AP's Kodachrome

An original targeting Kodachrome 64, the stock that made the world's picture of India. Retuned so its famous contrast falls on the scene rather than on Indian skin. Built for Mumbai. Derivation, the datasheet findings, and the lineage in the notes.

This is the third of the AP set, and the one with real ancestry here: Raghubir Singh, who effectively invented Indian colour photography with Bombay as a central subject, shot Kodachrome. So did Steve McCurry across India. Kodachrome is not a foreign look imported onto Indian subjects. It is the film that taught the world what Indian colour looks like.

The design principle

Kodachrome's drama should come from the scene going dark, not from the skin going dark.

That single sentence is the whole recipe. Kodachrome is a colour transparency film. Contrasty by nature, with deep blacks and no negative-film latitude to rescue you. That contrast is the point. Strip it out and you have Ektachrome. But Classic Chrome, the base every Kodachrome recipe uses, has a documented hard shadow end (Fujifilm's own words. The second-hardest tonality of any simulation), and Indian skin, per Kodak's own complexion-density data, sits lower on the tone curve than lighter skin does, right where that hardness bites.

So the fix isn't to flatten the recipe. It's to lift the skin clear of the compression and let the darkness fall below it, in genuine shadow, where it belongs. Hence Shadow −1 where the standard Kodachrome 64 recipe runs +0.5, a 1.5-notch swing, and the single most important change here.

Note this is less lift than AP's Portra's fallback uses (−2). Deliberately. Portra is a low-contrast negative that wants open shadows. Kodachrome is a high-contrast slide that wants deep ones. −1 clears the face and keeps the blacks.

Why the other values

• Classic Chrome, kept. Unlike AP's Portra. For Portra I moved to Reala Ace because a low-contrast negative needs soft shadows and Classic Chrome couldn't give them. Kodachrome is the opposite film: contrast is its identity, so Classic Chrome's hardness is a feature to place correctly, not a flaw to escape. Its other traits fit too, the lowest saturation baseline of any simulation matches Kodak's own description of Kodachrome as reproducing "subtle color naturally" (see below), and Fujifilm's official shooting advice for this base, "underexpose the shadows while preserving the highlights" Is slide-film technique, which is a genuinely lucky convergence.
• Highlight 0, not negative. Kodachrome is a slide: highlights clip, and that is faithful. Classic Chrome's highlight end is already soft, so pulling it down further would sand off the snap this film is supposed to have. DR200 supplies enough headroom for Mumbai without turning the top end milky. This is a deliberate break from AP's Portra, which runs Highlight −2 because Reala Ace has hard highlights that needed correcting. Different base, opposite problem.
• DR200. Slide-film latitude is narrow. DR400 would be more forgiving but would flatten exactly the contrast being preserved. DR200 is the honest middle, and it is what the existing Kodachrome 64 recipe uses. Floors ISO at 250.
• Color +2, CCE Strong. Kodak describes both Kodachrome 25 and 64 as reproducing "subtle color naturally": the hypersaturated reputation is largely about how the film was projected, plus the way it renders reds. So the saturation is moderate, and the richness comes from Color Chrome Effect Strong instead. That matters on this base: Classic Chrome has "suppressed magenta", which pulls reds toward brick and terracotta, a genuine mismatch with Kodachrome's most famous signature, deep rich reds. CCE adds density inside already-saturated colour, so it restores the red without inflating the whole frame. It also enriches "skin in warm light," per Fujifilm. Two wins from one setting.
• WB Daylight +3R/−3B. The existing Kodachrome recipes run −4B and −5B, a heavy yellow push. Maharashtra skin commonly carries olive undertones (yellow + green), and piling yellow onto olive is the direct route to sallow. So: −3B, warmth re-weighted toward red at +3R. That +3 is higher than AP's Portra's +2 for two reasons. Classic Chrome's documented cool shadows genuinely need the correction (Reala Ace didn't), and Kodachrome's warm richness is part of its identity. Fixed Daylight, because a colour-critical slide emulation must be repeatable.
• Grain Weak, Small. Datasheet: Kodachrome 64 is RMS granularity 10, Kodachrome 25 is RMS 9: "extremely fine grain" in Kodak's words. Among the finest in the bank.
• Sharpness +1, Clarity 0, and this is a real, named compromise. "Extremely high sharpness" appears in the datasheet for every Kodachrome, so it is definitional. But sharpening and mid-tone micro-contrast are both unkind to faces. The split: Sharpness models genuine optical acutance, which is what the datasheet is actually claiming, so it stays positive. Clarity models mid-tone local contrast, a different thing, and the setting that costs skin. The standard Kodachrome 64 and 25 recipes both run Clarity +3. Dropping to 0 is the second flagged fix, and it also makes this recipe HEIF-safe and fast-saving. For frames with no faces in them, +2 is there in the variants.
• Expose 0 to +1/3, metering the face. Slide film wants exposure discipline and slight underexposure deepens saturation, but darker skin needs light, and a muddy face is a worse failure than a blown sky. Meter the face and accept that a bright Mumbai sky may go. That is exactly the trade Kodachrome shooters made. McCurry's portraits are full of it.

