Is Old Fuji Color Better? Or Even Different?

Yawlen Z wrote:

At the top end of his January 2020 blog Mr. Roesch of fujixweekly does a fairly good job of dissecting the differences between X-Trans gens 1/2 and 3/4. I had to read some of that 5x over to grasp it right, in particular the section on tonecurve stepping differences that he found and the subsequent outcomes that he computed. It was worth the effort though as it did help me on. And at times I ended up cross-editing his advice for gen. II sensor WB shifts towards those that he lists for gen. III/IV

https://fujixweekly.com/2020/01/

From the article

"When I looked at the pictures that I captured with the different film simulations on my X-T1, it seemed like the results were different than with the same film simulations on my X-T30. Do the film simulations look different on different sensors? I did some tests to determine what is the same and what is not. I think there are some subtle changes, and you won’t get precisely the same results from X-Trans II as you will from X-Trans III and IV. It’s close, but not exact."

In my experience with every new X-Trans generation the older film sims get...

a) warmer

b) more saturated

c) more contrast-y

The newer film sims then tend to be cooler, less saturated and less contrast-y... or at least with a different highlight/shadow tone curve (I think the Color Chrome film sims - I forget the name) tend to have that "crushed but lifted blacks" look that's popular with some folks.

Interesting that the author points out that the white balance tends to shift between cameras. Depending on my purpose, I set white balance in camera to a grey card. An Expodisc if I want to do really critical work or a Lastolite collapsible grey card (which I use to set exposure) if I want to be less fiddly - plus I like the colors I get with the Lastolite (especially my old beat up one that's got a few grass stains on it - I attribute it to being part of my "look").

But I did find that white balance shifts when I go from the camera film sim to Adobe Color to a ColorChecker profile, so I'm not entirely surprised that he found something similar. An overall warming leads to a shift of where the center is (so to speak).

Re the 2nd article about Monochrome cameras:

A monochrome converted camera is necessarily more sensitive to light (the color filter arrays necessarily mean fewer photons reach the underlying photon detector). This means a higher base ISO, and better high ISO performance. Both because of the inherent increase in sensitivity to light and the lack of interpolation between CFA pixels.

Since Fuji cameras (and most APS-C and M43 cameras) tend to have a higher base ISO of 200 vs 100 for most full frame cameras, this means a Fuji camera that simply has the CFA removed will have a base ISO around 400, which would make them (IMO) worse certain kinds of photography.

Though I find that the X-Trans sensor has a different characteristic than a Bayer camera to begin with - the 2x2 green pixels do a great job at resolving detail. Though the default de-noising algorithms for X-Trans tend to have a surface blur effect - reduced microcontrast and heavy edge detection. I document it here for anyone who wants to read up on it.

For monochrome, I do like X-Trans but prefer something like RawTherapee that doesn't apply de-noising by default so I start with a grittier image and then can apply smoothing in post to my desire.