North Light

Creative design (field-test pending). See the notes, the reference recipe it derives from, and the Knowledge base.

Part 1: What the subject actually demands

Interiors and architecture are the one genre in this bank where the photograph has a client with an opinion about accuracy. An architect looking at a frame of their building is checking whether the concrete is the grey they specified. That reframes the whole exercise: most recipes here are chosen for the colour bias their base was designed with, and this one is chosen for the absence of one.

Four requirements, in order of how often they are the thing that ruins the frame:

1. Range. A window against a wall, or a sky against a facade, is a bigger brightness ratio than any other common subject. This is the failure mode.
2. Neutrality. Whites white, greys grey, timber the brown it actually is. A cast that would read as character on a street photograph reads as an error here.
3. Material texture. Board-marks in concrete, brick courses, oak grain, brushed steel, the pile of a rug. Buildings are surfaces, and a frame with no micro-contrast has nothing to look at.
4. Blacks that stay black. The most recognisable amateur interior look is the lifted shadow, where every corner is grey and the room reads flat and synthetic. It is what happens when someone solves requirement 1 by lifting everything.

Requirements 1 and 4 pull against each other, and most of the design work is in the settlement between them.

Part 2: What the reference got right, and where it stops

Interior Negativ by @thomiegraphy is the starting point and it is a good recipe. Two of its ideas carry straight over:

• The tone curve is aimed at the window. Highlight −2 with Shadow −1 softens both ends on a base Fujifilm documents as hard-contrast. That is the correct instinct for requirement 1.
• Micro-contrast is the point. Sharpness +2 with Clarity +3 is the most aggressive pairing on a colour base anywhere in this bank, and on a still-life of a cane chair it is exactly right. It understands requirement 3 better than most recipes do.

Four things do not transfer from an editorial interior to architecture:

None of that is a criticism of the reference, which is not trying to document a building. It is the list of things that change when the subject does.

Part 3: The base, which is the whole recipe

Requirement 2 disqualifies almost the entire lineup, because a designed colour bias is exactly what these simulations are for.

Why Reala Ace lost, since it nearly won

On paper it is the best fit in the lineup. Its documented "deep, soft shadows" is requirement 4 and requirement 1 at once: gradation retained in the low end while keeping density, which is precisely what documenting a room needs. And its nonlinear compression, "the more saturated the subject, the more muted the colour", would automatically stop a single red sofa dominating a neutral space.

It lost on a single line in this bank's own notes: recipes on Reala Ace "shouldn't fight it with heavy Highlight softening", because the hard highlight end is half of what makes the simulation read true to life. Fighting the highlights is not optional for this subject. A blown window is the failure mode, and any interior recipe has to push against the top end hard. A base whose character is damaged by the one correction the subject requires is the wrong base, however well the rest of it fits.

That is worth recording rather than burying, because the Reala Ace argument is strong enough that someone will make it again.

Why Pro Neg. Std won

Fujifilm's own positioning does the work. Std is "optimized for use in situations with good lighting and moderate contrast, such as studio photography", and its parameters "apply straightforwardly", which this bank's notes summarise as the honest base where adjustments land predictably.

Three consequences:

1. A lit interior is controlled light, which is the condition Fujifilm names. The manufacturer's own rule for choosing between the Pro Neg pair is that it is a lighting decision rather than a taste one: controlled light takes Std, uncontrolled outdoor light takes Hi. That rule is why the exterior variant is Pro Neg. Hi rather than a different set of numbers.
2. Predictable parameters matter more here than anywhere else in this bank. Elsewhere the tone curve is building a mood and an approximate response is fine. Here it is solving a measurable brightness ratio, and it needs to do what it says.
3. No stated bias means nothing to correct for. Every other recipe here spends settings compensating for its base's designed character. This one spends none, which is why the white balance sits at 0 and 0 and the Color dial barely moves.

Pro Neg. Std was unused in this bank before this recipe. That is a gap rather than a coincidence: it is the least expressive simulation Fujifilm makes, which is exactly why nobody reaches for it and exactly why it suits this subject.

Part 4: The settlement between range and blacks

The two requirements that pull against each other get separate tools rather than a compromise on one.

Stated as a rule: on this subject, take dynamic range from the Dynamic Range setting and leave the shadow end alone. It is the opposite of how most recipes in this bank are built, and it follows from requirement 4 being a real deliverable rather than a preference.

Part 5: Texture without the tell

Clarity +1 and Sharpness +1 look timid next to the reference's +3 and +2, and the reason is specific rather than cautious.

Clarity works on local contrast, so its visible effect scales with how fast tone changes across an edge. A cane weave and a carpet pile are dense small-scale variation, which is where Clarity flatters. A roofline against a bright sky is a single hard edge across the frame, which is where it fringes. The reference's subject is entirely the first kind. Architecture is full of the second.

