# Find a sound

`POST https://api.eesi.ai/v1/speech/transcriptions/{session_id}/sounds/search`

Find everywhere a sound happens: "applause", "a door slam", "a
phone ringing". Runs in the background; poll
`GET /v1/speech/transcriptions/{session_id}/sounds` for the spans.

Loud, common sounds are found reliably. For rare ones, check the score: in a
reverberant hall, "microphone feedback" can match the room itself.

Parameters:

- `session_id` (path, required)

Request body: application/json.

- `query` (string, required): The sound to find, described in words: "applause", "a door slam".

Responses: 202, 404, 422.

Schemas: https://docs.eesi.ai/api-reference/public.json