What Fujifilm Is Actually Selling

Most manufacturers sell sensor specifications, autofocus speed and frame rates. Fujifilm has all of those, but what it actually sells is something else: the colour processing in the body that decides what the JPEG looks like.

That sounds like an optional extra. What it really changes is workflow.

Why “straight out of camera” matters

The conventional digital workflow is: shoot RAW, import, adjust, export. It gives you the highest quality and the most control, and it charges you for both — every frame arrives owing a debt of editing. Plenty of people’s enthusiasm for photography dies inside an import folder that never empties.

Fujifilm’s proposal is the opposite: decide the look before the shutter, and the JPEG that comes out is the finished picture. What you see in the viewfinder is what you get.

Two real things follow.

1. You shoot more, because nothing is waiting for you afterwards.
2. The decisions move forward in time. You are forced to work out what you want this frame to look like before you press the shutter — and that, on its own, makes you better.

Which is why the fifteen chapters before this one are about composition and light. When the processing moves ahead of the shutter, your judgement in the field is the whole of it.

Why in-camera processing is a real thing

There is a common objection here: straight-out-of-camera is just the camera applying a filter, no different from a phone app. Dismantling that requires knowing how a digital photograph is actually made.

A sensor does not see colour. It is a grid of wells that count photons, each covered by a filter that admits only red, green or blue. Every pixel site therefore measures one third of the information, and the other two thirds are inferred from its neighbours. Getting from there to a viewable image takes at minimum:

1. Demosaicing — reconstructing full red, green and blue at every site from the single-colour grid.
2. White balance — deciding which white counts as white, and correcting the overall cast to match.
3. A colour matrix — converting what the sensor measured into a colour space matched to human vision. The filters’ spectral response is not the same as the eye’s cone response, so this translation is unavoidable.
4. A tone curve — deciding how the range from darkest to brightest gets distributed. This is what gives an image its contrast and its density.
5. Local hue and saturation work — nudging particular hue ranges sideways, lifting the saturation of some colours and not others.
6. Sharpening and noise reduction.

Here is the point: steps 3 through 5 have no single correct answer. The sensor measures a physical quantity; a person perceives a psychological one; the mapping between them is designed, not calculated.

So every digital camera — your phone included — is interpreting colour. Only two things vary: who tuned the interpretation, and whether you get to choose it in the field.

Fujifilm’s answer: it was tuned by a company that has been making film for ninety years, and it is on a dial.

X-Trans: even the filter layout is different

Most sensors use a Bayer array: red, green and blue repeating in a fixed 2×2 unit. That regularity has a side effect. When the scene contains fine detail close to the pixel pitch — shirt weave, distant brickwork, a screen’s dot pattern — regularity meets regularity and produces moiré, coloured banding that was never in front of the lens.

The industry’s usual fix is an optical low-pass filter in front of the sensor: blur the image slightly so no detail is fine enough to fight with the grid. Moiré goes away, and so does some resolution.

X-Trans takes the other road. Its 6×6 layout is irregular, with green in every row and every column, which breaks the periodicity. No periodicity, no moiré — and therefore no low-pass filter, so the detail survives.

Two costs come with that, and both are worth knowing:

• Demosaicing is harder. An irregular array needs different algorithms. For years third-party RAW converters handled X-Trans badly, turning foliage and grass into a smeared, waxy texture that photographers took to calling “watercolour”. Mainstream software has largely caught up, but this is a historical reason X-Trans owners are unusually willing to convert in-camera: the camera understands its own array.
• Fujifilm’s colour work was tuned against this sensor, which is why an in-camera JPEG and a third-party “Fuji-look preset” applied on a computer do not land in the same place.

Three workflows, three outcomes

The same camera gives you very different things depending on how you set it up. All three of these are legitimate; which one you want depends on what you are going out to get.

One: JPEG only

Record JPEG alone, choose the simulation in the field, send to your phone when you are done.

Good for travel, for daily shooting, for anything social. Cards last, transfers are quick, the pictures exist the same day. The cost is that there is no way back: if the white balance misread the scene or you chose the wrong simulation, that frame is what it is.

