Revert VOX preset to the NetTTS approximation

The tuned values (pitch 0, rate 15, head size 34, lowering 50) matched
the reference metrics better but sound far off the game voice, so restore
the published NetTTS approximation: pitch 3, rate 17, prominences 22,
head size 52, lowering 40, volume 70. Also revert the test expectations,
the published default metadata, and the tuning notes in the docs; keep
tools/vox_compare.py for future experiments.
This commit is contained in:
Fisto
2026-09-22 19:56:22 +02:00
parent de4631fe54
commit c6b5fb71e8
5 changed files with 29 additions and 74 deletions
+3 -5
View File
@@ -93,11 +93,9 @@ A short low-level leading hum (80 ms, 75 Hz, -48 dBFS) primes Discord's audio pi
so the first word is not gated; tune with `VESKSLOP_TTS_HUM_MS` (0 = off),
`VESKSLOP_TTS_HUM_HZ`, and `VESKSLOP_TTS_HUM_DB`.
**Half-Life presets** (Keith Bell): `VESKSLOP_TTS_PRESET=vox` applies the Valve VOX
profile (pitch 0, rate 15, head size 34, lowering 50) and enables VOX prosody;
`VESKSLOP_TTS_PRESET=hev` is a starting point for the HEV suit voice. The `vox` and `hev`
voice IDs select these directly. The VOX values were
tuned against the game's word assets with `tools/vox_compare.py`; details in
**Half-Life presets** (Keith Bell): `VESKSLOP_TTS_PRESET=vox` applies the approximated
Valve VOX profile and enables VOX prosody; `VESKSLOP_TTS_PRESET=hev` is a starting point
for the HEV suit voice. The `vox` and `hev` voice IDs select these directly. Values are in
`docs/TTS-DESIGN.md`.
`VESKSLOP_TTS_CRUNCH` (`hl90`, `hl90hq`, `telephone`, `adpcm`, `crush`) adds 1990s lossy