Audio Transcriptor
Start

Convert MP4 video to MP3 audio

Drop the video and keep only the sound. It happens in your browser: nothing is uploaded, nothing is watermarked, no account.

  1. 1File
  2. 2Settings
  3. 3Processing
  4. 4Download
  • MP4
  • MOV
  • MKV
  • WEBM
  • AVI
  • M4V

What this tool does

An MP4 is a box with two things inside: a video track and an audio track. Turning it into an MP3 is not really a conversion of the whole file — it's throwing away the picture, which is what takes up 95% of the space, and re-encoding what's left. That's why a 1 GB recording of a talk comes out as a 40 MB MP3 that sounds exactly the same.

The usual way to do this is to hand a lecture, an interview or a call recording to a website you know nothing about, wait for it to upload, and hope. Here the file never moves: the same ffmpeg those sites run on their servers runs inside your tab, which is why there's no queue, no size charged to us, and nothing for us to promise about your file.

What we handle that others don't

  • The video track is dropped, not converted. That's where the size goes, and dragging it along is what leaves you with an «MP3» some players refuse to open.
  • Nothing is uploaded, so there's no queue and no upload wait: on a normal laptop, an hour of video takes about a minute.
  • Files up to 500 MB. The video is read as it's needed instead of being copied into memory first, which is what usually kills these tools in a browser tab.
  • No watermark, no e-mail, no five-conversions-a-day. There's nothing to limit: it isn't costing us anything.

How it works

  1. 1

    Drop the MP4

    MOV, MKV, WEBM and AVI work too. It's read straight off your disk — nothing is uploaded.

  2. 2

    Pick the quality

    192 kbps is the sensible default. For a talk or an interview, 96 is plenty and weighs half.

  3. 3

    Download the MP3

    Ready for your player, your phone, or for uploading to be transcribed.

In detail

Why a one-gigabyte video comes out as forty megabytes

Because the picture is nearly all of the weight. In an ordinary MP4 the video track takes 90 to 95% of the file, and the sound sits in a separate track at 128 or 192 kbps that adds up to a little over one megabyte a minute. Keeping only the audio isn't compression working miracles: it's throwing away the big half.

The arithmetic is easy to do before you start: bitrate times duration. At 192 kbps that's 1.4 MB per minute, about 85 MB an hour; at 96 kbps, half of that. The size of the original video doesn't enter into it, so whether it was 300 MB or 4 GB, an hour-long talk lands on the same number.

How long it takes, and why

No server is doing this work: your own computer is, inside the tab, on a single core. A normal ffmpeg would use every core you have; the one that runs in a browser runs on one thread deliberately, because the multi-threaded build demands response headers that would break the payment form on the rest of the site. The result is roughly fifteen minutes of video processed per minute of waiting — an hour-and-a-half recording takes about five or six minutes.

The quality you pick barely moves that number: encoding at 320 kbps costs a hair more than at 96. What does move it is your machine, which is why the moment the bar starts, the time shown stops being an estimate and starts being measured on your hardware. Meanwhile the tab has to stay open — closing it stops the conversion the way closing a program does.

Questions

And if what you have is the video?

Upload it and you'll get the transcript with timestamps, the speakers separated by name and a summary with the key points. You can start without an account.

Transcribe my recordingSee prices

Related tools