Independent journalists operate under a constant tension between the richness of the story and the time available to capture it. A 45-minute interview contains far more usable material than any notebook can preserve - and the most powerful quotes rarely announce themselves as quotable in advance.
Transcription as part of the workflow
Traditional transcription - listen, rewind, type, correct - takes three to four times the duration of the original interview. For a freelance journalist filing to multiple outlets, that's hours that should be spent writing, not re-listening. With Listen, the transcript arrives automatically. The diarisation separates your questions from the source's answers, producing a document that maps directly onto a standard Q&A or narrative article structure. For multi-speaker formats like panel discussions, see our article on roundtables and panels.
Choosing the right engine
Listen offers two transcription engines: Whisper (OpenAI) and Gladia. For interviews conducted in standard British or American English, Whisper performs excellently. For sources with strong regional accents - think a Scottish councillor, a Jamaican activist, a Welsh academic - Gladia's accent modelling is demonstrably superior. See the full comparison in our article Whisper vs Gladia.
Field tip: position the iPhone between yourself and your source, microphone facing up. This balances the capture of both voices and produces a cleaner diarisation output.
Read the Listen guide before your first recorded interview.