Reads a board oral-exam transcript aloud as a two-voice exchange — examiner and candidate — for hands-free review.
Voices come from your browser/OS. The best-sounding ones are pre-selected; lower-quality system voices sink to the bottom of each list.
Pitch is honored only by local/system voices (e.g. macOS and Windows desktop voices). Many premium voices — Google, "Natural"/"Neural", and Siri — ignore pitch entirely, so this slider may have no audible effect when one of those is selected. Pick a ★ or ★★ system voice if you need the two speakers to differ in pitch.
Cloud TTS (not enabled in this version). This build uses your browser's built-in voices and costs nothing. A future revision can route synthesis to a cloud provider for more natural voices — the code has a single drop-in point marked CLOUD TTS DROP-IN POINT.
Off by default — fillers are kept so the read sounds like a real candidate under pressure. Repeated-word stutters are always preserved. The on-screen transcript is never altered; only what's spoken changes.