System Audio Capture

Status: Draft Author: Ryan Nel Date: 2026-05-01

Problem

The Meeting Recording feature captures audio from the user's microphone only. For users in remote meetings — Zoom, Teams, Google Meet — the other participants' audio plays through speakers or headphones but is not captured by the microphone input. The user hears the full conversation; WordLoop hears only one side. This makes the transcription incomplete and the extracted insights unreliable for the most common meeting format: remote calls.

Capturing system audio requires getDisplayMedia({ audio: true }), which uses a fundamentally different browser permission model (display media, not user media), a different audio routing pipeline (system audio loopback vs. microphone input), and potentially mixing two audio sources (mic + system) into a single stream.

Appetite

3–4 weeks. The browser APIs are well-documented but the permission UX differs across browsers and operating systems. The audio mixing pipeline (mic + system) needs careful handling to avoid echo and gain issues. This is the natural next capability after mic-only recording and the one users will expect most urgently.

Why Now

Remote meetings are the dominant use case. Mic-only recording covers in-person conversations and the user's own voice, but the highest-value scenario — capturing a full remote meeting — requires system audio. This should be the next recording capability after the initial mic-only release.