The Product Must Keep Its Color, but the Commercial Uses Colored Light
The Product Must Keep Its Color, but the Commercial Uses Colored Light
The promise usually arrives late in the treatment, in the paragraph describing the look: the pack's color will be brought back in the grade. Everyone in the room has seen the problem it answers. The kitchen is amber, the window is doing something magenta, and the coffee carton on the counter is neither.
The promise isn't stupid, and nobody is lying. Finishing tools can lift, pull and window a small object back toward something recognizable. But "brought back" assumes the image still holds a usable record of the pack's appearance, waiting to be released. Under a saturated light on a matte surface, that record may be thin. What you're actually deciding, before any of the grade, is whether the product stays inside the scene's light, gets an image of its own, or the scene's light changes. Those are three different commercials. Only one gets made, and the treatment should say which one you're proposing.
“Accurate color” is three questions wearing one coat
When a brand says the pack must keep its color, they mean something specific that they rarely state. Your first job is to make it sayable, and there are three parts.
Which appearance. Usually not the whole object. It's a band, a wordmark, a lacquer, the fact that the carton is matte rather than glossy. A brand that talks about "our color" is often protecting one surface and indifferent to the rest.
Which views. A carton sitting in soft focus at the back of a kitchen is not the same requirement as a carton in a hand, filling frame, at the moment the viewer is meant to recognize it. Write the beats down. You will find the requirement shrinks to two or three moments, which is enormously good news, because it means the rest of the sequence doesn't have to solve anything.
Which reference. Here the ground gets soft, and the treatment should stay honest about it:
- A flat artwork file is design intent. It has no light and no material, so it can't settle what a photographed carton looks like.
- A studio photograph from the brand's library is a photograph, with its own lighting and its own grade. It tells you how the pack is presented, which is useful, not how the pack reflects light, which is what you need.
- A printed carton from the current run is the closest thing to the material itself. Its appearance still depends on the light you view it under, so it becomes a reference only once somebody names that light.
These three can disagree with each other. The disagreement is information: it means the brand has never had to decide which one wins.
So ask the owner — the packaging lead, the brand manager, whoever actually holds the requirement — four things: which surfaces, which views, which reference, and what would count as a failure. "The band must not read as a dark gray-green" is testable. "It must look correct" is not. If someone accountable hands you a numeric tolerance and a viewing condition, use them; otherwise don't put a number in a treatment. A number you invented will be quoted back to you as a commitment.
One more thing worth saying out loud in the deck: a treatment frame with the pack corrected in an image editor is an illustration of intent, not an approval of the product's color. Say it, then put it in the deck anyway, because it will end up there regardless and at least the caveat travels with it.
Find the frame where the light and the requirement actually collide
The conflict is almost never the whole commercial. It's one surface in one beat, and naming that beat precisely is most of the pitch.
Why saturated light is hard on a matte surface: a rough, non-shiny material mostly returns whatever lands on it, filtered by what the surface reflects. Its apparent color is the arriving light's spectrum multiplied by the surface's reflectance — plus whatever else is bouncing around the room. Saturated light is light that is missing energy in some band. A saturated amber carries very little in the blue and green. A surface whose identity lives in blue and green returns almost nothing, and doesn't merely shift hue — it goes dark and dull, because the light it needs isn't there to reflect.
What finishing can do is real: raise exposure, window the pack, pull its hue. Two costs. First, exposure lifts everything, including the amber counter you liked. Second, a windowed pack is, by definition, no longer quite in the room's light, and the seam between a corrected object and the light falling across the surface beneath it is visible even to viewers who can't name what they're seeing. Pushing a nearly empty channel upward also brings along whatever is in it — noise, spill from the neighboring color. This isn't "impossible." It's "not free, and not promised."
That gives the pitch sentence a shape: surface, light, failure, and what a test would show. The teal band, under the amber key at the position we're proposing, is where we need to know whether it reads as the muted teal or as dark gray-green — so that's what the test should show. Compare that with we'll fix it in the grade, which is a location, not an answer.
