Continuo Video Lab

1. Your voice not set

Record about 60–90 seconds of natural speech (quiet room, normal pace). Reading the passage below works well. This creates an ElevenLabs instant clone — re-recording replaces the previous clone.

“Hello, my name is … and I’m recording this so my patients can hear information about their condition in my own voice. Clear information helps people feel confident about their treatment: what their medicine is for, how to take it properly, and what to look out for while they recover. If anything in a video is unclear, or symptoms change, patients should always contact the practice for advice rather than waiting.”

2. Your selfie not set

A well-lit, front-facing head-and-shoulders photo on a plain background. This becomes the HeyGen talking photo — uploading again replaces it.

3. Generate a video

Pick a condition, tweak the script, generate. The pipeline is: script → your cloned voice (ElevenLabs) → talking-photo render (HeyGen) → hosted on this worker. Renders take a few minutes. The voice sliders also apply to the test box in section 1 — tune there before spending render credits. Punctuation shapes delivery too: shorter sentences, commas and dashes add natural rhythm, and <break time="0.6s" /> in the script inserts a pause.

Expressiveness Style exaggeration Speaking speed Engine (HeyGen) Head motion (Avatar III only) Expression (HeyGen) Background colour

4. Videos

Links are unguessable but public — fine for testing, not for patient data.