Chrome Yellow

An original design for the high-chroma yellow-through-green register: the most saturated corpus measured in this bank, built on a palette that runs from chrome yellow into olive and viridian, with a real violet-blue minority and almost no red. Derivation in the notes.

Painterly reference: Vincent van Gogh, measured from sixteen public-domain reproductions against eleven other painters. Named for the pigment rather than the painter, which is this bank's standing rule. Chrome yellow is the lead chromate he actually bought, the one that made the sunflowers and the one that has been browning in museums ever since.

Provenance. Built 2026-09-17 and corrected the same day. The first pass drew on a corpus that admitted framed canvases and works on paper, which understated the cool half of this palette by a factor of five and led to a Color Chrome FX Blue setting that has since been changed. Every figure below is from sixteen Van Gogh oils cropped to the canvas, verified by eye, measured against eleven other painters.

What the corpus measures

The hue census is the blunt part.

Yellow, yellow-green and green together are 76.3% of every coloured pixel, and the warm-to-green run dominates at a chroma higher than any other well-sampled painter here. What the first pass got wrong is the rest: cool content measures 19.9% rather than the 3.5% a contaminated corpus suggested. Van Gogh has real blue in him, which anyone who has seen the night skies and the irises would have said, and the correction restores it.

The green lean is the second finding and the less obvious one. Mean a* is negative at both ends of the tone scale and strongest in the highlights at −2.78, where nine of the twelve painters measured sit on the positive-a* warm side at both ends. Most painting, like most photography, lives there, because skin, earth and wood all do.

Why Velvia

• Chroma 18.6 is the requirement, and Velvia is the base that reaches it without Color being pushed far enough to distort hue on the way up.
• Hard at both ends suits a black point of 22.1 under a white point of 70.4, with the curve then softened deliberately rather than manufactured.
• Its documented magenta lean turns out to matter after all, which the first pass got backwards. Fujifilm states Velvia adds a little magenta to blue, and this palette carries 11.2% blue-violet. Van Gogh's blues are ultramarine and violet-leaning rather than cyan, so a base that tilts blue toward magenta is rendering them the right way rather than distorting them.

The white balance runs the wrong way, deliberately

Nearly every recipe in this bank shifts white balance toward Red. This one shifts away from it.

Fujifilm's shift grid runs Red against Cyan and Blue against Yellow, with no green axis, so green has to be reached obliquely: a step away from Red moves toward cyan, and a step away from Blue moves toward yellow. Cyan and yellow together is green. That is how −1 Red and −2 Blue produce the measured a* of −3.2 with b* strongly positive.

The asymmetry matters. Two notches away from Blue against one away from Red keeps the result on the yellow side of green, which is where the corpus sits: yellow at 36.2% leads yellow-green at 31.4%, so the palette is a yellow that has travelled toward green rather than a green that has been warmed.

The rest

• Color 0 on the most saturated base in the system. Velvia at 0 already sits past where most simulations reach at +4, and the measured target is 18.6 rather than the 30-plus that pushing it would produce. Restraint here is what leaves Color Chrome Effect somewhere to work.
• Color Chrome Effect Strong, which is the setting that stops this becoming a poster. It gives tonal separation to colours that are already saturated, and a frame that is three quarters one quadrant of the wheel at chroma 18.6 is exactly the case where flat colour blocks are the failure mode. A wheat field, a sunflower head or a cypress needs gradation inside its yellow rather than more yellow.
• Highlight −2 and Shadow −1. Velvia is hard at both ends, so both values buy normal contrast rather than flatness. The white point at 70.4 needs the bigger correction. The black at 22.1 is only mildly lifted, so one notch covers it.
• DR200. One stop of highlight protection for a subject that is frequently sunlit, without the second stop's shadow lift, which would take the black past the measured 22.1. The floor lands at ISO 250.
• Color Chrome FX Blue Weak, corrected from Off. The first pass measured cool content at 3.5% and concluded there was nothing to act on. The clean corpus measures 19.9%, and Van Gogh's blues are deep ultramarine rather than pale accents, so one notch of depth is right. Strong is still wrong, because it would pull the night skies toward navy and cost them their violet.
• Clarity +1 and Sharpness +1. Brushwork is dense small-scale variation, which is what a local-contrast operator flatters. The caution from Austurbane applies in reverse here: Clarity fringes on long hard edges, and this subject is foliage, fields and impasto rather than architecture. Keep it at +1 rather than higher, because a frame this saturated does not need the extra edge energy.
• Exposure 0 to +1/3. Yellow at high chroma clips earlier than the histogram suggests, because the green and red channels run far ahead of blue. Check the individual channels rather than the luminance histogram on anything with a sunflower in it.