Keep two questions apart while you do this. What the camera records on the day is a capture question, owned by the cinematographer. What happens to that record afterward is a finishing question, owned by the colorist. They are related and they are not the same, and a treatment that blurs them will produce a shoot that assumes the grade will save it. And don't write a gel recipe into the treatment. The recipe belongs to the people who will test it.
Three routes, and what each one costs the film
The example below is constructed for this piece. No such shoot, grade, or reference review took place, and nothing here should be read as a tested result. It exists to show the shape of the comparison.
A fictional brand, Cinderhill, sells ground coffee in a 250-gram matte-coated carton — cream field, muted teal band, dark brown wordmark. The packaging lead is the person who can say what must be preserved. On hand: two printed cartons from the current run, a flat artwork file, and a studio photograph shot on a white sweep. The scene is a kitchen at night, with a saturated amber pendant over the counter and magenta coming through the window from a sign off-screen. A person opens the carton and spoons coffee into a mug.
Three beats matter. The carton sits on the counter in the background, out of focus. She lifts it and opens it in near focus, where the band is largest on screen. The pack is held at the end. For this example, the lead's requirement covers the second and third beats only — the background appearance is exempt, which is itself a decision worth getting in writing, because it means the sequence doesn't have to be solved end to end.
Keep it in the scene. The pack stays inside the stylized world's light, with a source added that the world can plausibly contain. She opens the carton beside an under-cabinet strip that exists in this kitchen: warmer than daylight, far less saturated than the pendant, and carrying just enough blue and green to bring the band back while the room stays amber and magenta. You keep the product inside the event, which is what the beat is for. The risk is a seam — one object lit by a different light than the surface it touches. Check the contact shadow and the pool of light on the counter next to the carton, because that's what gives the cheat away. Note the dependencies: the strip has to exist in the set, the framing has to allow for it or for another source that motivates it, and the shadow direction on the pack has to agree with something. This is an art department and camera problem as much as a lighting one.
Give the pack its own passage. The action keeps the carton turned, silhouetted, or partly out of frame; the recognition view happens in a separate passage, shot under controlled, less colored light, and cut in where the film wants a breath. Appearance becomes easy to evaluate, and the brand's requirement is met by an image built for it and delivered as a managed file. What you spend is continuity and emphasis. A separate passage is a statement that this object matters apart from the room; the cut interrupts the event; and the pack's state has to match — open or closed, coffee level, the surface it stood on, the spoon. Shoot that passage a week later on a different set and someone has to own that account. The light change also has to read as deliberate. In films whose grammar already includes product passages, it does. In a thirty-second spot that has been single-minded until then, it can read as a mistake.
Change the world around the pack. Pull the amber back to a warm but comparatively neutral practical and let saturated color live on the walls and in the window light, or move the beat down the counter where the magenta is doing less. The gains are the strongest of the three: the pack is genuinely lit by the room's light, there's no seam and no separate passage, and the requirement and the integration are satisfied by the same decision. That's why it's usually the right opening bid — it solves the problem at the level where the problem was created. The cost is the stylization you pitched. If the amber was the whole idea, you've spent the idea to save the carton.
These aren't exclusive. A less saturated key plus a small motivated source is the common compromise. What matters in the pitch is not which route looks more expensive but what each asks the viewer to understand: embedding keeps the product inside the event, separation presents it as a subject in its own right, redesign says the look bends around the pack. Each one is a claim about the product's place in the story, and the appearance decision rides on top of that claim.
Specify a test that can come back no
A test that only confirms the route is decoration. Name five things and one failure condition.
The product: a printed carton from the current run, not a mockup. The surface: whichever band, ink, or finish the lead named. The conditions: the proposed key and any rim, at the proposed position, at the intended exposure and camera position for the recognition beat — and, if the embedded route is on the table, the supplementary source as well, named by its colour quality, its position relative to the carton, and its level against the key, because for that route the added source is the variable under test. The reference: the same carton under a stated viewing light, so there is something to compare against rather than a feeling. The critical view: the framing of the beat that carries the requirement.