Sharpness gets the same restraint for a different reason: the X-T5 has no optical low-pass filter, so it already resolves close to Nyquist, and a given Sharpness value does more work here than the same number would on a OLPF camera. That is Sensor Architecture — X-Trans, Bayer, and what GFX actually is the bank's standing argument for keeping Sharpness low, and hard geometric edges are where over-sharpening announces itself first.

The grain change is the largest departure from the reference and the easiest to justify. Grain needs detail to sit in. Architecture supplies the opposite: broad flat planes of even tone. Strong and Large on a white wall is not film character, it is blotch.

Part 6: Where it sits in the bank

Both are worth having, and the recipe files say so in both directions. An interiors photographer shooting for a magazine wants the reference. One shooting for the architect wants this.

It is also the bank's first recipe on Pro Neg. Std, and the first written for a genre where accuracy is the deliverable rather than a constraint.

Part 7: Honest limits

• No frames. A designed look with no test shots. Everything here is reasoning from Fujifilm's published base behaviour and from the documented behaviour of the controls, which is the creative-design tier and nothing more.
• It inherits the reference's limit on people. Clarity on skin is texture nobody asked for, and the variant drops it to 0 with a Shadow lift, but this is not a portrait recipe and no variant makes it one.
• The neutrality claim is untested. Requirement 2 is the recipe's central promise and the white-wall frame in the test list is the only thing that can confirm it. If Pro Neg. Std turns out to carry a tint of its own, the argument above collapses and the base needs re-choosing.
• "Moderate contrast" is Fujifilm's phrase, not a measurement. Unlike the colour negatives in this bank, no characteristic curve exists for a film simulation, so the base's contrast is described rather than measured. The Measuring film curves from datasheets — method, calibration, and its limits curve work applies to film stocks, not to simulations.
• Exteriors are covered by a variant rather than by their own recipe. Pro Neg. Hi under uncontrolled light is Fujifilm's own rule and it should work, but a dedicated exterior build would probably differ in more than the base.

Last updated: 2026-08-09

An original design for interiors and architecture, built on the one base Fujifilm made for accurate colour under controlled light and which nothing in this bank had used. Named for the light architects and painters have always preferred: even, directionless, and honest about materials. Derivation in the notes.

Derived from Interior Negativ by @thomiegraphy, which is an excellent editorial interior look. This is not that. It keeps the two ideas that recipe got right, a tone curve aimed at the window and a heavy emphasis on material texture, and changes the four things that make an editorial look wrong for architecture: the colour cast, the halo risk, the grain on flat surfaces, and Auto white balance across a set.

Why Pro Neg. Std, and why nothing else

In architecture a colour cast is a defect rather than a style. The client is an architect or a designer, and the deliverable is a truthful account of a material: this concrete is this grey, this oak is this brown, this white wall is white. Every other base in the lineup carries a designed bias that becomes an error on that subject.

That last row is the whole argument. Fujifilm's own framing of the Pro Neg pair is that the choice between them is a lighting decision rather than a taste one: controlled light takes Std, uncontrolled outdoor light takes Hi. A lit interior is controlled light. And "parameters apply straightforwardly" matters more here than anywhere else in this bank, because the tone curve is being used to solve a measurable problem rather than to build a mood, and it needs to do what it says.

No recipe in this bank had used Pro Neg. Std before this one.

The window is the whole technical problem

Interiors and architecture share one failure mode: an enormous range between the bright thing and the room. A window against a wall, a sky against a facade. Three settings share the work.

• DR400 carries it. Two stops of highlight headroom is the largest single lever, and it pins ISO at 500, which costs nothing here. Architecture is tripod work, so a higher ISO floor is absorbed by a longer shutter rather than by noise, and ISO 500 is where the X-T5's conversion-gain step sits, making it the cleanest ISO in that region.
• Highlight −1.5, not −2. Settings are relative to the base. The reference recipe's −2 is a correction applied to Classic Negative, which Fujifilm documents as hard-contrast. Pro Neg. Std is already the moderate-contrast base, so it needs less. Going to −2 here buys little extra window and costs the room its snap. Use −2 only when the window genuinely dominates the frame.
• Shadow 0, and this is deliberate. The single most recognisable failure in amateur interior photography is the lifted-shadow look, where every corner is grey and the image reads flat and synthetic. DR400 already opens the low end, so an additional Shadow lift is what tips a clean image into that territory. Blacks stay black, and the room's shadow detail comes from the dynamic range rather than from the tone curve.

Texture without halos

Architecture is materials, so micro-contrast earns its place: board-marks in concrete, brick courses, the grain in oak, brushed steel, the weave of a rug. The reference recipe reaches Sharpness +2 with Clarity +3, and on a still-life of a chair it works beautifully.

It does not transfer to buildings, for a specific reason. Clarity is a local contrast operator, and architecture is full of hard high-contrast edges: a roofline against bright sky, a dark mullion against a lit window, a shadow line down a white wall. Local-contrast operators halo on exactly those edges, and a halo along a roofline is the tell that marks an image as over-processed.

