Echo 1.1 - transcription actually works

Fixes the three bugs that made 1.0's transcripts fiction.

What was wrong. 3h12m of capture produced 22,218 words, almost none of them spoken: "the security is over" seven times in a row, and one line reading मार žе дусов लाहथ प buffalovia hurry up — Devanagari and Cyrillic in the same sentence.

Three causes, all fixed:

1. Whisper was fed the silence. A chunk that was 3% speech went in whole, and the model invents fluent text over room tone. Only voiced regions are now sent, with timestamps mapped back to real clock time.
2. Language auto-detection reads only the first 30 seconds and applies the result to the whole call — across 99 languages. A chunk opening with silence had its language chosen from noise. Detection is now restricted to en/hi/mr.
3. The JNI never enabled whisper's own defences: suppress_nst defaults to false upstream, and per-segment no-speech probability was never read.

Plus a per-language initial prompt (without it, Marathi came back romanised), level normalisation, and repetition collapse at both transcription and summary time.

Throughput. Base measured 714 s to transcribe a 256 s chunk — 0.4x realtime, a queue that never drains. The cause was the temperature-fallback ladder retrying every degenerate window up to six times; it is now capped at three.

Model guidance, measured rather than asserted:

model	Hindi	Marathi	code-switched	English
Base	0.23	0.00	0.08	1.00
Small	0.54	0.13	0.38	1.00

Small is the right choice for Devanagari, at roughly 2.5x the compute.

63 unit tests pass. See docs/VERIFICATION.md for what remains unverified.

Echo 1.0 - prototype

Sideloadable release build. Signed with the standard Android debug key so it installs directly; native code is compiled -O3.

asset	size	for
echo-1.0-arm64.apk	24.5 MB	phones — this is the one you want
echo-1.0-universal.apk	28.8 MB	also carries x86_64, for emulator testing

Verified on a Pixel 9, Android 17 (API 37): 42 JVM unit tests and 11 instrumented tests passing; the Base model sustains 5.29× realtime on 4 threads, putting a 10-minute chunk at ~113 s.

Install

adb install -r echo-1.0-arm64.apk

On Windows, scripts\verify-on-device.ps1 also grants RECORD_AUDIO/POST_NOTIFICATIONS and exempts the app from battery optimisation.

First run: grant mic + notifications, then Settings → Speech model → download Base (57 MB), then tap record.

Audio and transcripts never leave the device. Note that the app records people who have not consented — see the privacy section in the README before using it around others. Hindi/Marathi transcription is not yet verified end to end; see docs/VERIFICATION.md.