It also quietly raises the stakes on backups. With RAW+JPEG a corrupted card still leaves you a negative to work from; with JPEG only, the file on the card is the entire photograph. If you shoot this way, get the frames off the card the same evening rather than letting a trip accumulate on one piece of plastic.

Two: RAW + JPEG — the default recommendation

Record both. Use the JPEG as the finished picture and keep the RAW as insurance.

IMAGE QUALITY SETTING → IMAGE QUALITY → FINE + RAW. Card consumption goes up roughly four- to five-fold, which is a much smaller price than one photograph you cannot get back.

Three: RAW-first

Worth doing when you already know the frames are going to a computer, or when someone else will ask for deliverables.

Here the film simulation becomes a previewing tool: you still use it to judge the scene, but the final colour gets rebuilt later. Nothing is wrong with working this way. You are simply not using the part of the camera you paid extra for.

X-T5 versus X-E4

You own both, so the differences are worth stating properly. They are not just a pixel count.

	X-T5	X-E4
Sensor	X-Trans CMOS 5 HR, approx. 40.2 MP	X-Trans CMOS 4, approx. 26.1 MP
Released	2022	2021
Controls	Three physical dials (ISO / shutter / exposure compensation)	Pared back, mostly menus and a command dial
In-body stabilisation	Yes	No
Screen	Three-way tilt	Flips up for selfies
Film simulations	Includes Nostalgic Neg.	No Nostalgic Neg.
Character	A tool you take out to shoot	A camera you happen to have

The image-quality menus are nearly identical. Compared item by item, the X-T5 has exactly two entries the X-E4 lacks: SELECT JPEG/HEIF and SMOOTH SKIN EFFECT. Everything else, from FILM SIMULATION down to EDIT/SAVE CUSTOM SETTING, is present on both and in the same place.

So almost everything in the next four chapters applies to both bodies. The one real exception is Nostalgic Neg., which arrived after the X-E4 shipped. Where later chapters use it, an X-E4 can approach it with Classic Neg. and a little warmth in the highlights — close, not the same.

One of those two extra entries deserves a word, because you will meet it in the menu and wonder. SELECT JPEG/HEIF lets the X-T5 write HEIF files instead of JPEG. HEIF carries more tonal precision and compresses better at comparable quality, which sounds like a straightforward upgrade and mostly is not — support outside Apple’s ecosystem is still patchy, and plenty of upload forms will simply reject the file. The tonal headroom matters least in exactly the workflow this chapter is about, where the point is a finished picture you can send immediately. Leave it on JPEG unless you have a specific reason, and know that the reason exists.

The “character” row is not mysticism. The X-E4 is light, flat and unobtrusive, so it leaves the house, and a camera that leaves the house is the one that takes pictures. The X-T5 has stabilisation and 40 megapixels, and it is heavier, so it tends to come out only on days you have already decided to photograph. That difference usually affects your results more than the sensor does.

Common failures and how to diagnose them

Straight-out-of-camera goes wrong in a small number of repeatable ways.

The whole frame has a cast, and changing simulation does not fix it. That is white balance, not simulation. The two live at different stages: white balance sets the reference white (step 2), the simulation interprets what comes after it (steps 4 and 5). If the reference is wrong, no interpretation downstream will be right. Fix white balance first, then choose a look.

A high-contrast simulation and the highlights are gone. Simulations do not blow highlights; exposure does. A contrastier rendering pushes highlights that were already near the ceiling straight through it, so under hard light you want to make negative exposure compensation a reflex. This is also why the dynamic-range settings exist, which chapter 19 takes on directly.

Two frames of the same scene come back noticeably different. Usually auto white balance re-deciding because the framing changed — one more red sign in the corner, one less patch of sky, and it gives a different answer. Pin it down in IMAGE QUALITY SETTING → WHITE BALANCE: pick the preset that matches the light, or dial in a colour temperature directly. This matters most when frames will be seen together, because inconsistent colour reads worse side by side than inaccurate colour does.

Other people’s Fuji photographs look better and no simulation closes the gap. Check whether you are comparing the same light. A simulation decides how light that already entered the camera is interpreted; it cannot manufacture light that was not there. Almost all of the difference in the photographs that look like they have something is the hour and the direction of the light — chapters 5 through 7 of this site — not the settings. The same parameters at noon and an hour before sunset are not comparable results.