Variants

Movie mode

Loses more than Open Air does. Movie mode drops Grain, Color Chrome Effect, Color Chrome FX Blue and Clarity, and three of those are doing real work here: Color Chrome Effect Strong keeps the dense yellows from flattening, FX Blue Weak holds the blues, and Clarity +1 is the brushwork stand-in. What survives is Velvia, the reversed white balance and the tone curve, which is the palette without the surface. Expect flatter colour blocks and grade for gradation in post if it matters. See Video / Movie Mode, settings, differences, and photo→video translation.

Validation tier: Creative design, field-test pending. Grounded in Fujifilm's published design statements for Velvia and for each base rejected above, in this bank's Dynamic Range & Tone Curve, the contrast controls tone-curve table and Color, Color Chrome Effect & FX Blue, the saturation/richness controls colour control notes, and in a CIELAB measurement with hue census across fourteen reproductions and a five-painter control group.

No frames have been shot with it.

Test priority: in order

1. A sunlit yellow subject, a field or a flower head. The whole recipe in one frame. Does the yellow hold gradation inside it, or has it gone to a flat block? Flat means Color Chrome Effect is not reaching it, and the answer is never more Color.
2. Check the red and green channels separately on that frame. Yellow at this chroma clips channel-wise while the luminance histogram still looks safe, and it is the most likely way to ruin an otherwise correct exposure.
3. A deep blue passage, a night sky or an iris. The setting that changed in correction, and therefore the one to check first. At CCFXB Weak the blue should read deep and violet-leaning rather than navy. If it has gone navy, drop to Off. If it reads washed, go Strong.
4. Anything with skin in it. The honest limit test, and the reason the variants table carries a portrait row. A green lean on skin is unflattering by design, and knowing where that line sits is worth one frame.
5. Dense foliage at 100 per cent. The Clarity +1 check. Look along any hard edge in frame, such as a roofline against sky, because that is the one place this subject can still fringe.

What would count as the recipe being wrong rather than needing a tweak: if the yellows clip before they reach chroma 19 at any usable exposure, then the measured target came from paint rather than from light, and the recipe has to settle for a lower-chroma yellow with the saturation carried by Color Chrome Effect alone.

Pro tip: This palette is three quarters one quadrant, and the remaining quarter is blue. Yellow through green is 76.3% of the reference corpus, blue-violet and cyan together are 19.9%, and magenta is 0.2%. Choose the subject accordingly: blue is a legitimate counterweight here, and a red or a pink is not. A pink in frame does more than sit outside the look. It takes the frame over.

Pro tip: The white balance goes the opposite way from almost every recipe here. Away from Red and away from Blue is a move toward green by way of cyan and yellow, because Fujifilm's grid has no green axis. If it looks like a typo against the rest of the bank, it is not.

Pro tip: Watch the channels, not the histogram. Saturated yellow drives red and green hard while blue lags, so the luminance histogram can look comfortable while two channels are already against the wall. Switch the camera to RGB histograms for this one.

Pro tip: Color 0 on Velvia is already extreme. The instinct on the most saturated corpus in this bank is to add Color, and the measurement says 18.6 rather than 30. Leave the headroom to Color Chrome Effect Strong, which adds depth instead of intensity.

Shoot JPEG. HEIF silently drops Clarity, which is the brushwork setting here. See Grain & Detail. Grain Effect, Sharpness, Clarity, High ISO NR.

Creative design (field-test pending). See the notes and the Knowledge base.

Part 0: method

Same instrument and same control group as Open Air, built in the same session: twelve painters measured in CIELAB on one day, with only the differences between them treated as findings. The reasoning for that design is in Open Air's Part 0, and it applies here without change.

The short version is that a single painter's reproductions cannot be read alone, because museum photography and centuries of varnish both leave a signature. A control group lets a result be attributed to the painter rather than to the pipeline.

Part 1: the corpus, and a correction

Sixteen Van Gogh oils, public domain, photographed cropped to the canvas, verified by eye against the manifest. The control group is eleven other painters gathered the same way.

This recipe was published once on a contaminated corpus and corrected the same day, and the correction changed a setting rather than only a number. The first harvest admitted framed canvases and works on paper, which understated the cool half of this palette by a factor of five.

The Off setting had been justified on the claim that there was nothing for the control to act on. That claim was an artefact. Anyone who has seen the night skies or the irises would have said Van Gogh has blue in him, and the clean measurement agrees.

One sampling caveat belongs at the top rather than buried at the bottom. The available open-access holdings lean toward the Arles and Saint-Rémy years, which is the period of the yellows. A corpus drawn from the Dutch years, when he was painting peasants in browns and greys, would measure differently and would not support this recipe. What is characterised here is the Van Gogh of the popular imagination, and the recipe is for that palette rather than for the whole career.

The images were fetched into a scratch workspace and are not in this collection.

