In a recent discussion thread on the Fuji Rumors site, I noted a comment that sounds right at first but actually needs to be challenged. Indeed, a lot of people believe some variation of that thought, so perhaps it’s time to address it generally. Here’s the point I’m referring to (somewhat edited by me for clarity):
"Using the same sensor in a series of different bodies gives you the same image quality outcome and thus you can use the same workflow independent of body. Let's say you shoot in the studio with your Fujifilm X-T4 and on the streets with the Fujifilm X100V: the image quality will be consistent in quality, color and overall look. Your workflow and outcome will be consistently the same. You want consistency.”
I agree and disagree with this comment. The problem here is conflation. I see that a lot when people try to make a point these days.
Let’s start with my agreement: using one maker’s products rather than mixing makers should give you a consistent workflow and outcome. Every now and then the maker will drop a technology that changes that until you standardize on products only with the new technology (e.g. what happened when Nikon debuted Picture Controls and EXPEED with the D3/D300 and dropped Color Modes).
But what’s the source of this consistency?
No, it’s not the image sensor. It’s the image signal processor. DIGIC, X-Processor, EXPEED, BIONZ, PRIME, etc. These processors take raw data (red filtered, blue filtered, and green filtered digital numbers) and turn it into finished RGB pixel data (typically JPEGs, but sometimes TIFF, or these days HEIC).
As most Nikon users can tell you, the only NEF (raw) converter that produces JPEGs that are consistent with the in-camera JPEG production is Nikon NX Studio (also earlier versions, such as Nikon Capture NX-D). That’s because the “EXPEED engine” is built into NX Studio, just as it is in the cameras.
The minute you bring another company into the mix—Adobe comes to mind—all bets are off. You might get a raw conversion that’s similar to what the camera will produce for JPEGs, but it is absolutely not the same. Adobe has spent some time recently trying to improve their profiles and settings recognition so that they better match the manufacturers’ results, and even did so in consultation with some of those camera makers. But no, “similar" is not the same as “the same” (pardon the recursion ;~).
Some raw converter providers do something different: they take all raw files, regardless of camera that produced them, and process them to provide the same final color and image attributes. CaptureOne does this. They have their own color model that they convert images to (by default), and spend time profiling cameras so that CaptureOne can provide consistent color across cameras.
Which, of course, has brought me to one of the biggest photography myths on the Internet: “Camera X produces the best color.”
Not a single camera on the market, nor any raw converter you can buy, produces exactly accurate color. I’m pretty sure I can test and find something that “isn’t accurate” to reality for anything you throw at me.
What every camera maker and raw converter does is try to create pleasing color. Canon and Fujifilm both tend to use double hue shifts. Fujifilm and Olympus tend to use higher saturation. Olympus tends to use higher contrast (hides the noise in the shadows). Nikon tends to de-emphasize contrast gains due to sharpening. And that’s just the beginning. There’s a lot going on in all those image processors.
So, no, Fujifilm isn’t getting its image consistency from the physical image sensor, it’s getting it from the X-Processor and what it does to the digital numbers the sensor produces. Indeed, anyone that went from the lower megapixel count Fujifilm’s to the higher ones could probably tell you that, as the JPEG images stayed perceptually the same despite a change in image sensor.
It’s important for all of us who are truly interested in nuanced results to understand where any differences and changes actually come from. If we mix that up with other attributes—megapixel count, for example—we start getting wrong thoughts and eventually wrong answers. We’ve seen that in discussions of noise, color, and much more.
