Oncology waiting rooms are architecture plates, not patient faces
Hailuo silent plate of the empty room you photographed; patients are not extras.
Hailuo silent plate of the empty room you photographed. Patients are not extras.
An oncology lobby is a room a real person will sit in. The ad is architecture: chairs, daylight, a reception desk, a corridor that is actually yours. It is not a generated patient in a gown, not a cousin of last week's infusion chair, not a photoreal oncologist explaining wait times. Photograph the empty room after hours. Then buy silent motion. Do not cast.
Photograph the empty room
Shoot the lobby when it is empty. Same lens height a visitor would have. Same millwork, same art, same seating run. That JPEG is the plate. It is also the rights file: nobody identifiable is in it.
If the generate must open on that millwork, upload the still to image-to-video. Frame one is the file. A paragraph about "calm modern cancer center waiting room, soft morning light" will invent a different lobby. Invented architecture is a different clinic.
Crop people out before you generate. A reflection in the glass, a name badge on a desk, a chart on a monitor — strip those in the still, not in a prompt. Do not ask a model to "remove the patient and keep the chair." It will rebuild both.
The plate is allowed to move: a slow push down the seating run, daylight sliding on the floor, plants that were already in frame. It is not allowed to seat anyone. An empty room you own is honest. Generated "community" is a cast you do not have grants for.
Hailuo 02 is silent architecture
Hailuo 02 is MiniMax text-to-video: prompt-to-action physical motion, audio false, 1080p, listed 8 credits. No image is required. The catalog does not publish a duration ladder on this slug. Do not copy a 4 / 6 / 8 list from another row.
Silent is the point in a clinic. Hailuo 02 will not mint a waiting-room murmur, a nurse call, or a voice explaining labs. Plan music later if you need a bed. Do not pick this row hoping for native audio. Audio is off.
Because the row is text-to-video, it will not ingest the lobby JPEG. Use it when the job is empty-room physics from a sentence you wrote against the photo: camera locked, no people, no mouths. Check the take against the photograph. If the seating run is not yours, throw it out. If the millwork must be the file, you already wanted image-to-video, not this Hailuo.
1080p is the ceiling. Cinematic / Documentary / Abstract are looks, not soundtracks and not identity locks. Documentary will not add location audio. Cinematic will not add a clinician.
Patients are not extras
A deepfake of a private person is not a style. A patient is a private figure. TikTok's AI-generated content help forbids likeness of adult private figures used without permission, and of anyone under 18. A label does not create a grant. "Generic person in a waiting room" is not a defence if you seeded a real headshot, a review photo, or a staff Slack crop.
Do not prompt a photoreal patient in a gown. Do not generate background sitters "for scale." Do not invent a caregiver holding a hand. If someone who knows them would say "that is them," you have a likeness problem. Invented patients who look comforted by your lobby are fabricated testimonials.
Minors are a hard stop. No synthetic child in a pediatric bay, no "aged-up" school photo, no generated grandchild in a chair. Shoot empty rooms that are yours, or staff who signed a replica grant for this use.
A generated nurse or oncologist who advises is a second kill. YouTube's inauthentic-content third bucket is AI-generated personas that present as human experts on health. A constructed specialist plus "here is what you should expect from infusion" is that shape. Labelling the face AI does not leave the bucket. Keep clinical lines off synthetic mouths.
This is general information, not legal advice.
Hours as overlay, not a clinician mouth
What the clip needs to say — hours, a booking URL, "results vary," a campus name — is type you typed after picture lock. Burn one authored line with a text overlay: whole clip, top / center / bottom. It does not transcribe. It does not invent a clinician so the captions have a face.
Urgent care clinics already have a nearby audience page: walk-in queue, hours, a facility plate without a diagnosing host. Oncology waiting rooms are not that page. This is a cancer-care lobby, often a multi-hour sit, often a person who did not consent to be scenery. Urgent care is a different building and a different wait. Do not copy walk-in prompts onto chemo chairs.
Staff on camera only with a grant that covers generation, not last year's open-day photo release. No grant: empty architecture, or a named credentialed human you actually filmed. Generation can still do the empty room. It cannot be the consult, and it cannot be the patient.
FAQ
Can we generate blurred patients in the background for "realism"?
No. Blur is not a permission structure. Photoreal sitters are still people, and a lookalike of a real private person is still that person. Keep the room empty.
Does Hailuo 02 ingest our lobby photograph?
No. Hailuo 02 is text-to-video, silent, 1080p, 8 credits listed. It does not require an image and will invent a cousin lobby from a sentence. If the millwork has to be the JPEG, use image-to-video. Use Hailuo 02 when you need silent empty-room motion and you will check the take against the photo.
Can a synthetic oncologist explain wait times if we label it AI?
Not as a YouTube-monetised health adviser. A constructed specialist prescribing what to expect is the YPP persona pattern. Overlay the hours. If someone must speak, put a real named clinician on camera.
How is this different from the urgent-care audience page?
Urgent care is walk-in facility marketing. This page is an oncology lobby as architecture. Same rule on faces: no generated patients, no diagnosing avatar. Different building, different wait, different still.