Unfortunately our machines are only as good as the audio that is submitted. If your audio or video file contains background noise, slurred and swallowed words, and has people talking over each other, the resulting automated transcription will be negatively affected.
Many of our users have taken it upon themselves to figure out ways to capture higher quality audio so that they remove the transcription headache as much as possible.
One of the ways you can improve quality in Sonix is by using our multitrack feature. If you record multiple speakers on different tracks, you can upload those separately. This improves the quality of the transcript and as a bonus you can automatically label speakers. You can find more about multitrack here: https://sonix.ai/resources/automatic-speaker-labels-multi-track-transcription
Here is a blog post that provides some tips & tricks to capture better audio: https://sonix.ai/greataudio.