Both audio and transcripts are quite routinely cleaned up for broadcast and publication. Enough so on NPR / public radio that it's a signature. With some exceptions, but those are exceptions:
It’s no secret that pre-taped interviews on public radio are edited, sometimes considerably. Short silences and stumbles are “cleaned up.” Digressions or parenthetical remarks within an answer are removed in what are called “internal edits.” Often entire blocks of Q&A are excised. The cuts are made to bring the conversation down to time and present it in its most compelling form....
“Cleaning up” tape. This refers to removing ums and ahs, as well as stumbles and pauses, from an answer. Most of the time this is fine, as long as you are not removing something that suggested the interviewee was reluctant or unable to respond to a question....
My exposure to the darkroom began when I edited my first story. I sat at the digital editing console with a producer, listening to an interview with a source I had recorded earlier. With just strokes of the keyboard, he cleaned up and tightened the sound bites I was going to use, taking out sentences, words and even some of the pauses, making what are called "internal edits." Then the various thoughts were woven together technically in a way that would be totally hidden to the listener.
An intelligent verbatim transcript attempts to capture what was said rather than how it was said. The main priority being content and “voice”, while leaving out repeated words and fillers.
The aim of an intelligent verbatim interview transcript has less to do with depicting accents or involuntary vocalization. It has more to do with accuracy of the substance of the research interview, considering the meanings and perceptions created and shared during a conversation. And this is why this style of transcription is so popular, because it provides a more readable transcript while staying true to the voice and intended meaning of the participants.
It’s no secret that pre-taped interviews on public radio are edited, sometimes considerably. Short silences and stumbles are “cleaned up.” Digressions or parenthetical remarks within an answer are removed in what are called “internal edits.” Often entire blocks of Q&A are excised. The cuts are made to bring the conversation down to time and present it in its most compelling form....
“Cleaning up” tape. This refers to removing ums and ahs, as well as stumbles and pauses, from an answer. Most of the time this is fine, as long as you are not removing something that suggested the interviewee was reluctant or unable to respond to a question....
https://training.npr.org/2019/11/12/deciding-what-parts-of-a...
My exposure to the darkroom began when I edited my first story. I sat at the digital editing console with a producer, listening to an interview with a source I had recorded earlier. With just strokes of the keyboard, he cleaned up and tightened the sound bites I was going to use, taking out sentences, words and even some of the pauses, making what are called "internal edits." Then the various thoughts were woven together technically in a way that would be totally hidden to the listener.
https://www.wnycstudios.org/podcasts/otm/segments/130565-pul... (2003) (see Transcript)
An intelligent verbatim transcript attempts to capture what was said rather than how it was said. The main priority being content and “voice”, while leaving out repeated words and fillers.
The aim of an intelligent verbatim interview transcript has less to do with depicting accents or involuntary vocalization. It has more to do with accuracy of the substance of the research interview, considering the meanings and perceptions created and shared during a conversation. And this is why this style of transcription is so popular, because it provides a more readable transcript while staying true to the voice and intended meaning of the participants.
https://weloty.com/intelligent-verbatim-transcription/