Can A Fashion Film Use A Made Up Model?

Yes. We can generate a synthetic fashion model. Whether you can publish that film where you want depends on approvals, disclosure and the rights and licences required by the markets and media you target. Record those needs in the brief so they can be priced and agreed.

Short answer

Short answer. We make films where the people in our work are generated rather than filmed, so a fashion film can feature a model who does not exist. We treat a generated model as a character and hold look, face and wardrobe steady through the cut.

Every sequence is broken into directed shot sheets before a single frame is generated. Blocking, lensing and lighting are decided in writing so continuity is fixed up front, and you see the shot sheets before anything is built.

Try this. If you want a sense of approvals, versions and where generation is allowed, use the Every Market Calculator to compare published shoot ranges with the computing column for generation and lip synced dubbing, then send the brief so we can price the actual campaign. Nothing is quoted at that stage. The calculator runs in your browser and sends nothing unless you request the estimate by email.

How we make the people on screen

We build each shot from structured prompts after a shot sheet is agreed. A director decides blocking, lensing and lighting in writing before a single frame is generated, and characters are held steady with reference sheets so the same face reads across a whole production. See the method.

Production runs on JSON driven prompt control and frame accurate lip sync is part of our pipeline rather than an afterthought. Versions are produced and compared side by side so you can pick looks, poses and timings before anything is committed, which reduces last minute surprises. Nothing is needed from you while production runs, unless something in the brief changes.

We are not tied to one platform. The stack changes and a project is never designed around a single provider, which helps us port the same character across wide, upright and square masters composed for each frame. See examples on the work page.

When a real person is needed

Some briefs need a real person or a real shoot. A shoot is the better answer when the brief needs a real, named person on camera, a live event, a physical product exactly as it is, or a regulatory approval that requires actual footage.

Where a project needs a real person's likeness or voice deliberately, that needs their documented permission in hand before production rather than after. A standard consent form has not been drafted, so permission is agreed in writing per project.

If your campaign rests on a recognisable face, put that requirement in the brief so it is part of scope and price. Also include any approvals or media owner and retailer conditions, because those requirements can determine whether generation is acceptable for a market.

What we can confirm

Every project runs through our creative director from first prompt to master delivery. You work with that director rather than a team of generalists, and grading, logo lock ups and the masters are checked by that director before anything leaves the studio.

Deliverables are set at the brief. You receive colour graded masters with logo lock ups in the aspect ratios and formats we agree at the start, and versions are reviewed side by side so the final master is the option you chose rather than an on the day decision. Masters are produced from one concept, not re-edited per placement, so format and aspect ratio requirements are set at the brief and applied in finishing. Finished masters are delivered as agreed.

The Every Market Calculator is explicit that its computing column shows published rates for generation and lip synced dubbing only, and that the studio's own work and fee are not included. Use it to see how languages, formats and versions add up before you send the brief.

What we have not decided

The responsible production page is a working draft and it names gaps rather than filling them. It notes there is no written approval procedure with named sign off stages, no standard retention rule for supplied files, and no automatic per file disclosure or embedded provenance on delivered masters.

It also says there is no process here for checking whether a generated face or voice resembles a real person closely enough to matter, and it would be wrong to imply one exists. Where that question matters, state any tolerance and remedy in the brief.

Production runs on third party platforms and their licence terms can affect how finished work may be used. The studio is bound by those platform terms and where a licence would affect your rights in the masters, that belongs in the project agreement rather than this site.

What to put in the brief

Start by saying where the film will run, how long it needs to be, which formats and which languages you need, and whether you require a named or recognisable person on camera. Length drives the shot count, and the shot count drives everything behind it. The treatment stage is the cheapest point to change the creative approach, before any shot design begins. The brief form captures that and more; use the brief so the first conversation covers what the piece has to do and nothing is assumed.

Record any approvals or media owner rules that matter to you, who signs likeness releases, how long assets are kept, whether you want an on screen AI disclosure, and which platform licences matter. These decisions affect scope and price.

Send the brief rather than trying to resolve these items by email. The studio's creative director reads the project brief and replies within 24 hours. A project is priced from the brief and confirmed in writing before anything starts.

Sources

• The people in the studio's portfolio are generated rather than filmed: https://www.vsnry-motion.com/responsible-ai-production, read 14 October 2026.
• Every sequence is broken into directed shot sheets before a single frame is generated: https://www.vsnry-motion.com/method, read 14 October 2026.
• Each shot is built from structured prompts rather than prose, using JSON driven prompt control: https://www.vsnry-motion.com/method, read 14 October 2026.
• Characters are held steady across the production and dialogue matched with frame accurate lip sync: https://www.vsnry-motion.com/method, read 14 October 2026.
• Every VSNRY project runs through the creative director from first prompt to master delivery: https://www.vsnry-motion.com/about, read 14 October 2026.
• Versions are produced and compared side by side rather than committed to on the day: https://www.vsnry-motion.com/method, read 14 October 2026.
• The responsible production page names gaps rather than pretending they do not exist: https://www.vsnry-motion.com/responsible-ai-production, read 14 October 2026.
• Where a project needs a real person's likeness or voice it needs permission before production: https://www.vsnry-motion.com/responsible-ai-production, read 14 October 2026.
• The Every Market Calculator excludes the studio's own work and fee from the computing column: https://www.vsnry-motion.com/tools/calculator, read 14 October 2026.
• The studio's creative director reads the project brief and replies within 24 hours: https://www.vsnry-motion.com/brief, read 14 October 2026.
• No single platform is good at everything and the good ones change every few months: https://www.vsnry-motion.com/method, read 14 October 2026.
• A project is priced from the brief and confirmed in writing before anything starts: https://www.vsnry-motion.com/services, read 14 October 2026.