Kodak UltraMax 400

Datasheet-derived original. Primary source: Kodak E-7023, February 2016 (Kodak Alaris). Archived locally at X-T5/_reference-sources/datasheets/kodak-still/ultramax-400_e7023.pdf, 4 pages, read in full. Secondary: the bank's one real UltraMax scan. Settings in the notes.

---

What the datasheet actually says

Property	Kodak's wording
Audience	"a worry-free, easy-to-use high speed film designed for snapshooters"
Core claim	"the flexibility you need to take consistently better pictures in more picture taking situations — better low-light picture quality with fewer underexposures, better results with zoom lenses, greater flash range, better 'stop-action' photos, and reduced impact of camera shake"
Sharpness & grain	"Even when enlarged, ULTRA MAX 400 Film delivers excellent sharpness and fine grain for crisp, clear pictures"
Colour & skin	"Optimized color precision technology provides consistently bright, saturated colors with accurate skin-tone reproduction for natural-looking people pictures"
Technology	Advanced KODAK T-GRAIN Emulsion; antenna dye sensitization; advanced development accelerators; optimized color precision technology
Benefits table	"Highest effective speed of any consumer 400 offering"; "Better underexposure protection"; "Long flash range"; "Fine grain and high sharpness"; "Great skin tones and superb color"
Compatibility	"It is printing compatible with KODAK GOLD Films"
Grain metric	Print Grain Index 46 (135 format, 4×6 in print, 4.4× magnification)
PGI scale	"A Print Grain Index rating of 25 on the scale represents the approximate visual threshold for graininess"; "a change for four units equaling a just noticeable difference"
Mixed light	"this film is tolerant of mixed-lighting situations" — plus full CC-filter tables for six fluorescent types and four HID types
Long exposure	"You do not need to make any exposure or filter adjustments for exposure times of 1/10,000 second to 1 second"
Process	C-41 (KODAK FLEXICOLOR)

The complexion-density table

Area measured on the negative	Density reading
KODAK Gray Card (gray side), same illumination as subject	0.80–1.00
Lightest step (darkest in the negative) of a KODAK Paper Gray Scale	1.20–1.40
Normally lighted forehead, light complexion	1.10–1.40
Normally lighted forehead, dark complexion	0.85–1.25

Kodak's caveat, which is part of the data and should never be dropped: "Because of the extreme range in skin color, use these values only as a guide. For best results, use a KODAK Gray Card (gray side)."

---

The central finding: PGI 46 is the number that settles the grain argument

The bank has argued about UltraMax's grain twice, the reference recipe runs Strong/Large + Clarity —, a 2026——audit flagged that as grittier than the emulsion spec, and AP's UltraMax moved to Strong/Small. All of that was reasoning from the phrase "fine grain." The datasheet also gives a number, and it points the other way.

Print Grain Index 46, on Kodak's own uniform perceptual scale where 25 = threshold of visibility and 4 units = one just-noticeable difference:

Film	PGI @ 4×6	JNDs above visible	Bank recipe grain
Ektar 100	< 25	0 — invisible at album size	Weak, Small
Portra 400	37	~3	Weak, Small
UltraMax 400	46	~5	Strong, Small

So UltraMax is about 2¼ JNDs grainier than Portra 400 at the same box speed, and roughly five JNDs above the point where grain becomes visible at all. It is the grainiest colour emulsion in this bank by a clear margin. Strong roughness is not an aesthetic choice; it is what the number says.

But the size axis goes the other way. The emulsion is T-GRAIN. Kodak's tabular crystal technology, the same family this bank documents for Ilford's Delta films in the B&W README. Tabular crystals are flatter and more regular than conventional cubic grain, which is why Kodak can claim "fine grain" at PGI 46 without contradiction: the grain is prominent but well-structured, not clumpy. That is exactly the distinction the X-T5's Grain Effect encodes as Roughness × Size. Hence Strong, Small, and it is why the reference recipe's Large overstates it. Large models clumping, which is a conventional-grain behaviour this emulsion does not have.

