The Test Runs Before The Render Is Funded

The brief was 15 s, vertical, and it never reached a renderer. That is the point of the exercise. A 15 s teaser for the X-150 gasifier is not small work: 9:16 at 1080x1920, about 6 shots, 1 hook inside 30 s, and a score sheet carrying 100 points. The pipeline that produces it has 12 stages. Stages 0, 2, 4 and 11 were run end to end on a real brief with no image generated anywhere, because a test whose pass condition is that a picture appeared proves nothing about whether the decisions upstream were right. I have paid for that mistake. In 2008 I signed a EUR 20 million personal guarantee against a plan that looked finished and was not, and the company folded on Christmas Day. A render is the cheapest thing to produce and the most expensive thing to get wrong late.

Triage is 1 decision that constrains the other 11. The register here is bright-modern: clean, fast, optimistic, no doom. Every register carries an anti-signature, which is the list of things that are correct in some other film and wrong in this one - moody, dark, doom, climate-crisis imagery, smog, grey, slow pans. That line is worth more than 20 lines of positive prompt. A renderer filling an empty gap reaches for the moodiest reading of industrial it has seen most often, and 6 shots rendered separately do not agree on anything you did not write down. The same shape held for a EUR 75 million industrial group I restructured: the standard is set once at the top, or it gets renegotiated in 12 departments and 200 small decisions. Naming the register took 4 min. Rescuing a moody teaser in the grade takes a day and never fully works.

The story stage carries 20 of the 100 points, and it runs before any image exists. 5-line cards in a fixed order: situation, desire, conflict, change, result. Waste is burned or landfilled and the energy is lost. Cities and industry want fuel without the guilt. Gasification sounds like a failure from the 1970s, so trust is the barrier. The X-150 turns waste into hydrogen-rich syngas at 2,500+ validated operating hours. Your waste becomes your fuel, on site. Then a title with specificity in it: X-150, Your Waste, Your Fuel. Then a hook inside 30 s, opening on a landfill and cutting hard to a reactor hall that is running. Then the body re-ordered 2-1-3-4 by shock rather than by chronology, because chronology is the order the facts were discovered in, not the order they land in. 1 CTA, placed at the peak of the film: book the site tour.

Stage 4 is the stage most pipelines skip, and it is the one that carries the money. 1 card holds 16 fields: subject, prompt with the identity line, negative, impact, detail, framing, lens, lighting, place, plausibility, character line, cut, camera move, continuity, body and emotional beat. Shot 03 cuts on the word factory, which is how a drone pull-back over a landfill slams into a locked-off hall instead of drifting into it. The lens is held at 35mm for the whole beat-group, because changing it mid-group is how 1 scene reads as 3. Plausibility is a written field, not a feeling: the light comes from a clerestory that exists in this hall, and the card says which hall. A field left blank is a decision handed back to the renderer, and the renderer will make it differently every time.

A test whose pass condition is that an image appeared proves nothing about the decisions upstream.

Stage 11 tallies the sheet before any render is funded. Story and script 20 of 20. Register and style 15 of 15. Identity 12 of 15, because the reactor identity is locked and no character block is needed when the facility is the protagonist. Lighting and grade 10 of 10. Anti-AI and anatomy 10 of 10. Motion and cut 9 of 10. That comes to 76 of 76 scorable points. The other 24 do not sit inside the film at all: 10 on sound and voice, 5 on spec and export, 5 on distribution, and 4 in mechanical stages that can only run once footage exists. Reading 76 as a shortfall would be the most expensive possible misreading of that sheet, because it would send 24 points of work back into a stage that was already finished. 200 wood-gasification machines were built across the UK and Europe and 210 energy systems were deployed across Africa and Asia, and in both programmes the order of the work was the part that had to be right before the first unit was switched on.

The test returned 1 sentence worth keeping: every words-first gate fires, and a full production run is the next step rather than the first one. That is a different claim from the film looks good. It says the decisions are in place before the money is, and that the remaining 24 points are mechanical stages rather than defects. 16 years of field work across 12 countries, from rural Nigeria to post-conflict Serbia to Kazan, taught me the same arithmetic every time: the cheap place to decide is on paper, and the expensive place to decide is in the grade, on set, or in front of a client. The probability that a pipeline nobody wrote down survives its first funding round intact is not zero. Most people price it at zero, then pay for it 1 render at a time.