How to Make a Color Script for a Commercial Treatment

A palette tells the client which colors are in the commercial. A color script tells them when.

That difference is small to say and large to use. A palette page answers what; the sequence answers what happens next, and what the change costs. When a client is handed four swatches and asked to approve them, they can only argue about taste, because nothing on the page shows the commercial moving.

So the working method is this. Choose the moments where the color or value relationship genuinely changes. Reduce each one to large fields before any surface detail. Read the studies in order, asking what each frame prepares and what the next one needs. Then run the same story a second time with a different distribution of light, and compare the two passages rather than the two prettiest frames. The comparison is where the useful information lives.

What a palette cannot say

A swatch carries three properties: hue, value and saturation. That's it. It has no area, no order, no duration.

The same swatch can be sixty percent of a frame or a four-pixel accent, and those are different decisions. The same swatch can open a film or close it, and those are different decisions too. A palette flattens all of that into a row of chips that look equally important because they are equally sized.

Try it directly. Take five colors from a palette and put them in a row: nothing in that row tells you which one the viewer sees first, how long they sit in it, or which one is being withheld so that its arrival can do something. Area and sequence are the two things a color script adds, and they are usually the two things the treatment is actually trying to decide.

The practice is established, and it came from somewhere else

Simplified sequence studies are not a new idea. In a published interview, Pixar artist Lou Romano describes starting from broad thematic images and then drawing sequence tableaux to map how color progresses, treating continuity and the larger intent as related but separate jobs. A MoMA audio transcript for his Color Script, The Incredibles (2002) has him describing simplified shapes and color carrying the sequence, with each scene responding to what comes before and after it.

Two limits matter here. Both records come from feature animation, and feature animation is not a commercial pipeline; they establish the practice, not a procedure you can lift. And neither authorizes reusing artwork, which is why nothing from them is reproduced or paraphrased as a visual.

What they do support is the part that surprises people new to this: the sequence is the unit of work, not the frame.

Choose the moments where the relationship changes

You are looking for four kinds of change: a change in attention (someone starts looking somewhere else), a change in setting, a change in action, or a change in light. Any one of them can require a new color and value relationship, and a shift in attention can require one even when the room itself hasn't changed at all.

The test for each candidate moment is a question, not a feeling: would anything about the picture need to be different on the other side of this moment? If the answer is no, you have two passes at the same study, and you only need one.

These moments are not every shot, and they are not evenly spaced. Resist the tidiness of one study per scene. A sequence where nothing changes for four beats and then changes hard in the fifth is using the interval as material, and even spacing would throw that away.

Keep enough neighboring moments to explain the transition. This is the failure mode that looks like success: a folder of five finished hero frames is not a color script. Hero frames are the peaks. The script is the run-up, and if you only keep the peaks you have thrown out the two beats before the turn and the beat after it — which is exactly the interval the script exists to resolve. Eight to twelve studies can carry a thirty-second spot; six is enough to learn the method, and it is what the example below uses.

Reduce each moment to large fields

Work in a fixed order, and don't skip ahead.

Value first. Flatten the batch to greyscale. If the sequence disappears in grey, the colors are carrying work that value should be carrying, and it will collapse the moment a shot is lit differently than you drew it.

Then area. Decide how much of each frame each field covers. A warm note that occupies four percent of the frame and a warm field that occupies sixty percent are not the same event, even when they are mixed from the same color.

Then hue.

Detail last, and mostly never. These are studies. If a study is reading as a finished style frame, it has stopped doing its job, because it will now be argued about as art direction rather than as an argument about sequence.

While you block the fields, state openly what has to stand apart and what has to belong together. That sentence is the actual content of the study.

Product color needs one extra piece of discipline, because three separate things get collapsed whenever someone says "make the bottle orange":

1. The object's own color, which belongs to the client and is not yours to redesign.
2. The light falling on it, which raises its value and pulls it toward the light's hue.
3. The field behind it, which decides whether it separates at all.

A study may decide the second and third. It may not decide the first. If you fix a composition by changing the product's color, you haven't solved the frame, you've quietly redesigned the product, and the client will find out at the shoot.

Six beats, two distributions

The example below is invented for this exercise: a paper comparison rather than a painted sequence, and nothing in it has been produced, rendered or tested. It also simplifies in a way the rest of this piece argues against: one beat per scene, in order, so the comparison stays legible. A real script is less even than this — two studies inside one scene, nothing new across another — and the even version here exists to make the two routes argument-able side by side, not to model a working sequence. The product is a fictional amber-orange insulated bottle, and the six actions are held identical across both routes.

1. Bedroom, before the alarm. Near-black cool field. The phone screen is the only bright note. A sliver of the bottle sits unlit at the frame edge.
2. Kitchen, coffee. The room stays cool grey. The mug's steam catches a small warm patch. The bottle on the counter is the most saturated object in frame.
3. Street, pre-dawn. Blue-grey field. Streetlights are small warm dots, scattered. The bottle rides in the frame cage, saturated but tiny.
4. The bridge.
5. The lobby.
6. The desk, where the bottle is put down.

Beats 1 through 3 are the same in both routes, including the small warm pockets. The divergence starts at 4.

Read the studies as neighbors

Now take Route A and walk it in order, because a sequence is only legible in motion.

Beat 2 gives you the warm note at small scale: steam catching light. Beat 5 repeats that note at small scale: an indicator light. Between them, beat 4 opens the value range without opening the temperature. That beat isn't wasted. It widens the contrast so that beat 5's flat fluorescent has something to be flat against, and it holds the warm change in reserve.

Then beat 6 delivers the same note as beats 2 and 5, at full scale, across the desk.