Variants

Handling

• Meter the face, not the scene. This is the recipe's one non-negotiable technique. Slide film discipline says protect the highlights. Darker skin says protect the face. When they conflict, and in Mumbai midday they will, the face wins. A blown patch of sky is a normal Kodachrome photograph. A muddy face is a failed one. McCurry's portraits are full of exactly this trade.
• Shoot it in hard light. Kodachrome 64 is a daylight transparency and the recipe inherits that. Its design condition is direct sun with distinct shadows. Kodak's own daylight table for K64 lists 1/125 at f/11 in bright or hazy sun. Flat overcast is where this look goes limp. That's what the overcast variant is for, and even then AP's Portra or Mumbai Monsoon will usually serve better.
• Don't stack warmth on warmth. WB +3R/−3B is already a strong warm bias. In golden hour, drop Color to +1 rather than adding more, the light is doing the work, and Classic Chrome's terracotta lean will tip orange-heavy if you push both at once.
• Fixed Daylight WB, always. Colour-critical slide emulations need repeatability. If you find yourself reaching for Auto WB, you want AP's Portra, that recipe is built to float with the light. This one is not.
• Expect the sky to be pale, and don't fight it. Fujifilm explicitly removed the magenta component from Classic Chrome's sky blues, so skies here render airy rather than deep. That is a base-level design decision. No amount of Color, WB shift or Color Chrome FX Blue recovers a Velvia sky without wrecking everything else, if a frame genuinely needs deep blue, change simulation.
• Clarity is 0, which also makes this fast. Non-zero Clarity slows shot-to-shot saving on the X-T5 and blocks HEIF. Keeping it at 0 is a portrait decision first, but it makes this the more practical street recipe of the AP set as a side effect. The street variant raises it only because there are no faces to pay for it.
• Where it sits in the set: AP's UltraMax is for when the city is the subject. AP's Portra is for when a person is. This is for a person in the city: the documentary middle, where both have to hold.
• Video: transfers as Film Sim + WB + DR (manual) + Tone Curve + Color. Grain, Color Chrome Effect and Clarity are all unavailable in movie mode: and CCE is doing real work here (it is the entire red-restoration strategy), so a video version of this recipe will render reds noticeably flatter than the stills version. Compensate with Color +3 and accept it is an approximation. See Video / Movie Mode, settings, differences, and photo→video translation.

Test priority: in order, and what each frame is actually asking

This recipe has never been shot on Indian skin, which is its entire premise. These are the frames that would settle it, ranked by how much each one would change.