(One methodological caution, stated because it matters: PGI and RMS granularity are different scales and cannot be compared, Kodak says so explicitly. So UltraMax's 46 cannot be ranked against Pro 400H's RMS 4 or Acros's RMS 7. The comparison above only works because Ektar, Portra and UltraMax are all Kodak stills films quoted on the same PGI scale at the same print size.)

---

The second finding: this is a skin-tone film

Read the datasheet cold, without the drugstore-print associations, and the design intent is startlingly explicit:

"...consistently bright, vibrant colors with accurate skin-tone reproduction for natural-looking people pictures." "Great skin tones and superb color."

Kodak said it twice, once in prose and once in the benefits table. UltraMax was engineered as a people film, a snapshot film whose job was faces at birthdays, weddings and holidays, in whatever light was available. The "better low-light picture quality with fewer underexposures" pitch and the exhaustive fluorescent/HID filter tables are all in service of that.

The popular rendering. Crushed shadows, heavy yellow WB, Clarity —. Works against every one of those claims. It is not wrong as a look; it is a faithful model of a 1998 minilab 4×6 print, which had its own contrast curve, its own paper, and its own auto-correction. This recipe backs that pipeline out and keeps only the emulsion. The practical consequences are the WB (—/— rather than —, weighted toward red rather than yellow, because piling yellow on a skin-accuracy film defeats its purpose) and the tone curve (Highlight —/ — rather than —/—).

---

Bases considered

Classic Chrome: chosen

Two pieces of evidence, not habit:

1. Kodak states UltraMax is "printing compatible with KODAK GOLD Films." Print compatibility means the two emulsions are designed to land in the same place on the same paper, the same colour-balance target. This bank's Gold 200 is its most scan-validated recipe (~35 frames across all light conditions, WB and Clarity refined from evidence), and it sits on Classic Chrome. Building UltraMax on a different base would discard the strongest colour-balance evidence the bank owns for a Kodak consumer negative.
2. Classic Chrome's low saturation baseline makes +Color land on accents. This is the behaviour the one real UltraMax scan documents (below), and it is the film's actual character, a muted frame with one thing in it that pops, not a uniformly saturated picture.

Classic Negative: rejected

Structurally it looks like the better match: it is Fujifilm's consumer colour negative simulation, modelled on Superia, where Classic Chrome is modelled on no film at all (Fujifilm calls it "an imaginary film" based on printed magazine pages, see official design notes). Rejected for two reasons:

• Classic Negative carries a documented cyan-green shadow / magenta highlight signature. That is a Fujifilm consumer-negative fingerprint, and stamping it onto a Kodak emulsion imports the wrong manufacturer's colour philosophy. On skin specifically it doubles the sallow and ruddy risks, the same reasoning that rejected it for AP's UltraMax.
• The bank already has Superia 400 as the cool green-leaning 400-speed consumer negative. Putting UltraMax on the Superia-derived base would collapse a genuine Kodak-vs-Fuji distinction the bank exists to preserve.

Nostalgic Neg. / Pro Neg. Std: not seriously considered

Nostalgic Neg's amber highlights are a 1970s American-print look, further from the emulsion than Classic Chrome is. Pro Neg. Std is a studio portrait profile with soft gradations, the opposite of a punchy consumer film.