A JPEG came out wrong, so it needs reshooting. Not if you kept the RAW. In-camera conversion lives in the playback menu (▶) as RAW CONVERSION — not under the shooting-side IMAGE QUALITY SETTING, which is why people fail to find it the first time. Faster: bring the RAW up in playback and press Q. You can reassign the simulation and every parameter, save a new JPEG, and leave the RAW untouched.

A film company that makes cameras

Fujifilm was founded in 1934 to make film. The simulation names are not marketing adjectives; they are products the company actually sold:

• Velvia — a 1990 transparency film known for extreme saturation and contrast, the landscape photographer’s signature.
• Provia — the neutral transparency in the same family, Fujifilm’s own idea of standard.
• Astia — lower saturation, gentle skin tones, made for portraits.
• Acros — black-and-white negative film, prized for fine grain and long-exposure behaviour.
• Superia — the consumer colour negative that filled every high-street print shop.
• Eterna — motion picture stock, low contrast and wide tonality, built to leave room for grading.

So Fujifilm is not, strictly, imitating someone else’s film. It is translating its own formulations into digital processing — a position no other camera manufacturer is in.

Why a film formulation translates at all

This works because film and digital processing are structurally parallel.

A colour film has three emulsion layers, sensitive to red, green and blue. Each layer has its own speed, its own contrast, and its own way of responding from shadow to highlight — all of which a manufacturer plots as a characteristic curve. The differences between stocks live in the shapes of those curves: Velvia’s are steep, which is where the contrast and saturation come from; Eterna’s are shallow, which is where the latitude comes from. On top of that, chemical interactions between layers determine which colours get pushed where.

Put that in digital terms: the characteristic curve is the tone curve, and the inter-layer behaviour is the colour matrix plus local hue work. Steps 4 and 5 from the beginning of this chapter.

Fujifilm is therefore not inventing a look from nothing. It is porting something its laboratories already had numbers for. This is also why the simulations do not substitute for one another — they correspond to different curve shapes, not to different strengths of one filter.

The turning point was the X100 in 2011: a fixed-lens camera with a retro body and physical dials for everything. In an era when every manufacturer was competing on specifications, it sold handling and colour. Its success is what the whole X series grew out of, and both of your bodies are descendants of it.

Decisions you can make in the field

Which body today

Situation	Take	Why
Just walking, may not shoot at all	X-E4	The camera that comes with you is the one that gets pictures
Low light, slow shutter speeds	X-T5	Stabilisation buys two or three stops
Might crop hard or print large	X-T5	40 MP gives you the room
Want Nostalgic Neg.	X-T5	The X-E4 does not have it
Would rather not be noticed	X-E4	Smaller body, quieter shutter

Four settings before you start

1. IMAGE QUALITY SETTING → IMAGE QUALITY → FINE + RAW
2. IMAGE QUALITY SETTING → FILM SIMULATION → choose one deliberately; do not leave the house on the default
3. IMAGE QUALITY SETTING → WHITE BALANCE → pin it when the light source is fixed
4. Check you have card space — RAW + JPEG runs four to five times the JPEG-only rate

Two checks after the frame

• Are the highlights intact? Contrastier simulations need more headroom.
• Is the colour what you wanted? If not, press Q and reconvert now rather than at home. Doing it on the spot is also the fastest way to learn what each simulation does to a scene you are still standing in front of.

Film simulation does not lock you in

One thing has to be clear: film simulations affect the JPEG only, never the RAW.

A RAW file holds the sensor’s own values; the simulation rides along as a tag. Which means:

• Experiment freely in the field. The RAW is still there.
• A bad JPEG can be reconverted in-camera (playback menu → RAW CONVERSION) with no computer involved.
• Or convert on a computer with X RAW STUDIO, which hands the processing back to the camera over USB, so the result matches the in-camera one exactly.

Recommended setting: RAW + FINE JPEG together. You get the pleasure of shooting finished pictures and a way back.

The next four chapters take the imaging parameters apart one at a time. If you would rather play while you read, the film simulation lab is right here.