Notice what is doing the work here. Not the color of the warmth. The bottle is the through-line: from the kitchen through the lobby, four beats, it is the only saturated object in a cool world, and in the sixth beat the world finally catches up to it. A viewer doesn't need to be told that amber means arrival. They've watched one object stay itself while everything around it stayed cold, and then stop being the exception. That's a relationship across time, and it is precisely the thing a palette page cannot show.

This is also where hue-emotion short circuits tend to sneak in. If you annotate the studies "blue = isolation, amber = comfort," the client will argue about your adjectives instead of your sequence, and you will deserve it. Color meaning here is manufactured by composition, action, and what the viewer looked at one beat ago.

Run Route B the same way and the loss is structural. The bottle is inside warmth by beat 4, so it stops being singular three beats early. The thing that made it the through-line is spent at the midpoint, and beat 5 has to be pushed cool to protect any change at all.

What each route spends too soon

Route A's cost is real and worth stating before a client finds it. Five beats of cool, low-contrast, low-temperature-variety imagery can read as one long stretch, and then the change lands with very little time left to be felt. Beat 4's value move and the two small pockets are what keep the stretch from going inert, and they have to be doing that work on purpose rather than by accident.

Route B's cost is the mirror image. The film comes alive earlier and the middle reads as more varied. But the closing product beat has to hold on framing, composition and performance rather than on light, and if the spot was written so that the last beat is the point, Route B has quietly moved the point.

Neither of these is a prettiness contest, and that's the reason to compare passages rather than frames. The right question is not which study do I like but which of these two runs would you rather watch, and where does each one spend its biggest change.

There is also a scale move available inside either route. A change can be local or broad: a pocket of warmth on one object, or a field that covers the frame. Keeping every change local until the end and then going broad is one way to spend late. Going broad in the middle and then local at the end is another. Both are legitimate; they prepare different things.

Annotate the route you keep

A color script proposes relationships. It does not certify physical lighting, color reproduction, or a grading workflow, and the annotation should make that clear rather than paper over it.

For the Route A sequence, the questions handed forward are concrete:

• The sun at beat 6. Is a horizon sun position that lays a band across a desk achievable on the shoot day, at that location, at that hour? The study proposes a direction of light and a coverage, not a schedule.
• The lobby at beat 5. What are the fixtures actually doing, and is the cool the study assumes available in the space, or does it need to be created?
• The desk at beat 6. Is the warm band a window, a practical, or a fixture? Each answer changes how much of the frame it can plausibly cover and how quickly it can change.
• The bottle. Does the finish reproduce the value the study assumes, and can it hold separation against a warm field without altering the product color? If not, the value step has to be built into the field instead.
• The grade. Whether this structure survives shots cut from different setups, times of day and lighting plans is a question for a later pass, not something the studies settle.

Notice that every one of these is a question rather than an answer. That is the correct state for a treatment. A script that answers them has stopped being a proposal and started being a claim it can't support.

Annotate the source of each consequential change too: at beat 6, the change is motivated by the sun clearing a building, which is a light change, not a mood change. Someone painting the frames up needs to know that, or they'll reach for the mood.

Where the change belongs

Keep Route A, and the reason is where the spot wants its weight. The commercial ends on the product being put down, so the last beat has to carry the release, and holding the broad warm change until beat 6 is what gives that beat anything to do. Beats 2 and 5 rehearse the same note at small scale so the ending reads as a culmination rather than a surprise, and beat 4's neutral brightening keeps the middle from flattening without spending the temperature change.

The condition that flips the choice is simple. If the memorable moment of this commercial is the commute itself rather than the arrival — if the audience is meant to remember the bridge and not the desk — then Route B's early light is the better spend, and the closing beat should be rewritten to hold on composition instead of light. That's a decision about where the spot puts its emphasis. It isn't a decision about which sequence looks better in a folder, and a folder is the wrong place to make it.

One practical note for the handoff. Keep the rough studies and the comparison table together. The moment a set of these gets painted up and mailed as individual frames, each one starts getting judged on its own beauty, and the argument you built across six beats — the one that shows what beat 4 prepares and what beat 6 finally spends — evaporates. Finished painting should clarify that argument, not replace it.

Frequently asked questions

What does a color script add that a palette cannot?

Area and sequence. A palette shows hue, value, and saturation but not how much of a frame a color occupies, what the viewer sees first, how long they sit in it, or what is withheld for later.

How should I choose which moments get a study?

Look for changes in attention, setting, action, or light, and ask whether the picture would need to be different on the other side of the moment. If not, one study covers both sides. Moments need not be evenly spaced or one per scene.

Why reduce studies to greyscale before adding detail?

If the sequence disappears in grey, the colors are carrying work that value should carry, and it will collapse when lighting differs. Value comes first, then area, then hue, with detail last and mostly never.

What are the three separate things to keep distinct when placing a product color?

The object's own color, which belongs to the client; the light falling on it, which raises value and pulls toward the light's hue; and the field behind it, which decides separation. A study may decide the light and field, not the product color.

In the six-beat example, why does Route A hold warmth back until the desk?

The spot ends on the bottle being put down, so the last beat needs the release. Beats 2 and 5 rehearse a small warm note so the final broad warm band reads as culmination. Beat 4's neutral brightening keeps the middle from flattening without spending the temperature change.

Further reading

• What a Beautiful Reference Frame Hides About the Shot — Explains why a folder of striking stills misleads: the force often comes from the cut or a later shot, so composition is shown and the sequence's debt hidden.
• Describe What Lighting Makes Us Notice—not Just a Lighting Look — Use when annotating the light change at a beat: name what the audience must read and tie the change to a cause in the world, not a mood.