How to apply this guide
Capture a sample that covers the third mural after the stone arch, then check the beginning and end before transcribing the whole recording.
Start with
Check whether the first words name the subject. For the sample “Turn left after the stone arch, then stop beside the third mural.”, keep enough audio from before it to explain who or what is being discussed. Make sure the input meter responds during a short practice recording.
Check before you finish
Use the smallest approved portion that preserves context. Keep the original and name the clip by date, subject and start time so the tour editor can replay it later.
Evidence to keep: Confirm the app edition, permission and destination together; an allowed permission alone does not prove that insertion will work. This page’s key limit is: Capture a sample that covers the third mural after the stone arch, then check the beginning and end before transcribing the whole recording.
Start here
Before turning a neighbourhood audio tour into text, decide which parts another person needs to review. For this example, start with the cleared tour recording and its route notes. A clear recording still needs the context that identifies the topic and what happened next.
Start before the identifying sentence
Check whether the first words name the subject. For the sample “Turn left after the stone arch, then stop beside the third mural.”, keep enough audio from before it to explain who or what is being discussed. Make sure the input meter responds during a short practice recording.
Keep a complete ending
Play back the last complete statement and check whether it addresses whether the path is open in winter. If the source does not answer that question, retain it as an open point. Start the next clip at a complete sentence so a reader can place it in the timeline.
Choose what to send for transcription
Use the smallest approved portion that preserves context. Keep the original and name the clip by date, subject and start time so the tour editor can replay it later.




