The capture SDK, running for real.
This is the same @attunio/capture module you'd drop into a Zoom, Doxy.me, or Daily session — pointed at your microphone instead of a video stream. Speech rate and response latency are computed on-device; the result comes back from the live API.
Run it live
This page uses your microphone, extracts telemetry in your browser, and POSTs only the derived numbers to the live /v1/dmse/analyze endpoint. No audio is uploaded or stored.
Create one under Publishable keysin the dashboard and add this page's origin to its allowed origins.
Video-derived signals aren't extracted in the browser. Supply your own here to see them merged into the same call.
On-device telemetry
Speech rate
—
Response latency
—
Speaking time
—
Session
—
// Click Analyze to call the live dMSE API
// and see structured Digital MSE output here.Only the four numbers above are transmitted. Audio stays in your browser.