Part 2: the measurement

Finding 1: the chroma

Mean chroma 18.6, the highest of the well-sampled corpora and roughly 1.6 times Monet's and 1.7 times Rembrandt's. For scale against this bank's other reference work, the Pink Chrome corpus measures 22.5 and was described there as high.

Finding 2: the palette is three quarters one quadrant, and the rest is blue

Yellow through green is 76.3%, and the cool families together are 19.9%. The comparison with Rembrandt is still the instructive one: both are yellow-led and nothing alike, since Rembrandt's yellow sits over red-orange with zero cool content while Van Gogh's sits over green on one side and ultramarine on the other.

Finding 3: the green lean lives in the highlights

Mean a* is negative at both ends but unevenly: −1.41 in the shadows against −2.78 in the highlights. Nine of the twelve painters measured are positive at both ends, because skin, earth and wood all are. A body of work that goes negative at all is unusual, and one that does it most strongly in its lit passages is what the reversed white balance exists to produce.

Part 3: the base

The requirement list is short and the first item is decisive.

1. Chroma 18.6 natively. Velvia gets there without Color being pushed far enough to distort hue on the way.
2. A curve that can hold a 22.1 black under a 70.4 white, which is a moderate range with both ends pulled in from where Velvia sits natively.
3. Nothing that mutes saturated subjects, which rules out one base outright.

Velvia takes it on the first requirement and survives the other two.

A note worth making, and a correction. Pink Chrome was built on Velvia two recipes earlier, and its argument leaned on Fujifilm's statement that Velvia adds magenta to blue. The first version of this page called that property inert here, on a measurement that showed almost no blue. The clean corpus shows 19.9% cool content, and Van Gogh's blues are ultramarine and violet-leaning rather than cyan, so a base that tilts blue toward magenta is rendering them the right way. The property that won Velvia one job turns out to be doing quiet work in the other.

Part 4: reaching green on a grid that has no green

Fujifilm's white-balance shift runs Red against Cyan and Blue against Yellow. There is no green axis, which is a fact this bank has used before in the opposite direction: Reading Room reaches magenta by moving toward Red and Blue together.

Green is the other diagonal. A step away from Red moves toward cyan, a step away from Blue moves toward yellow, and cyan plus yellow is green. So −1 Red with −2 Blue produces a mild green lean with strong yellow, which is the measured highlight a* of −2.78 against a highlight b* of +18.17.

The asymmetry is the design. Two notches off Blue against one off Red keeps the result on the yellow side of green rather than in the middle of it, matching a corpus where yellow at 36.2% leads yellow-green at 31.4%. This is a yellow that has travelled toward green, not a green that has been warmed, and the shift has to say which.

Part 5: the two settings that stop it becoming a poster

Color stays at 0 on the most saturated simulation Fujifilm makes. The measured target is 18.6, and Velvia at 0 is already past where most bases reach at +4. Pushing Color from there overshoots the corpus and, more importantly, spends the headroom that Color Chrome Effect needs.

Color Chrome Effect Strong is the one doing the work. It finds gradation inside colours that are already saturated rather than raising saturation, and a frame that is three quarters one quadrant of the wheel at chroma 18.6 is the textbook case for flat colour blocks. A wheat field has to have wheat in it rather than being a yellow shape. That is what CCE is for, and it only has room to act because Color was left alone.

Part 6: honest limits

• No frames have been shot with it.
• The corpus leans to the Arles and Saint-Rémy years. A Dutch-period corpus would measure brown and would not support this recipe. Stated at the top of Part 1 as well, because it is the limitation most likely to matter.
• These are reproductions, so the endpoints belong to a photograph of a painting rather than to pigment. Chrome yellow in particular is known to darken with age, so the measured yellows are if anything duller than what was painted.
• The green lean is hostile to skin and is strongest in the highlights, which is where a lit face sits. The variants table in the notes carries a portrait row that backs it out.
• A reference tells you what a look is and never whether a recipe reproduces it.

Part 7: where it sits in the bank

Fourth creative original on Velvia, after BewareMyVelvia, Fresh Install and Pink Chrome, and the four occupy different parts of the wheel on the same base: drama, blue-green, magenta and now yellow-green. Fresh Install is the one to check it against, since both are saturated outdoor looks that carry blue. That recipe runs CCFXB Strong for an impossible sky. This one runs Weak, because Van Gogh's blue is a counterweight rather than the subject.

It pairs with Open Air as the two painter-derived recipes built in one session from one control group, and they disagree about almost everything. Chroma 18.6 against 11.9, a black point of 22.1 against 34.9, shadows that drain colour at −14.50 against shadows that barely do at −2.24, Velvia against Astia. Two painters working outdoors in the same decade in the same country, measuring as opposites on every axis except one: both carry real violet, 11.2% against 26.9%.

Last updated: 2026-09-17 (corrected same day against a clean corpus)