Essential oils: dropper identity and burned type; skip the clinician VO

Vial still is I2V. Claims stay captions. No medical talking head.

An essential-oil listing is the dropper you sell: this glass, this bulb, this botanic name on the label. Motion is a bead at the tip, a slow tilt, light on the meniscus. A constructed clinician walking "what you should take for that headache" is a health persona YouTube already named. Animate the bottle. Typeset the claim. Do not hire a mouth the model invented.

The vial still is the SKU

Photograph the bottle you actually ship. Same dropper, same label crop, same rubber colour. Text-to-video will invent a prettier apothecary with a cousin font and a herb you do not distill. That is a failed SKU, not a cinematic upgrade.

AI image to video keeps frame one as the file you sent. Labels, fill line, and the cardboard behind the vial survive because they were already in the picture. The prompt is for physics — a drop forming, a slow rotation, a bulb that does not grow a new bottle. Re-describing the brand in the prompt is how the model starts drawing a second label.

If the still includes a face, crop it. You do not need a person for this ad. You need the dropper. A generated patient, a dummy in scrubs, a "dermatologist" holding your vial — those are people. Health content that invents a body is a hard stop. No residual limb, no clinic extra, no child with a cold. The glass is the subject.

Do not send "photoreal lavender oil in a spa, wellness ritual" as text-to-video and hope the SKU appears. Average is a cousin dropper with a Latin name the model invented.

I2V the dropper; do not invent a protocol

Once the vial still is locked, the generate is allowed to do three things:

• Move the camera around the bottle. Slow orbit, a gentle push to the dropper tip, a hold on the botanic name.
• Add diegetic liquid. A bead, a tilt, oil catching light. Not a new herb.
• Keep the label. If the letters crawl, the take is unusable.

It is not allowed to "improve" the INCI, invent a wellness protocol, grow a spa, or put a clinician in frame. Those are claims. A buyer and a reviewer will notice them before they notice your logo.

Prompt camera and physics only. "Locked vial, drop forms at the tip, label holds, no person." Stop. "Aromatherapist explains anxiety protocol" is a second job. It is a private-person job if a named founder is on payroll. It is a health job if the line is treatment.

If the still is a lifestyle crop with a diffuser in the background, crop to the vial or shoot again. Do not prompt the room out and hope the model rebuilds the glass.

Claims are type you wrote, not a mouth

Skip the clinician VO. A mute dropper still has nothing to transcribe. Add captions to a video is speech in, styled type out: 165 languages, BASIC at 1× and DYNAMIC at 2×. It listens. It does not invent a protocol. Running it on silence hoping "lavender" appears is a spend on a blank track.

The botanic name, the volume, the "not a drug" line — those are strings you already print. Add a text overlay burns one authored line, whole clip, top / center / bottom. Timed overlays if the beats change. Neither tool belongs in the I2V prompt. Prompted glyphs on glass are a misspelled Latin name you regenerate to fix.

If a real named person ever speaks — a founder, a formulator on payroll — caption that take. Do not generate the take so captions have something to listen to. Captions are a finish on speech you already have. Skip the clinician VO so you never need a synthetic mouth.

Do not ask the video model to typeset a health claim on the vial. Do not lay a closed-mouth read under the dropper. The voice is still a presenter. Wellness remedies are in YouTube's Help example on purpose.

/for/dermatologists is the clinic; this is the bottle

AI video for dermatologists is the nearby audience: a board-certified clinician correcting a viral claim, a procedure room with no patient, ingredient education that is not a diagnosis. It is written to a practice. It is not an essential-oil SKU hold. Do not rewrite that page as a dropper commercial, and do not paste its "when a mole needs checking" voice onto this still.

Synthetic presenters on health and finance are a YPP kill. Help's third inauthentic-content bucket is AI-generated personas that present as human experts on health. A constructed clinician, herbalist, or "aromatherapist" prescribing a dropper is that shape. Labelling the face AI does not leave the bucket. Fan funding and ads sit behind Partner Program eligibility.

A silent vial is not that shape. Authored type on the plate is typesetting, not a mouth. A real, named, credentialed person on camera is a different day — and then you caption that take. Generation can still do the drop. It cannot be the consult. If the first three seconds look like a consult, rewrite until they look like glass.

FAQ

Can we generate a clinician so the dropper has a protocol?

No. That clinician is a health persona. Photograph the vial. Animate the still. Overlay the botanic name and the line you already print. A real named person, on a different shoot, if you need a voice. Do not invent a body so the bottle has a scene.

If the clip is silent, should I still run add-captions-to-video?

No. Captions transcribe speech. A mute dropper has nothing to listen to. Burn the claim as overlay. Use captions only when a real named person actually spoke, and then caption that file.

Why not text-to-video a "spa ritual" and put our label in the prompt?

A sentence is not a plate. The model will invent glass, a font, and a herb. Upload the vial you ship and describe only the drop. Identity of the dropper is the listing.

Is /for/dermatologists the page for essential-oil content?

It is the clinic page: a named clinician, a room with no patient, myth correction. This page is the vial still plus burned type. Do not rewrite the dermatology audience as a catalog clip, and do not replace the dropper with a talking explainer.