graphic_eq Online Voice Recorder star

How Recording Actually Works

Once record is pressed and microphone access is granted, the browser's MediaRecorder API takes over — the very piece our Microphone Test tool deliberately never touches. Instead of just reading live levels, MediaRecorder writes the incoming audio stream out in chunks, firing a dataavailable event periodically with each new piece delivered as a Blob object. When recording stops, all those chunks are concatenated into a single Blob representing the complete take, which is what gets handed to the player and the download link below.

Which container and codec that Blob uses depends entirely on the browser, not on this tool — no server-side conversion is involved at any point. Chrome and Firefox default to WebM containing Opus audio; Safari defaults to MP4 containing AAC. Both are well-supported, modern formats, so the file downloads exactly as the browser recorded it, with no re-encoding step in between.

What You Should Know

Opus is worth knowing about because it was purpose-built for exactly this kind of use case. It was created by combining Xiph.org's Vorbis codec with Skype's SILK codec and standardized by the IETF as RFC 6716 in 2012, specifically optimized for low-latency speech rather than general-purpose music streaming — it handles the pause-and-burst rhythm of spoken voice particularly well at small file sizes. AAC, Safari's default, is a more general-purpose codec that also performs solidly for voice, just from a different design lineage.

• Recordings live only in the page's memory until they are downloaded — closing the tab discards anything not yet saved to disk.
• An hour of speech is roughly 30–50 MB, so length is rarely a practical concern on a modern device.
• If a specific app insists on MP3, running the downloaded WebM or MP4 file through any converter works fine — both source formats carry enough quality for that conversion to be effectively lossless.