I’ve been busy the last few days trying to see what is similar and what’s different about the JPEG output from my new Fujifilm X-T5 camera. How does X-Trans V compare with X-Trans IV? Are my X-Trans IV Film Simulation Recipes really compatible with the new Fujifilm cameras? Technically they are because the options are the same (except that X-Trans V has the new Nostalgic Neg. film simulation), but do they render the same, or at least similar enough?
I was surprised by something that I discovered. While many of the colors are extremely close in rendering, blues are not. Take a look at the comparison below. Both images were captured with identical settings and even the same lens—one with an X-Trans IV camera, and the other on X-Trans V—but the blue sky is not the same. If you study close enough you might notice some other extremely subtle differences, but the rendering of blue is a clearly not the same between the two sensor generations.
I looked very closely at all of the different film simulations, and I noticed that this difference in blue is film simulation dependent. Not all film sims render blue differently, and some vary more than others. Here are my discoveries:
Eterna Bleach Bypass on X-Trans V renders blue a little darker than on X-Trans IV with Color Chrome FX Blue set to Weak.
Classic Negative on X-Trans V renders blue identically to X-Trans IV with Color Chrome FX Blue set to Weak.
Classic Chrome and Eterna on X-Trans V renders blue just barely lighter than on X-Trans IV with Color Chrome FX Blue set to Weak.
Velvia, PRO Neg. Hi, and PRO Neg. Std on X-Trans V renders blue halfway in-between Color Chrome FX Blue set to Weak and Off on X-Trans IV.
Provia and Astia on X-Trans V renders blue identically to X-Trans IV with Color Chrome FX Blue set to Off.
In other words, with the exception of Provia and Astia, blue is different on X-Trans V than X-Trans IV. With Classic Negative, if an X-Trans IV Film Simulation Recipe calls for Color Chrome FX Blue set to Weak, if you set it to Off instead, it will render the same; if it call for Color Chrome FX Blue set to Strong, if you set it to Weak instead, it will render the same. You can also do that with Eterna Bleach Bypass, Eterna, and Classic Chrome, and it will be pretty close, but it won’t be identical. With Velvia, PRO Neg. Hi, and PRO Neg. Std, you can go either way (adjusting Color Chrome FX Blue or not), and it really doesn’t matter because it will be wrong by about the same amount, either too light or too dark.
Does it matter? it’s not a huge difference, but it is a difference. You might prefer the rendering of X-Trans IV or you might prefer the rendering of X-Trans V. I think, personally, I’m leaning towards preferring the X-Trans V rendering—many of my recipes use Color Chrome FX Blue set to Weak or sometimes Strong, and with X-Trans V that’s already built-in now on some of the film simulations. But it also means that many X-Trans IV recipes will render differently, and, while technically compatible, aren’t truly compatible with X-Trans V cameras.
Why did Fujifilm do this? I don’t think this is anything new. For example, Fujifilm has tweaked Velvia film, because the original Velvia emulsion was a “mistake” (although many photographers didn’t think so—if it was an accident, it was a very happy one). I remember reading once that Fujifilm, with each sensor generation, reevaluates each film simulation to see if it’s possible to get them closer to the intended aesthetic. Fujifilm likely decided that the blue of previous generations wasn’t quite “right” for many of the film simulations and so they fixed it. You might not thought that it was “broke” so you’re wondering why they felt the need to “fix” it; however, the folks at Fujifilm must have thought something wasn’t quite right, so they adjusted it.
What else is different? I’m still looking closely, so I don’t have all the answers yet. I think the rendering of cool colors is slightly different, but blue is by far the biggest change—even with a close side-by-side comparison it’s difficult to spot the differences of the greens and purples, but blue is obvious. Shadows and luminance in general seem to be just a hair dissimilar, but it’s close. Warm colors seem to be pretty much identical; if they’re different it’s tough to spot—maybe yellow is the most dissimilar of the warm colors, but it is still very similar, and I think it might have more to do with general luminance than a color difference. Really, blue rendering is the only significant difference I’ve found so far between X-Trans IV and X-Trans V.
With the Fuji X Weekly App, I don’t have X-Trans IV recipes currently listed as compatible with X-Trans V cameras. There are some—those that use Provia or Astia—that could be listed as compatible because they’ll render the same. For others, a Color Chrome FX Blue adjustment will make them compatible. For others, they won’t ever look the same but will still look pretty similar. I’m still deciding how I’ll handle this. The easy route would be to just say they’re all compatible—that they’re close enough—but I don’t think that’s the right path. I ask for patience as I wade through these waters—the rendering of that blue water on X-Trans V—and how to best present it to you in the form of Film Simulation Recipes.
