WAV to text

WAV is uncompressed, which is great for quality and hard on uploads. Transcribe WAV files free in your browser without uploading anything, or fast when you need speaker labels.

Why WAV files are different

WAV stores audio uncompressed, so it keeps every detail and takes a lot of space. An hour of CD-quality stereo WAV (44.1 kHz, 16-bit) is about 635 MB, several times larger than the same hour as an MP3. That is why WAV is common with audio interfaces, field recorders and studio software, and also why uploading WAV files can be slow on an ordinary connection.

Free, with no upload at all

The browser lane in Skeleton Fingers reads the WAV file on your own computer and transcribes it there with the Whisper speech model. The file is never uploaded, so its size does not turn into upload time, and your audio never leaves your device. It is free, unlimited and needs no account.

What you trade for that: transcription runs at the speed of your computer, so it is slower than the fast lane. The first run downloads the speech model once. The browser lane does not label speakers.

Fast, with speaker labels

On the fast lane, your browser converts the file to audio, uploads that audio (about 115 MB per hour), and our servers transcribe it in seconds. The first two steps depend on your device and connection, so the total time varies. It labels speakers and detects the spoken language automatically, which is useful for interviews and meetings captured as WAV. It needs a free account, and your recording is uploaded and stored with that account until you delete that transcript or your account.

How to transcribe a WAV file

1. Open Skeleton Fingers in your browser.
2. Choose "In your browser" or "Fast".
3. Drop in the WAV file. No need to convert it to MP3 first.
4. Review the text and check names and numbers against the audio.

Getting the most out of a WAV recording

• Good WAV audio helps, but recording matters most. An uncompressed file of a distant, noisy room still sounds distant and noisy.
• Trim long silences. Field recordings often start and end with minutes of nothing. On the fast lane, minutes are counted by the length of the audio, so trimming saves minutes.
• Multitrack sessions: if your software can bounce a single mixed file of the voices, transcribe that mix rather than one track at a time.

What it costs

The browser lane is always free. On the fast lane, a free account transcribes files up to 30 minutes free, 3 a day. Longer files, or more in a day, use pack hours: an hour pack is $5 for 3 fast hours, paid once, with hours that never expire, or Pro is $9.99 a month for 10 fast hours. No subscription is required to get started.