Mato does not have a self-service "report an audio problem" button. What it does have is a regenerate flow with a few behaviors that explain most of the audio issues people run into. This article covers what the product can actually tell you and what to try before contacting support.
A regenerate did not fix the section I heard a problem in
By default, Regenerate audio is incremental: it only re-renders the sections whose script text has changed since the last successful render, and reuses the existing audio for everything else. If you heard a glitch in a section you have not edited, a plain regenerate will not touch it, because Mato does not know that section needs redoing.
Two ways around this:
- Make a small edit to that section's script text (even a punctuation change) so Mato treats it as changed and re-renders it.
- Use Force full regen, which rebuilds every section from scratch instead of reusing anything. This option is limited to workspace owners, admins, and Mato staff.
The regenerated episode lost its chapters or background music variety
This usually means the script does not use section markers. Mato splits a script into sections using markers like [INTRO], [SEGMENT 1: Title], and [OUTRO]. Without them, a regenerate treats the whole script as a single block: no chapter markers, no per-section parallel generation, and no background overlays per section. Check the regenerate confirmation dialog; it warns you when it detects a script missing these markers. Add them to the script and regenerate again.
The waveform or chapters look wrong, or the transcript skips the ads
If the podcast has dynamic ad insertion turned on, this is expected rather than a bug. Ads are stitched into the audio per listener at play time, so a single fixed waveform and a single fixed set of chapters cannot represent every listener's version. The transcript you see is derived from the script itself, with ad reads excluded. If you need a fixed waveform and chapter set, this is a tradeoff of dynamic ad insertion rather than something to fix per episode.
There is no visible error, but something still sounds off
Mato does not currently surface an automatic loudness or quality score to listeners or creators. If nothing in the episode's Overview tab flags a failure (see Why did generation fail? for what a real failure looks like), the audio generated successfully by Mato's own pipeline standard, even if you are not happy with a specific line reading or pacing choice.
For a pacing or delivery issue rather than a technical one, the fix usually belongs on the talent's voice settings (speech speed, director notes) or in the script itself, not in a retry. See Configure voice and delivery.
Before you contact support
Have these ready:
- The episode's URL.
- Whether the section in question was edited before you regenerated, or untouched.
- Whether the script uses section markers.
- Whether dynamic ad insertion is on for the podcast.