset - choosing a preset notifies `voicePath` observers - **`udg_parity`:** passes. - **Build:** `xcodebuild` for the iOS Simulator succeeds. - **Not tested visually:** the o1nk host app doesn't show the AU's view, so I haven't seen the menu rendered. It needs a look in an AU host such as AUM or GarageBand.

peterswimm added 1 commit

2026-10-14 23:31:38 +00:00

au: voice picker in the header d1b5582313

The AU view's header gets a menu of the HEXAWE voices: Init (the
wavetable engine), then a submenu per release (per song for albums)
listing its instruments. It sits beside the UDG pages, not on them, so
udg_parity is unchanged.

The header follows the AU: voicePath is now KVO-observable, so host
preset changes and state restores update it, and a restore also sets
currentPreset to the factory preset it matches. Voices carry their
release group and instrument separately; the preset name is built from
them as before.

Co-Authored-By: Claude <noreply@anthropic.com>

peterswimm merged commit a7fa73f3a6 into main

2026-10-14 23:33:24 +00:00

peterswimm deleted branch feat/au-voice-picker

2026-10-14 23:33:24 +00:00

peterswimm referenced this pull request from a commit

2026-10-14 23:33:25 +00:00

Merge pull request 'AU: voice picker in the header' (#7) from feat/au-voice-picker into main