With the Cinderhill example, the failure condition is concrete. If the band reads as dark gray-green at the intended exposure, and the under-cabinet strip needed to bring it back also visibly separates the carton from the light falling across the counter beside it, the embedded route is out and the choice moves. Worth checking while you're there: the magenta from the window will land on the band's edge and will not do anything for its face — look at that rather than assume it helps. The cream field will go orange, and the wordmark's separation from it narrows, because the light remaining is the part of the spectrum the dark ink reflects best. Whether that's acceptable is the packaging lead's call, not the colorist's.
Two boundaries keep this honest. A still that has been corrected in an image editor shows what finishing would need to achieve; it does not show that the capture supports it. And a pack shot can be produced, delivered, and reviewed against a stated viewing condition as a managed file. A commercial cannot promise that the same teal arrives intact on every viewer's screen — that claim isn't established by a treatment reference, and the pitch shouldn't make it.
Finally, state the unanswered question in the treatment instead of burying it. "Fixed in post" is not an answer. It's a place where an answer might eventually be found by someone who isn't in the room yet.
Choose conditionally, then ask one question
For the constructed example, the opening bid would be the redesigned world: a less saturated key, the saturated color carried by the window and the walls, and the pack lit by the same light as everything it touches. If the stylization is the reason the spot exists, the fallback is the separate passage, with the action beats framed so no recognition view depends on amber-lit teal. The in-scene source is the narrow option — worth keeping for a beat short enough that two keys on one object don't have time to announce themselves.
And one question goes to the specialists, bounded enough that a no is useful: in the recognition beat, under the proposed key and rim at the intended exposure and camera position, does the band read as the muted teal, or as a dark gray-green that no correction can separate from the light falling on the counter beside it?
Keep the cost to the film next to the appearance decision when you write all this down. A route that keeps the carton accurate by removing the reason the commercial existed isn't a solution — it's a trade, and a trade is fine as long as the people approving it can see both sides of it. The grade can do some of this work. It just can't be the place where the decision was avoided.
Frequently asked questions
Why is fixed in the grade not a safe promise for product color under saturated light?
Finishing can raise exposure, window the pack, and pull hue, but under saturated light on a matte surface the image may hold a thin record of the pack's appearance. Exposure lifts everything; a windowed pack is no longer quite in the room's light, making a seam; and pushing a nearly empty channel brings noise and spill. It is not impossible, but it is not free or promised.
What does a brand mean when it says the product must keep its color?
The body separates three questions: which appearance or surface matters, such as a band, wordmark, lacquer, or matte finish; which views or beats carry the requirement; and which reference is being used. Flat artwork is design intent, a studio photograph shows presentation, and a printed carton is closest to the material but still depends on the light it is viewed under.
What are the three routes for handling colored light and product appearance?
Keep the pack in the scene with a source the world can plausibly contain; give the pack its own passage under controlled, less colored light; or change the world around the pack by pulling the saturated light back and letting saturated color live on walls or window light. Each route makes a different claim about the product's place in the story.
What makes a product-color test able to come back with a useful no?
Name the product as a printed carton from the current run, the surface, the proposed key and rim at the intended position, exposure, and camera position, the supplementary source if the embedded route is being tested, the reference under a stated viewing light, and the critical framing. Then state a failure condition, such as the band reading dark gray-green and the added source visibly separating the carton from the light on the counter.
What limits should a treatment state about color accuracy and deliverables?
A still corrected in an image editor shows what finishing would need to achieve, not that the capture supports it. A pack shot can be produced and reviewed against a stated viewing condition as a managed file, but a commercial cannot promise the same teal on every viewer's screen. Unanswered questions should be stated in the treatment rather than buried as fixed in post.