Plumbing OEM: SKU still, not a plumber talking twin

I2V the fixture. Plumbing OEM video is a SKU still put into motion. It is not a generated plumber who talks like a brochure. The catalog is the product. A talking twin is a different audience.

Fixture buyers already know the valve. What they need is: this is the SKU, this is the finish, this is a handle that still matches the cut sheet. That is an image-to-video job. It is not a script.

The fixture is frame one

Photograph the approved unit the way you would print a spec sheet: square-on, finish true, markings readable, no invented type on the body. Animate a slow orbit, a handle turn, water sheeting off a spout you actually ship. Do not prompt a new faucet.

Seedance 2.5 Image to Video is the hold: required still, first frame plus optional last frame, native audio, 4s–30s, 480p / 720p, listed 47 credits, billed resolution-based per second. Catalog copy: one frame into continuous motion without the drift or stitching of shorter multi-clip workflows. If the shot must land on the model number, feed that square-on still as the last frame.

720p is the family ceiling. A 30s pass of the wrong finish is a long wrong finish with a soundtrack. Approve the still the way you approve a print proof.

Image-to-video keeps the upload as frame one. Keeping a physical SKU believable is the same gate for a valve as for a bottle: freeze, crop the marking, compare word-for-word to the canonical still.

Native audio on this row is a mix — a faint water bed, a cartridge click — if you write it. It is not a plumber who speaks.

The plumber playbook is a different audience

AI video for plumbers is a trade-business page: the satisfying fix shot, a clear pipe, a diagnostic walkthrough, hidden-leak checklists. That playbook is for shops that bill labor. It is not an OEM cut sheet in motion.

Do not take that talking-walkthrough instinct and generate a plumber who does not work for you. A synthetic tech pointing at your SKU is a presenter job plus a likeness problem. OEM catalog video does not need a face. If a trade partner wants the fix-shot format, that is their feed. Your job is the fixture they would recognize in a box.

Overlay the SKU; do not generate a twin

Model numbers, finish codes, and "not for potable" lines will change in the next revision. They do not belong in a video prompt.

Burn typed lines with timed text overlays or a single text overlay. Then you can swap the finish name without regenerating the brass. On-screen text in generated video is why: a still has to be right once; in-frame type has to be right every frame.

A talking twin is the expensive way to typeset a spec. If you need a host, that is a different row after the plate exists — a real person or a disclosed presenter, not a fake plumber. Keep Movie Mode for a brand film about water. This unit is a still, a Seedance hold, a few timed lines.

FAQ

Why not text-to-video the faucet from a prompt?

Because customers already own the drawing. Text-to-video will invent a spout. Image-to-video starts from the photo you would ship. Seedance I2V requires an image. Use it.

Is the plumbers audience page the wrong link for OEM?

It is the wrong playbook. Plumbers is for trade shops filming the fix. OEM is the fixture. Link it so you do not accidentally run their talking-walkthrough on your SKU.

Can I I2V every finish and pick later?

You can. You will pay video rates to discover a still fact. Approve finishes as stills. I2V the chosen unit. That is the cost shape.

When is a talking installer acceptable?

When the brief is a partner-facing how-to with a real person, or a disclosed presenter who is not a fake plumber. The SKU in the listing photo still has to match. Generate the host on a talking row; keep the product plate separate.