SimuLook · Notes from the camera
Read a recipe out of a photo you already shot
You liked how a frame came out eight months ago and you have no idea what was in the camera at the time. The usual answer is that the information is gone.
It is not. Your Fujifilm wrote it into the file.
Every JPEG carries its own recipe
When the camera makes a JPEG it records the settings it used in the file’s metadata — film simulation, dynamic range, both tone controls, colour, sharpness, noise reduction, clarity, grain, both Color Chrome controls, the white balance shift. The recipe you shot with is sitting inside the photograph.
That means you can recover it with no cable and no camera present. Point the app at the photo and it reads the settings out.
And unlike reading over USB, this is not limited to one camera model. The metadata format is the same across bodies, so it works on any supported original Fujifilm JPEG. We check every read against frames from more than one body.
What it cannot do
The catch is the word original.
The metadata lives in the file the camera wrote. Screenshots do not have it. Neither do photos exported by another editor, sent through a messaging app, or re-saved after a crop. Those all produce a new file, and the new file carries the editor’s metadata rather than the camera’s.
If you have the original on a card or in a library that keeps originals, you are fine. If all you have is what came back from Instagram, the settings are gone and no app can recover them — the honest answer, rather than a plausible guess.
When we cannot read a file we say specifically that, instead of showing a half-filled screen.
Two questions, kept apart
Once the settings are out, the app answers a second question: which of your own recipes is this?
Those two things carry very different certainty and the app keeps them visually separate. What settings shot this frame is a fact read out of the file. Which recipe it matches is a judgement, with a score and possibly a near miss.
Three rules make that judgement worth reading:
Film simulation is a gate, not a weight. If the photo is ACROS and the recipe is Classic Chrome, there is no match. Not 85%. They are different looks and averaging them into a percentage would be nonsense dressed as arithmetic.
A near miss names the difference. “87% match” tells you nothing you can act on. “Photo +2, recipe +1 on shadow” tells you exactly what to change. So the app lists the settings that differ.
A field only one side knows about is skipped, not scored. If your recipe does not specify grain and the photo does, counting that as a mismatch would punish the honest recipe, and counting it as a match would flatter every recipe equally. It is left out, and the app tells you how many fields it actually compared — so a thin match reads as thin.
What it is good for
The obvious use is recovering a look you lost. The better one is checking your own work: shoot a frame with a recipe, read it back out of the JPEG, and confirm the camera held what you thought it held.
That is also how we verify the app itself. The seven slots on our test camera were read two entirely different ways — over the cable, and out of photographs of each slot — and the two agreed on every field. Two encodings of the same seven recipes, neither derived from the other.
Reading a recipe from a photo is free in SimuLook, on every camera the parser supports. It needs nothing but the file.
Written while building SimuLook, an app for reading and writing Fujifilm film simulation recipes. See which cameras are supported · Read the other notes