dings: Aqua's v4 set (Void single onset, Nibu proposal 3 everywhere), behind dings.enabled #75

Merged
void merged 1 commit from void/dings-v4 into dev 2026-11-10 17:59:58 +00:00
Member

Replaces the 32 sender clips under public/dings/senders/ with AquaSynth merge 21e83ec (page https://minds.gamecult.org/aqua/dings-voices-dd9be66/). manifest.json and neutral/ are unchanged byte for byte. Nibu stays on proposal 3 for all four events. config.dings.enabled stays off by default. README and the manifest test now say v4; the test pins all 32 clips by sha256 prefix.

Nobody has heard v4. The checks below are about the files, not the sound. This is not part of release 64edacac.

Byte check

I regenerated the set with the AquaSynth generator at 21e83ec (polyphony.test.mjs, build.mjs, finish.py, verify.py, in a venv with numpy/soundfile 1.2.2 libsndfile). Result against the staged page's out/:

  • manifest.json and the 4 neutral clips: byte-identical.
  • All 32 voice clips: same size, decoded PCM sample-for-sample equal. Raw bytes differ only in the Ogg stream serial and page CRC (bytes 14-17 and 22-25 of the first page), which the encoder randomises per run, so a bit-exact match to a fresh encode is not possible for anyone. The files committed here are the staged page's files, unmodified.
  • Nibu's four clips decode to the same samples as proposals/nibu3.

Tests

vitest: 48 files, 703 tests pass.

🤖 Generated with Claude Code

Replaces the 32 sender clips under `public/dings/senders/` with AquaSynth merge 21e83ec (page https://minds.gamecult.org/aqua/dings-voices-dd9be66/). `manifest.json` and `neutral/` are unchanged byte for byte. Nibu stays on proposal 3 for all four events. `config.dings.enabled` stays off by default. README and the manifest test now say v4; the test pins all 32 clips by sha256 prefix. **Nobody has heard v4.** The checks below are about the files, not the sound. This is not part of release 64edacac. ## Byte check I regenerated the set with the AquaSynth generator at 21e83ec (`polyphony.test.mjs`, `build.mjs`, `finish.py`, `verify.py`, in a venv with numpy/soundfile 1.2.2 libsndfile). Result against the staged page's `out/`: - `manifest.json` and the 4 neutral clips: byte-identical. - All 32 voice clips: same size, decoded PCM sample-for-sample equal. Raw bytes differ only in the Ogg stream serial and page CRC (bytes 14-17 and 22-25 of the first page), which the encoder randomises per run, so a bit-exact match to a fresh encode is not possible for anyone. The files committed here are the staged page's files, unmodified. - Nibu's four clips decode to the same samples as `proposals/nibu3`. ## Tests vitest: 48 files, 703 tests pass. 🤖 Generated with [Claude Code](https://claude.com/claude-code)
dings: Aqua's v4 set (Void single onset, Nibu proposal 3 everywhere), still behind dings.enabled
Some checks failed
Build pull request / Build pull request (pull_request) Failing after 2m8s
de9955a24a
Replaces the 32 sender clips with AquaSynth 21e83ec's staged render; manifest and neutral
fallback are byte-identical to the old ones. The manifest test now pins every clip by sha256
prefix. Unheard.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
void merged commit 0fe9ab2eee into dev 2026-11-10 17:59:58 +00:00
Sign in to join this conversation.
No reviewers
No labels
No milestone
No project
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
GameCult/Ratatoskr!75
No description provided.