Hi Ritchie! Thank you for this article and your work around!
Can’t wait to play/test different recipes in same way 😉
BR Thomas
Awesome! Soon, right?
Unless everyone is standing in a gallery looking at the same physical print illuminated by well placed lighting, no two people will perceive the color cast the same. This is especially true for a screen photo (JPEG) since most everyone is looking at a differently calibrated screen, what with all the varieties of phones, tablets and computer monitors.
It’s like the old saying about watches. If you only have one, it is always “correct”. If you have two, with a different time displayed on each, well… I doubt the variation in blue would have even been seen if you didn’t have two “watches”. However with the concept of time, there is an absolute truth. I use the Atomic clock to set my many watches. Color is a judgment.
Hope you are enjoying your X-T5.
Your argument would be correct only if both images were displayed in different screens. Here both are displayed at the same time in the same device, so any difference in the monitor will translate equally to both, translating in the same degree any underlying difference. Digital media, besides, has the screen as the source of illumination, is less, not more, dependent of external variables as prints are, whose source of illumination is external.
For what is worth in both of my calibrated screens, one computer monitor and the other a smartphone from different brands, the shift in the blue is noticeable in the same way. Also, as you yourself pointed out, the more brands one use the more differences one notices in interpretation of color, regardless of how much variation our subjectivity introduces. In this time when smartphones give too much importance to the camera module people in general is more aware that some companies give them colors or renderings that are pleasant, even if they don’t consciously are looking for it. The same happens with camera brands where some offer more neutral colors, and other introduce measurable shifts, measurable and absolute truth because although admittedly subjectivity and proximity between colors alter their perception colors have a certain wavelength frequency. And time is relative, satellites have to use relativist physics to compensate for the differences in the surface, otherwise, if technology would work considering the time as an absolute measurement, internet wouldn’t be possible.
Hello, Francis. I’m not saying that you can’t see the difference, only that it doesn’t matter. For one to be “right”, then one has to be “wrong”, and one can’t suddenly be wrong because of a new model hitting the streets. Blue is blue, the shade may be different, but one is not better or worse.
Without the comparison, you’d never know one is different, thus my two watches analogy.
I doubt any photo would suffer regardless of which camera you use.
Hello, Albert. Of course there is no right or wrong choice in preferences, and I don’t recall in the article or in my reply any indication about it, in fact in the article Ritchie mentions that it is a personal choice. Personally I prefer the blues in the sky to shift not to the magenta, even if it is not accurate is, what I find pleasing, not better. Because color is very important to me; and maybe I have edited and shot too many photographs, maybe influences that I had drawing and painting as a hobby, but actually: yes, I am aware how colors are different even without comparing because I am aware of the choices by the engineers about the way they shift some colors to get a palette, likewise with raw converters, or even smartphones.
I wouldn’t say a photo would suffer according to a camera, just, as quoting you “not better or worse” is a preference. If some equipment helps me to achieve my preferences is a good tool and I can enjoy more taking photographs. which of course is different for you and that’s fine, but is okay as we both have different experiences and backgrounds.
Where I think it matters is that I select a certain setting (such as Color Chrome FX Blue) for a reason. If someone follows an X-Trans IV recipe (which calls for CCEB Weak) on their X-Trans V camera, it might (depending on the film simulation) produce a look more like Color Chrome Blue Strong. Maybe a lot of people wouldn’t notice the difference. Maybe some people even prefer the divergent rendering. And in that way perhaps it doesn’t matter. Maybe a lot of the subtle settings don’t matter, such as a shift of -4 Blue vs -5 Blue. I will say that I noticed something was not quite “right” before I did this comparison (it’s actually what prompted me to do it), but maybe my eyes are a bit more “trained” than most? I don’t know if it matters to most people—it probably doesn’t—but it does matter to me, so I shared my findings.
Interesting…. 🧐
We have five watches now… the X-Trans I, II, III, IV, and now V. I’m not sure which one is most “correct” (because color is more subjective than time). I do think it is interesting that Fujifilm tweaked the blues on X-Trans V.
Thanks a lot for your hard work and detailed comparison, it’s very welcomed and I’m waiting for your new recipes with XT5 sensor. Thanks again!!
I appreciate your kindness and encouragement!
I agree whit Albert Smith that “Color is a judgment”. So I think that to study close enough we need at least pictures with full resolution.
