The two things a read-through is usually missing: a stop where there should be one, and the breaths and laughs that belong in it. In DOTV both come from markers you drop into the text — the two buttons at the top of the input bar, Pause and Interjection — so the AI voice-over doesn't come out as one flat, breathless newsreader take. This section covers how to insert both kinds of marker, and how they divide the work with the settings panel.
Two rules come from the synthesis engine, and breaking either one kills the whole take: a marker has to sit between words that can be spoken (not at the very start or the very end), and two markers cannot be written back to back.
Put the cursor where you want the stop, click Pause, and pick a length to insert.
This is the only set of markers the engine knows. Write your own — [laughing], *sighs* — and it is treated as text and read out exactly as written, which is why you pick them here instead of typing them.
Again, put the cursor in place, click Interjection and pick one.
19 in all: Laugh, Chuckle, Sigh, Breath, Pant, Inhale, Exhale, Gasp, Cough, Clear throat, Sniff, Snort, Sneeze, Groan, Burp, Lip smack, Hum…
Use them together: set the speed of the take with the settings first, then use pauses to open space where it counts.