1. A face in bright midday sun, the Singh condition. Does Shadow −1 clear the face? The single most consequential setting here is Shadow −1 against the reference recipe's +0.5. The whole argument is that Classic Chrome's documented hard shadow end compresses precisely where darker skin sits, and that −1 lifts it clear. If the face still reads muddy or the cheek-to-jaw modelling collapses, go to −1.5 before touching anything else. Do not start adjusting Color, WB or DR, because they are not the problem. If −1.5 also fails, the base is wrong for the subject and the honest move is Reala Ace, as in AP's Portra.

2. A Mumbai midday sky. Does Highlight 0 hold? Highlight sits at 0 deliberately. Kodachrome is a slide, highlights clip, and that is faithful. Classic Chrome's highlight end is already soft, so pulling it down would sand off the snap. But Mumbai's midday dynamic range is brutal and DR200 is narrow. If skies are going beyond recovery on ordinary frames, the escape hatch is DR400 + Highlight −1, in that order. Accept that this costs slide snap. It is a real trade, not a free fix.

3. A red garment. Sari, sindoor, marigold. Do the reds land, or does the base win? This is the one place the base genuinely fights the film. Classic Chrome has a documented suppressed magenta that pulls reds toward brick and terracotta. Kodachrome's most famous signature is deep, rich red. CCE Strong is the entire counter-measure. If reds still land brick rather than blood, that is Fujifilm's design decision beating the setting, and per their own line ("no matter what you do in Velvia mode, it will never turn into Fortia") the honest answer is a different simulation, not more Color. Raising Color past +3 will inflate the whole frame before it fixes the red.

4. The same face on this and AP's Portra, back to back. These two are deliberately opposite solutions to the same problem. Keep the hard base and place the skin above it, versus escape to a soft base entirely. One comparison frame is worth more than a dozen single-recipe tests, because it isolates the base decision from everything else.

5. Golden hour on Marine Drive, and a festival frame. Lower priority, but they check the two variants that change the most values. Golden hour tests whether WB +4R/−4B tips orange-heavy. A Ganpati or Holi frame tests whether Color +3 holds together when the scene is already saturated.

What would count as the recipe being wrong rather than needing a tweak: if faces read well only at Shadow −1.5 or lower and reds refuse to leave brick even at CCE Strong, then Classic Chrome is being asked to do two things it is designed not to do, and the recipe should be rebuilt on another base rather than tuned further.

Validation tier: Original design. Field-test pending. Grounded in the Kodak E-55 datasheet (RMS 9/10, "extremely high sharpness," "reproduces subtle color naturally", K-14 transparency process), Fujifilm's published tonality and colour-design data for Classic Chrome, the complexion-density findings inherited from the Portra and UltraMax datasheets, the Mumbai 11-image place study, and the documented Singh/McCurry lineage. Not yet shot on Indian skin.

An honest gap specific to this recipe: unlike Portra and UltraMax, the Kodachrome datasheet contains no complexion-density table. Those exist for colour negatives, where you judge printing density, not for transparency film. So the skin reasoning here is inherited from the other two Kodak sheets plus Fujifilm's base-curve data, not drawn from Kodachrome's own paperwork. The logic holds (where skin sits on a tone curve is a property of skin and the curve, not of the film stock), but it is one step further from the source than the other two AP recipes. Worth knowing.

What to check first, in order. (1) Does Shadow −1 clear the face, or does Mumbai's contrast still bury it? Go −1.5 before touching anything else. (2) Is Highlight 0 holding, or is midday sky gone beyond recovery? DR400 + Highlight −1 is the escape hatch. (3) Do the reds land. Sari red, sindoor, marigold, or does Classic Chrome's suppressed magenta still pull them to brick despite CCE Strong? If so, that is the base fighting you, and the honest fix is a different simulation, not more Color.

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

Why this recipe exists

Flagged as the obvious next build at the end of AP's Portra. The argument there: Portra is the right film for people, but if the question is what the greats actually shot for India in colour, the answer is Kodachrome, and the bank's existing Kodachrome recipes carry two settings that are actively wrong for Indian skin.

