Recording is the easy part, getting a usable file is not
Every operating system has a voice recorder buried in it somewhere, and most of them save into a format that something else then refuses to open. The web equivalents usually want an account, cap the length, or upload what you recorded so they can hand it back as a download.
This page does the recording and the conversion in the same tab. Your microphone is captured, held in memory, and written out as either an MP3 at a quality you choose or an uncompressed WAV. Nothing leaves the machine, there is no limit imposed, and the only thing that reaches your disk is the finished file.
Making a recording
- Choose Download as, either MP3 or WAV.
- If you chose MP3, pick an MP3 quality, ignored when WAV is selected level.
- Start recording and allow microphone access when the browser asks.
- Speak. The page shows that recording is in progress.
- Stop. A brief conversion runs, then the download is offered.
The file is named from the date and time you stopped, so a run of takes sorts itself chronologically without you naming anything.
Why there is a pause after stopping
Browsers do not record straight to MP3 or WAV. They record into their own working format, which is efficient to capture but is not what most people want to end up with.
So the conversion happens after you stop, and it is a real step rather than an instant one. The WAV path decodes the recording and rewrites it as plain uncompressed samples. The MP3 path runs it through an encoder at your chosen bitrate.
Knowing that explains the wait and it also explains why the quality setting cannot affect the capture. By the time the setting is applied, the recording already exists. The setting decides how it is written out, not how it was taken in.
Choosing between MP3 and WAV
The rule of thumb is about what happens next.
If the recording is finished when you download it, take MP3. A voice memo, a note to yourself, something to send to somebody. At 192 kbps it is transparent for speech and a small fraction of the size.
If the recording is raw material for something else, take WAV. Anything you are going to trim, level, denoise, transcribe or feed to another tool is better handled uncompressed, because each lossy encode in a chain is one the next stage has to work around. Convert to MP3 at the end instead, which WAV to MP3 does with a bitrate you choose.
The three MP3 qualities in practice
192 kbps is the default and the right answer nearly always. It is well past the point where speech through an ordinary microphone shows any difference.
320 kbps is worth choosing when the file is heading into further processing and you want to give the next stage as much to work with as possible, though WAV is the better answer for that case if size permits.
128 kbps is for when size genuinely dominates. A long meeting recording that has to fit an attachment limit is the honest use case, and speech survives it perfectly well.
What to do with the recording
If the level is too low, Normalize Audio or Boost MP3 Volume fixes it. If there is constant background hiss, Remove Audio Noise is the pass to run. If there are long gaps, Silence Remover tightens it up, and Cut MP3 trims the ends.
To turn the recording into text, Transcribe Audio runs a speech model on your own device. To check your microphone before recording anything important, Webcam and Microphone Test shows a live level meter. The rest are on the audio hub.