BRAND FILM
PRODUCTION

Brand film production has always been rationed by geography. The idea wants a tulip field at dawn and an orbital view of the same coastline eight seconds later, and the budget answers with one afternoon in a rented warehouse. Ambition gets negotiated downwards in a meeting nobody enjoys, and the finished piece describes a company two sizes down from the client who commissioned it.

Generative methods remove that negotiation, which turns out to be the dangerous half. When anything is available, most output becomes a showreel: nine unrelated gorgeous shots with no argument running between them. Capability is not a story. A picture that can travel anywhere still needs a reason to move, and that motive has to come from what the business claims about itself.

So we write before rendering. The narrative is built around a single spine, usually one transformation stated in a sentence a chief executive would repeat unprompted. Shots earn their place by advancing that line. Anything merely spectacular is cut at the treatment stage, where losing it costs a conversation instead of a fortnight.

Then the world gets specified. Locations become written documents: architecture, era, palette, one key direction, a stated colour temperature. Continuity survives because paperwork enforces it. A hero object becomes a master plate referenced by every angle. Where a face recurs it is a trained identity, locked before the first shot, so whoever appears at ninety seconds is recognisably the same individual at twelve.

The grade decides whether a generated sequence becomes cinema or stays a collection of clips. Everything travels through a single look: matched contrast, a shared black point, grain applied over the whole timeline. Sound follows identical logic, scored as one arc across the whole cut. Most of what separates our work from an impressive demonstration happens in these two stages, after the interesting part has finished.

You receive the graded master at delivery specification, the cutdowns your channels need, a stills set pulled from real frames, and the sources underneath: plates, world documents, identities, project files. Your next picture starts from that library. Companies commissioning one a year find the second costs a fraction of the first, which is the entire argument for keeping your materials.

Approval is where these projects run late. So the treatment carries frames, and the previsualisation exists before anybody expensive is committed. A stakeholder sees roughly what will arrive while changing it is still cheap, and the awkward conversation about whether the opening works happens in the first week. Nothing about generative tooling fixes a decision-making culture, but showing pictures earlier is the single intervention that moves it.

Rights are settled up front and written into the contract. Model licences cleared for commercial exhibition before rendering, music either scored originally or properly licensed, any recognisable likeness consented in scope and duration. Broadcasters and large advertisers increasingly ask for that paperwork, and assembling it retrospectively is unpleasant. Producing it alongside the work costs almost nothing.

What we will not do is pretend the method suits every brief. A documentary resting on genuine testimony should be filmed with a camera and a person. So should anything where the value lies in proving something unrepeatable happened. Where the subject is aspiration, scale, transformation or a place that cannot exist yet, this approach wins on its own terms. Saying so out loud matters, because a supplier who never declines anything is selling you something.