How to apply this guide
Export the meaningful channel to mono and verify the entire utterance from a preserved original.
Start with
Use an audio editor with a solo control. Play the left and right tracks separately and record which has the speaker. Check near the end as well as near the beginning in case routing changed.
Check before you finish
If the mono file works and the original does not, use the known speech channel for that recording. If both are equally wrong, check for clipping, low level or an unexpected language.
Evidence to keep: Keep the target field open and verify that the complete reviewed result reached the intended destination. This page’s key limit is: Export the meaningful channel to mono and verify the entire utterance from a preserved original.
Start here
A live microphone and an imported stereo recording can be processed differently. If a file has speech on just one side, a deliberate mono export of that channel provides a useful comparison.
Identify the side
Use an audio editor with a solo control. Play the left and right tracks separately and record which has the speaker. Check near the end as well as near the beginning in case routing changed.
Export a safe test copy
Take a short clip of the speech side and export it to standard mono PCM WAV. Keep the stereo original unchanged. Import both with the same speech language and plain Dictation mode.
Compare exact words
If the mono file works and the original does not, use the known speech channel for that recording. If both are equally wrong, check for clipping, low level or an unexpected language.