The lineage: this one isn't borrowed

Most film emulations applied to Indian subjects are a foreign look imported onto a foreign subject. Kodachrome isn't.

Raghubir Singh (1942-1999) is regarded as the pioneer of colour photography in India, working from the early 1970s when serious photography was assumed to be black and white and colour was still a marginal art form. Bombay was one of his central subjects. He shot Kodachrome, handheld 35mm, documenting working-class daily life, festivals, rituals and weddings.

His reasoning for colour is worth recording, because it is effectively a design brief for this recipe: he held that only colour could capture Indian reality, partly because the streets "pulsate with the entire chromatic spectrum," and more because "the somberness of black is shunned by Indians, and the dark, guilt-ridden ethos of Western art is alien to their joyful vision of life." He was influenced by Cartier-Bresson and broke with him precisely on this point.

One technical detail matters for the recipe: Singh worked with Kodachrome 25, the slowest and finest-grained of the family, and had to reconcile that with his need for precise depth of field. ISO 25, handheld, stopped down. That tells you his design condition was bright Indian daylight, and it is why the K25 variant in the notes exists.

Steve McCurry's India and South Asia work, including the most recognised colour portrait of the twentieth century, is Kodachrome 64, the more practical documentary standard, and the primary target of this recipe.

What the datasheet actually says (Kodak E-55, archived)

The combined Kodachrome 25/64/200 sheet, and it corrects a popular myth:

Kodak's own daylight exposure table for K64 puts "bright/hazy sun, distinct shadows" at 1/125 at f/11, roughly half a stop more generous than sunny-16 at ISO 64, and an useful reminder that Kodachrome shooters were not underexposing by default the way slide-film folklore suggests.

Three constraints in that table matter more than they look:

1. It cannot be pushed. Kodak declines to recommend push processing for either 25 or 64, only the 200 pushes. So the ISO ceiling on any faithful Kodachrome recipe is a design property, not a taste call.
2. It is a daylight film, absolutely. Under tungsten, K64 drops to EI 16 with a 80A filter. There is no honest night version of this film.
3. It fails at long exposure. Beyond 1 second, Kodak says "not recommended" outright. Combined with (2), that closes off night and astro work entirely, a real limit, and the reason this recipe's night row sends you elsewhere rather than offering a variant.

"Reproduces subtle color naturally" is the line that should change how this is built. Kodachrome's reputation as a saturation monster comes mostly from how it was projected, a backlit transparency at 5000 K is a different experience from a print, and from the specific way it renders reds. Kodak's own copy describes a film aiming at natural, subtle colour with extreme sharpness and fine grain. So this recipe keeps Color at +2 and sources richness from Color Chrome Effect instead of cranking saturation.

Two further useful facts from the sheet: the film is a colour transparency (reversal), which drives the contrast and latitude decisions below. And Kodak's own stated replacements after discontinuation are Ektar 100 and Ektachrome E100, both of which exist as datasheet-derived originals in this bank (Ektar, E100), making them the natural A/B partners for anyone who wants to hear Kodak's own second opinion.

The two settings that were wrong, and why

The bank's existing Kodachrome 64 and Kodachrome 25 recipes are good recipes. Both carry two values that work against darker skin:

Neither change is a criticism of the source recipes. They were built as general-purpose daylight looks, and Clarity +3 is doing real work there, creating an etched crispness that suits landscape and street. They just aren't portrait recipes, and nobody claimed they were.

Where the complexion evidence comes from: and the honest gap

The reasoning below is inherited, not native, and that distinction should be stated:

• Kodak Portra 400 (E-4050): grey card 0.77-0.87. Light-complexion forehead 1.08-1.18 (+0.31 above grey). Dark complexion 0.93-1.03 (+0.16).
• Kodak UltraMax 400 (E-7023): grey card 0.80-1.00. Light 1.10-1.40. Dark 0.85-1.25.

