MP3 to text for podcasts, recorders and voice memos
Why MP3 to text works well for speech
MP3 is one of the most common audio formats, so almost every podcast host, recorder and editing app can save one. It keeps speech clear at small file sizes, which makes uploads light. You can transcribe MP3 files from any of them.
Turning an MP3 audio file to text takes one upload, and you go from MP3 to transcription with nothing to install. WAV, AAC, OGG, FLAC and the other common formats work the same way.
M4A to text from an iPhone
To transcribe M4A files from Voice Memos, open the memo, tap the More button and choose Share or Save to Files. Then upload the file from your phone's browser. The M4A goes up as it is.
Getting the MP3 off your device
A handheld recorder keeps its MP3s in a folder you can open once it is plugged into a computer by USB. On a podcast host, the episode page in your dashboard usually lets you download the file you published. Drop the file in with its original name, since the name has no effect on the text.
Does bitrate matter?
For speech, bitrate matters less than the recording itself. A low-bitrate MP3 of a clear speaker usually transcribes cleanly, while a high-bitrate file with music under the voice can lose words.
What hurts most is noise, clipping and distance from the microphone. Clipping is the harsh distortion you hear when a voice was recorded too loud. Upload the MP3 as it is, because converting it to WAV first adds file size and no detail.
Long MP3 files
A free account transcribes the first 5 minutes of each MP3 and marks where the locked part begins. Unlimited takes files up to 10 hours, so a full episode or a long talk goes through as one file. It also runs 3 files at once, which helps when you upload a season of episodes.
Podcast show notes from the transcript
A podcast transcript gives you more than show notes. Pull quotes for social posts, write chapter titles from the timestamps and publish the full text on the episode page for listeners who prefer to read.
Interview recordings for research and articles
Researchers who record qualitative interviews often code the text in NVivo, MAXQDA or ATLAS.ti. All three import Word files, so a DOCX export from Unlimited goes straight into your project. Before that, mark who speaks. Use the timestamps to locate where each turn begins, then type initials such as I and P at the start of it in the editor.
Journalists can check every quote against the recording. Jump to the timestamp in your own copy, listen to the sentence once more, then copy it.
Music, lyrics and spoken word
Kotosay is made for speech. Songs with vocals over loud instruments come out patchy, because the music covers the words. In a podcast with an intro jingle, the spoken parts transcribe normally.
Phone call recordings
Call recorders often save MP3 or M4A, and both upload directly. Phone lines carry a narrower band of sound than a microphone in the room, so expect a few more names and numbers to fix in the editor. Laws on recording calls differ by country and state, so ask before you record.
From MP3 to a finished document
- Fix names and terms in the editor. The timestamps stay where they are.
- Copy the text into show notes or an article on any plan.
- Download TXT, DOCX or PDF with Unlimited.
- Export SRT or VTT if the MP3 belongs to a video.
TipFor a podcast, note the timestamp of each topic while you read. Those times make quick chapter marks for your show notes.