mavsttd/whisper_handler.go: name the no_speech_prob confidence-zeroing
floor (0.9) and the whisper thread count (4), both previously bare
literals with no reason attached.
mavttsd/piper_handler.go: name piper's render rate (22050) and the
canonical wire rate (16000) used by the resampler, instead of repeating
the two numbers inline four times.
mavenclient/main.go: remove the strconv/io/net/time imports and their
`var _ = ...` keep-alive lines — dead weight with no caller, not future
scaffolding.
Behaviour-preserving; no test changed. go test -race ./internal/...
./cmd/... is green.
The file comment named four regressions caught here. Three were not.
Nothing on this path resamples, because PCMFromWAV refuses anything that
is not already 16 kHz mono s16. Nothing exercises language selection,
because the hint comes out of the manifest already correct. And a bad
model path was the one condition that made the whole test vanish behind
a skip nobody reads. The comment now claims the two things that are
real, an explicitly set MAVEN_WHISPER_MODEL that does not exist is a
failure, and a missing fixture is a failure rather than a skip.
looseWordMatch accepted a different word. Four retained runes of "воды"
is "вод", so whisper hearing "выпил водки" satisfied the ru_fact
keyword, and "dis" let display, distance and discuss all stand in for
"disk". A case ending adds a rune, not a syllable, so the hypothesis is
capped in length as well as matched on prefix.
The spoken text lived in the generator and in the manifest with nothing
tying them together. Editing one left the other describing audio that no
longer existed, and at a flat ceiling of 0.34 over a five-word reference
a one-word drift passed silently. The script reads text out of the
manifest now, and the ceilings are set just above what each case really
measures against ggml-small, with the measurement recorded beside them.
Also: the test carried its own copy of the PCM to float32 conversion, so
a regression in the daemon's copy left the silence-gate assertion green,
and the manifest was validated for keywords but not for text, where an
empty reference makes every hypothesis score a WER of 1.
Found in review of #75.
Whisper hallucinates subtitle-credit boilerplate ("Редактор субтитров …") on
silence/room-noise, which then got stored as tap:voice facts. Gate before the
model: drop clips shorter than -min-ms (default 300) or below -silence-rms
(default 0.01 normalized RMS). Both are flags — the mic floor is hardware
specific. Returns empty transcript (same as whisper's no-segments path), so
nothing downstream changes.
gateReason is pure and unit-tested (silence/short/quiet → dropped, loud+long →
passes). ponytail: energy gate, not a real VAD; upgrade to WebRTC VAD or
whisper no_speech_prob if too blunt.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- resolveQuietToggle runs in HandlePushToTalk before the router so
'тихий режим' works regardless of classifier confidence.
- whisper_full() runs in a goroutine with ctx.Done() select so the
handler returns promptly on timeout/shutdown.
- StubReplier.IntentQuery no longer claims query is unimplemented.
claude
kami