Resolution is not the same as color, so full resolution won’t give you any relevant data or any advantage to notice differences in the plain color of the blue sky as a blue pixel is not different from three or one million of blue pixels; and even so you won’t even will be able to watch it as monitors at most, go to 4 megapixels in resolution.
Your argument is a bit strange as well because in both my color calibrated LG ultrawide and my Samsung S20 FE in natural mode (color calibrated for SRGB) the difference indicated is clear and visible.
I have four monitors (two Dell, NEC, and Philips tv) and all of them show a different picture. What is interesting, calibrated Dell shows the minimum difference on blue. So again “Color is a judgment”, but on the way, it would be interesting to study all differences between XT4 and XT5. That was my main ‘argument’.
So far, from what I’ve been able to tell, the rendering of blue (on most film simulations) is the biggest difference, and the only one of significance, if it even is significant. I’ll keep studying, though, and report what I find.
Due my professional background, not in photography, color is very important so I only purchase color calibrated equipment, IPS screens, I take it in account when setting the shift and the white balance for the Fujifilm recipes, is probably that your Dell doesn’t cover 100% or close to it, and if so maybe it is a TN panel which are more for videogames but not that much for color accuracy. To have a calibrated screen today is easier than just five years ago, so easy that even if you don’t get a new modern monitor you will have still a cellphone from Apple, Samsung, or any other big company with screens professionally calibrated by default.
I get requests from time-to-time to display full-resolution images, which I completely understand, because you just can’t get into the details with a low resolution copy, and you’re also dealing with compression artifacts. But there are several reasons why I don’t: data costs money, it makes the website run slower (which reduces the experience and also makes it rank lower on Google), and it makes the pictures more prone to theft (not by the person asking for full-resolution, but by those looking for “free” pictures when they should be paying for them… it happens much more often than it should). Unfortunately, it’s just not “worth it” for me to add full-resolution pictures to this website.
There’s this: https://www.dpreview.com/reviews/image-comparison
I’m not sure which film sim they use for the JPEG, but you can tell a minor difference in the blues between the X-H2 and X-T4.
Nice catch. Clear as day difference on my iPhone 14. I much prefer the truer blue of the XT-5 in that photo — definitely can see a purple tint with the other. I’d say that’s an improvement! Hope to hear and see more about the differences you notice.
I think the red/purple tint is built-in to that shade of blue on that film simulation (Classic Negative, if you hadn’t guessed already). When adding Color Chrome FX Blue Weak to the X-Trans IV picture, it becomes identical to the X-Trans V picture. If you look at the X-Trans V picture, you can still see some of that tint in the bottom-right. I’ll definitely report whatever else I find.
Maybe it’s time to get with Dave Etchells from Imaging Resource and dig into the simulations on XTrans-V sensors as was done with XTrans-IV:
https://www.imaging-resource.com/news/2020/08/18/fujifilm-film-simulations-definitive-guide
-Eric
He might have to update it. I’m sure that was a heck-of-a-lot of work to publish in the first place, though.
time for a poll to post. What IV recipes do you want most on V? My first vote is Positive Film. 🙂
Awesome! I like that recipe, too. 😀
yup, as noted the blues are the biggest change with the V sensor, so thought Positive Film is a good one here, along with Pacific Blue 🙂
I’ve definitely have a lot to look through. I thought I found a weird WB anomaly, but it only appeared (so far) in one side-by-side comparison, and I haven’t been able to replicate it since. I used two different lenses in that particular test, so I’m thinking it is the lens. Otherwise, it seems the blue rendering is the only significant difference thus far.
Hi Ritchie. Do you have any idea about differences in rendering between the new stacked x-trans V vs the 40 MP non-stacked x-trans V? Thank you for your awesome work and everything you do for the Fuji community!
I don’t. I’m assuming that Fujifilm worked to make it as identical as possible, but I have not come across any direct comparisons to verify it.
pal2tech just made a video comparing the X-T5 and X-T4 on SOOC jpegs and RAW files in Capture One as well
very interesting
https://youtu.be/KHMUlj7Ui5g
That’s very interesting. Thanks for sharing!
Thank you for the detailed analysis! Looking forward to learning what the new sensor has to offer.
What film simulation did you use for the sky comparison picture in this article?
Oh, geez. I believe it was Fujicolor 100 Gold, but with Grain turned Off, and then I made versions with Color Chrome Effect and Color Chrome FX Blue set to Off, Weak, and Strong, so that I could spot any differences.
Apologies if my question is too simple, but will the simulation above work on the X Pro3 and the X100V?
The Nostalgic Neg. film simulation is not on the X-Pro3 or X100V… I believe that is what you are asking. Sorry.