Decide which transcription conventions your project requires before editing the words. Use consistent labels for speakers, keep a route back to the audio and mark unclear passages instead of guessing. Transcribe Audio to Text provides a speech transcript to review; it does not automatically apply an academic transcription system or identify the speakers.
Which rules should I use?
Start with your project or course instructions. A readable record of the wording and a detailed analysis of pauses need different levels of notation. Write down your choices before editing the first interview. If several people transcribe the material, give them the same example and resolve differences early rather than changing conventions halfway through.
Should I remove hesitations and repeated words?
Keep them when the chosen method requires them or when they affect the meaning. Do not make a quotation sound more certain by silently deleting a hesitation. A separate reading copy can be edited differently, but label it accordingly. Clean read rewrites speech and is not the source for a verbatim research transcript.
What could a simple working example look like?
This invented example keeps a hesitation and marks an unclear word with a time noted from playback. Speaker labels and the uncertainty note are manual additions. Adapt this working convention to the rules required by your project.
Interviewer: When did you start?
Participant 1: In May, I think.
Participant 1: The first week was [unclear, 02:14].How do I review and deliver the transcript?
Tap the relevant words in Transcript to check them against playback. Export DOCX to add your project’s speaker labels and notation in a word processor. Keep an unchanged reference copy and record meaningful edits. Use the same label for each speaker throughout; a paragraph break alone does not show who spoke.
- Check the selected conventions and document identifier.
- Review names, numbers, negatives and every quotation.
- Mark unresolved words with a consistent note and time reference.
- Open the exported file and verify the layout required by your project.
Questions you may have
Does the app add speaker labels?
No. Listen to the recording and add labels yourself. A paragraph break does not establish that a different person is speaking.
What should a transcription convention sheet include?
Write down speaker labels, treatment of hesitations and pauses, an uncertainty marker and how you note playback times. Add a short example so everyone working on the project follows the same rules.
Should a pause always be written down?
That depends on your conventions and research question. Decide consistently; do not claim precise pause measurements from the speech transcript alone.
Can I replace the original with Clean read?
Keep the original wording and audio for research checks. A rewritten reading copy has a different purpose and may contain changed phrasing.