instagram transcript generator file too long
Trim or split the recording before upload, then transcribe one saved part at a time. Every part must be no larger than 25 MB and no longer than 10 minutes: The 25 MB limit is checked when a file is selected or dropped; the 10-minute limit is checked after audio decoding in the form's submit handler, before speech recognition starts. Listen across each cut and check the adjoining transcript lines for words that were lost or repeated.
Size and duration are separate checks
Passing the file-size check does not clear the recording's duration, because the audio is checked after you start transcription.
Here, starting transcription means submitting the form; the duration check happens before speech recognition begins.
The size warning is a reason to change the saved file; the duration warning is a reason to shorten the recording it contains.
Reducing a file below 25 MB does not make it eligible if its decoded audio still exceeds 10 minutes.
Prepare each saved part
Check the size and playing time of each saved piece, rather than assuming equal cuts will produce equally sized files.
Cut between complete spoken phrases, then replay both sides of the cut before selecting the piece for transcription.
Prepare the cuts before selecting the files here; the uploader processes a saved audio or video file one at a time.
Finish reviewing the current piece before replacing it with the next file in the upload box.
Check where the parts meet
At each join, compare the preceding transcript's last words with the following transcript's first words while listening to the recording.
Resolve a missing or repeated phrase at the cut before treating the separate transcripts as a complete account of the recording.