• Clarity +1 gets material texture and stops short of the edge artefacts.
• Sharpness +1 rather than +2, and the reason is the sensor. The X-T5 has no optical low-pass filter, so it already resolves close to Nyquist and a given Sharpness value does more work here than the same number on a OLPF camera. Restraint is the house rule for that reason and it applies doubly to hard geometric edges.
• Grain Weak, Small. The largest change from the reference. Architecture is full of large, flat, evenly-toned areas: white walls, concrete slabs, clear sky. That is the worst case for large grain, which reads as blotchy noise rather than as film once it has nothing to sit on. Weak and Small keeps a film signature on the materials without dirtying the planes.

The rest

• White Balance fixed at Daylight, 0 and 0. The reference uses Auto with a cool shift, and for a single hero frame that is fine. Architecture is shot in sets: a room from six angles, a building through a day. Auto white balance follows the changing mix of window light and lamps, so the frames drift apart and will not cut together. Fixed is the professional requirement and it is White Balance & WB Shift — the colour-cast control the bank's own rule for colour-critical work. Zero shift because any shift is a cast, and a cast is the thing this recipe exists to avoid. The variants below cover artificial light.
• Color +1. Pro Neg. Std is faithful and low-saturation by design, and its parameters apply predictably, so this is a small lift to keep materials alive rather than a push. It stops at +1 because there is no clever compression on this base to protect a red door from a bigger number.
• Color Chrome Effect Weak. A grey interior has little saturated colour for a gradation control to act on, so Strong is mostly wasted processing. Weak still catches the one saturated element when there is one, a red door or blue signage, and stops it flattening.
• Color Chrome FX Blue Weak. Skies matter on exteriors, and Strong on this sensor risks a synthetic navy against a pale concrete facade.
• Exposure. Meter for the room at 0 to +1/3. Drop to −1/3 when a window or a bright sky occupies a large part of the frame, and let the tone curve and DR400 do the rest.

Variants

Validation tier: Creative design, field-test pending. There is no datasheet for an original look, so this is grounded in Fujifilm's published design statements for the Pro Neg pair and for the bases rejected above, in the documented behaviour of local-contrast operators on hard edges, in the Sensor Architecture — X-Trans, Bayer, and what GFX actually is sensor architecture notes on the missing low-pass filter, and in the reference recipe it departs from.

No frames have been shot with it.

Test priority: in order

1. A room with a window in frame. The whole recipe. Does DR400 with Highlight −1.5 hold any window detail, and does the room still have blacks? If the window is gone, −2 is the next step. If the room looks flat, the problem is Shadow and it should go positive rather than negative.
2. A white wall. The neutrality test, and the one an architect will apply. Is it white, or has it picked up a tint? If it has, that is the base failing at its one job and the recipe needs rebuilding rather than tuning.
3. A roofline against bright sky. The halo test for Clarity +1. If there is a bright fringe along the edge, go to 0 and take material texture from Sharpness alone.
4. A large flat surface at 100 per cent, concrete or a plain wall. Confirms Grain Weak/Small stays clean where the reference's Strong/Large would blotch.
5. The same room on this and on Interior Negativ. The clearest way to see what neutrality costs and what it buys, which is the honest way to choose between them.

What would count as the recipe being wrong rather than needing a tweak: if materials read lifeless at Color +1 and only come alive past +3, then Pro Neg. Std is too flat to carry a look and the answer is Reala Ace with the highlight problem solved by DR alone, rather than more Color on this base.

Pro tip: In architecture a cast is a defect, not a style. Every other recipe in this bank is chosen partly for the colour bias its base was designed with. This one is chosen for the absence of one. If you find yourself adding a white balance shift to make an interior look better, you are making an editorial image, which is fine, but use the reference recipe for it instead of bending this one.

Pro tip: Clarity haloes on the edges architecture is made of. It is a local-contrast operator, so it does its most visible work where tone changes fastest: rooflines, mullions, shadow lines on white render. That is why this recipe sits at +1 where its editorial parent sits at +3, and why the detail variant is safe to push, because a close-up of brick has no long hard edge to fringe.

Pro tip: Fix the white balance before the shoot, not after it. Architecture is shot in sets, and Auto white balance follows every change in the window-and-lamp mix, so a room shot from six angles comes back as six slightly different rooms. Pick Daylight, Incandescent or a Custom reading, lock it, and keep it for the whole space.

ISO 500 is the sweet spot, not a compromise (measured 2026-07-27). DR400 pins the ISO floor at 500, which is exactly where the X-T5's conversion-gain step sits, so it is the cleanest ISO in that region rather than a tax. On a tripod the floor costs nothing at all: it is absorbed by a longer shutter. See Dynamic Range & Tone Curve — the contrast controls.