---

Recipe-vs-datasheet audit: this build against the other two

Setting	Reference recipe	AP's UltraMax	This (faithful)	What E-7023 supports
Grain	Strong, Large	Strong, Small	Strong, Small	PGI 46 → Strong ✅; T-GRAIN → —. Large overstates it
Clarity	+3	0	0	No micro-contrast, edge-contrast or acutance language anywhere in the sheet
Highlight	+1	−1	−1	"Better underexposure protection" + —; —is print punch, not emulsion
Shadow	+1	−1	0	Underexposure is the stated failure mode. +1 crushes it; −1 is a skin-tone interpretation; 0 is neutral
Color	+4	+3	+4	"bright, saturated colors" ✅ — and the scan confirms accent-pop at this level
WB	Auto, —/—	Daylight, —/—	Daylight, —/—	Warm ✅, but "accurate skin-tone reproduction" argues against a heavy yellow push
Sharpness	0	0	0	"excellent sharpness" says don't go negative; PGI 46 says don't go positive
DR	DR-Auto	—	—	"more picture taking situations", "fewer underexposures", full bad-light filter tables
CCFXB	Off	Weak	Off	Classic Chrome has magenta removed from skies by design

Two honest notes on that table. First, the reference recipe is right about Color —and this build follows it, that value has pixel evidence behind it, and copying an evidence-backed value is not the same as copying a recipe. Second, the reference recipe's deviations are not errors; they are a different target (the minilab print), correctly identified as such in the 2026-07-15 validation audit and preserved as a variant here.

---

What the one real scan actually establishes

The bank holds a single confounder-checked UltraMax frame (Wikimedia Commons, analysis-only; Pentax ME Super + SMC Pentax 50 mm f/1.4, Porthtowan Beach, Cornwall, flat overcast daylight, 35 mm, no styling). Recorded 2026-07-21 against the reference recipe; the findings apply to the emulsion, so they carry here.

What it evidences: - Warm lean overall: sand and cliff as warm buff/pinkish-tan, with a slight magenta lean in bright sand. Supports a warm WB; the magenta lean is why this recipe weights toward red rather than piling on yellow. - Accent behaviour: "the single saturated accent (an orange jacket) reads strong and clean without going neon, against an otherwise muted scene." This is the direct evidence for Color —on a low-saturation base. - Shadows hold detail in rock crevices, with a warm lean → supports —, contradicts —. - Highlights close to blowing but not clipping harshly → supports Highlight —+ — rather than a harder curve. - Greens muted and yellow-olive; grain visible and moderate, appropriate to a 400-speed consumer stock.

What it does not evidence, and must not be claimed to: it is one frame in flat overcast light. It says nothing about skin, the film's headline claim. Nothing about strong sun, and nothing about golden hour. Those are exactly the test-priority items in the notes. (A second candidate frame, a São Paulo interior, was examined and rejected as lit entirely by a TV screen, recorded as seen-and-rejected in the scan index.)

---

Where this sits in the bank

	Speed	Grain	Character	Evidence
Gold 200	200	Strong, Small	Warm sunny nostalgia; softer (Sharpness —)	~35 scans — best in the bank
UltraMax 400	400	Strong, Small (PGI 46)	Saturated, faster, accent-pop, skin-accurate	Datasheet + 1 scan
Portra 400	400	Weak, Small (PGI 37)	Professional restraint; latitude over punch	Datasheet + 1 scan
Ektar 100	100	Weak, Small (PGI <25)	Sharpest, most saturated C-neg	Datasheet + 2 scans
Superia 400	400	—	Cool green-leaning Fuji consumer 400	Datasheet + 2 scans

Gold and UltraMax are Kodak's own stated print-compatible pair, the same colour target at two speeds. Shoot Gold for nostalgia and softness, UltraMax for speed and pop. Portra is the professional opposite of UltraMax at the same box speed: the two bracket the entire C-41 400 category.

Validation status

Datasheet-validated + Scan (1 frame, flat overcast, inherited). Every setting traces to E-7023 or to the scan analysis, and the two places this recipe deviates from the popular build (grain size, tone curve) are argued against specific datasheet language above rather than asserted.

The gap is skin. The film's headline claim is skin-tone accuracy, this recipe is built to restore it, and there is no skin evidence at all, for this or for any colour recipe in the bank, which validation-methodology.md already names as the bank's largest structural gap. A colour portrait scan on UltraMax would be one of the two highest-value contributions anyone could make to this project. Drop it into references/.

Derived: 2026-07-25