To put a doctor on screen, we need a set of clear reference photographs of them, and if they speak in the film, a short voice recording. Your team takes the pictures. We build the lighting, the wardrobe, the environment and the film around that likeness. Here's what good reference looks and sounds like, and how to send it.
Tim and Gavin Devereux, co-founders of Continuas AI, in an investor film made with this process. Neither was filmed. We modeled both: likeness, delivery, lip sync.
Watch the mouths and the light on their faces. That's the standard we hold every Screen Test to, doctors included.
We develop the photographic look of each character through specific prompt structures, making deliberate choices about virtual camera and lens settings, camera angle, depth of field, lighting and composition. Those choices shape how the face is presented, how light falls across the features, and how the doctor stands out from the background.
We can explore softer, more flattering lighting, natural skin tones and different environments to give the image depth and a cinematic finish.
The reference photographs establish the doctor's appearance. Our work builds the visual treatment around that likeness, giving the finished character the attention to lighting and composition you'd expect from a professionally directed portrait.
For a narrative :60, those decisions run across a whole sequence. We hold continuity shot to shot, then shape the edit, pacing, sound and finish.
Dr. Simon Oh of Smart Arches Dental Implants. His practice sent us interview footage they'd already shot in 4K. We pulled a front-facing frame and a turned frame, and that was the reference. Everything on the right was built from it. That's enough for stills; a doctor who speaks and moves on screen needs the full series below.
Same person, three directions. New lighting, new wardrobe, new setting, new framing, and still recognizably him. The reference only had black scrubs; the lab coat is wardrobe we added. His embroidered name carried through because it was legible in the reference.
For whoever's behind the camera. This is our creative team's capture spec, and every point shows up in the finished film: a soft photo becomes a soft animation, and a shadow across the face becomes something the model has to guess at.
Window light or a softbox, ideally in daylight. Keep the key light on the front and sides of the face with subtle fill. No harsh directional light, and no backlight or rim light: it looks great in a normal portrait but hides the features we need.
Every shot from about eye level. Chin neutral with a slight downward tilt, never tipped up. A warm but neutral expression, subtle confidence rather than a big emotion. Mouth slightly open, not pressed shut and not a full smile; that's what lets us build natural lip sync.
Glasses, hats and hair over the brow cover the landmarks we map expression from. Shoot one close-up with glasses on, then the rest with them off. If hair normally covers the forehead, take 3 to 4 medium close-ups with their usual styling, then shoot the series with it pulled back.
Usual grooming, minimal makeup, real skin texture. No beauty filters, no skin smoothing, no retouching. Heavy makeup and extreme hairstyles don't hold up well across expressions.
If a particular lab coat, uniform or embroidered name needs to appear, give us 2 to 3 full-height shots of it, front and back, plus a side if there are details you can't see from either.
An uncluttered wall with no one else in frame. Keep the same lights and camera height for the whole session, so every photo matches the others.
Not one perfect photo: the same person from every side, at three distances, under the same light. The camera stays put and the person turns. About 18 photos plus a few extras, all from eye level.
Solid marks are used in every framing. Dashed 120 to 135° turns are for the medium close-ups only.
Head to feet, with arms, hands, hair and clothing all in frame. Hands relaxed at the sides, slightly away from the body.
From the waist to just above the top of the head.
Top of the head to the shoulders. Eyes open, lips slightly parted. The face fills about 60% of the frame, with the hair and both shoulders in and nothing touching an edge.
The full series is for anyone with a speaking role. Someone who only appears in a group shot and isn't featured doesn't need the glasses and hair steps.
A real reference set from our creative team. Top row: full body, front, both sides and back. Below: the medium close-ups and mediums turning through the angles, then two full-body shots with the arms held slightly away from the body.
Notice what's missing: no glasses, no hair over the brow, no shadows, no props, nobody else in frame. And the uniform details, like the name tag, read clearly from the front, which is what we need if they have to appear in the film.
A full-frame camera and a proper portrait setup. If you're handing this to a photographer, this is the part to forward.
Can't get a shoot together? Clean 4K video of the person facing camera can work. It's how Dr. Oh's portraits above were made. The full series gives us much more to work with, especially for a doctor who speaks on screen.
If the doctor speaks in the film, we model their voice too. That needs a clean, unprocessed recording of them reading aloud. Same rule as the photos: clarity beats polish, and the room matters more than the mic.
Come with your usual grooming. Keep makeup minimal. The film should look like the doctor patients meet.
Glasses or hair over your forehead? We take a few shots your usual way first, then the rest with glasses off and hair back, so we can see your brow.
Bring the coat or scrubs you want on screen. We'll photograph it front, back and side.
Stay relaxed and warm. Lips slightly parted rather than a big smile, while the photographer turns you through a set of angles. No lines to learn. For voice, about two minutes reading from a favorite book and a short passage from a professional journal.
Organize the RAW photos and WAV audio by doctor and by framing. We want uncompressed files, so send them through a proper file transfer service: Box, Dropbox, WeTransfer, Google Drive or SharePoint. Zip the folder if that's what it takes to get it to us intact.
Name the folders so we can tell who's who without opening anything.
Send us the link