Both show darker skin rendering at roughly half the separation from mid-grey that lighter skin does, i.e. Sitting in the lower-mid tonal zone.

The Kodachrome sheet has no equivalent table. Complexion-density targets exist for colour negatives, where a printer judges density to set exposure. They aren't published for transparency film, which is viewed directly. So the skin argument here rests on the other two Kodak sheets plus Fujifilm's published curve data for Classic Chrome.

That is defensible, where skin sits on a tone curve is a property of the skin and the curve, not of which film stock is being emulated, but it is one inferential step further from the source than AP's Portra or AP's UltraMax, and it should not be presented as if Kodak published it about Kodachrome. Flagged in the recipe too.

Why Classic Chrome stays: the deliberate contrast with AP's Portra

For AP's Portra I moved off Classic Chrome to Reala Ace. Doing the same here would be pattern-matching rather than thinking.

Three more reasons Classic Chrome is right specifically for Kodachrome:

1. Lowest saturation baseline of any simulation. Matches "reproduces subtle color naturally" better than any other base would.
2. Fujifilm's own shooting advice for it is "underexpose the shadows while preserving the highlights", which is slide-film technique, verbatim. The base and the film want the same discipline.
3. The bank already documents why the pairing works: "a coincidence of character (subdued palette, suppressed magenta, printed-page contrast), not shared ancestry." Classic Chrome is based on no analog film at all. Fujifilm calls it "an imaginary film" modelling printed magazine pages, so this is an approximation by character, and should always be described that way.

The one place the base genuinely fights the film: Classic Chrome's "suppressed magenta" pulls reds toward brick and terracotta, and Kodachrome's most famous signature is deep, rich red. That is why CCE Strong is non-negotiable here. It adds density inside already-saturated colour, restoring red richness without inflating the whole frame. If field-testing shows reds still landing brick rather than blood, that is the base's design decision winning, and per Fujifilm ("no matter what you do in Velvia mode, it will never turn into Fortia") the honest answer is a different simulation, not more Color.

The AP set: three recipes, three mechanisms

All three drop Clarity to 0 and lift shadows relative to their source recipes. That convergence isn't laziness. It is the same two findings (darker skin sits lower on the curve. Clarity punishes skin) applied to three different films, and the rest of each recipe diverges sharply because the films do.

Validation status

Original design, field-test pending. Grounded in:

1. Kodak E-55 datasheet (archived). RMS 9/10, extreme sharpness, "reproduces subtle color naturally," K-14 transparency process, Kodak's own suggested replacements.
2. Complexion-density data inherited from Kodak E-4050 (Portra 400) and E-7023 (UltraMax 400), explicitly not from the Kodachrome sheet, which has no such table.
3. Fujifilm's published tonality and colour-design data for Classic Chrome (asymmetric curve, suppressed magenta, lowest saturation, official shooting advice).
4. The Mumbai 11-image place study.
5. The documented Singh / McCurry lineage, including Singh's stated reasoning for colour and his use of Kodachrome 25.

Not yet shot on Indian skin. The full ranked test plan, with what each frame is actually asking and what result would mean the recipe is wrong rather than merely untuned, is in the notes. In short: (1) a face in bright midday sun. Does Shadow −1 clear it. (2) a Mumbai midday sky. Does Highlight 0 hold. (3) a red garment. Do the reds land or does the base's suppressed magenta win. (4) the same face on this and AP's Portra back to back.

How this relates to the faithful Kodachrome 64 original

This recipe is an interpretation, not an emulation, and the bank now holds both. kodachrome-64 is the datasheet-faithful build. Neutral, no Indian-skin tuning, derived from E-55 on its own terms. The two are meant to be read together:

If a frame disappoints on this recipe, shoot the same scene on the faithful one before changing values. It separates "the interpretation is wrong" from "Classic Chrome cannot do Kodachrome", two different problems with two different fixes.