Newly purchased GFX100S has red tint

[rdeloe]

Aren’t the Fujis famous for their film color simulation? So, there should be a way to reproduce a “natural” color rendering.
Never used Fujis. But the consistency of color brought me from Leica to HB.

The film simulations are a JPEG thing. They are implemented in LR with Adobe's versions, and in Capture 1 with their versions. Some have said that the C1 "Fuji" film simulations are more faithful to what Fuji intended than the Adobe versions.

I have not noticed a difference. But I also don't use them as a rule. I made my own profiles with a ColorChecker Passport, but didn't find them very good. For the last year or so I've used profiles for the GFX 100S made by Color Fidelity. The creator of those takes a different approach. Rather than trying to simulate a film stock, he aims for a neutral starting point. That approach works for me.

I don't use Hasselblad but I spent some time exploring output from a CFV 100C to see what it is like. Hasselblad's "Natural Colour Solution" produces pleasing files, but it's not what I call a "neutral" solution in that it's very much how Hasselblad thinks the world should look. If that lines up with how the photographer thinks the world should look, then it's a good match.

 

[usm]

Hasselblad thinks the world should look. If that lines up with how the photographer thinks the world should look, then it's a good match.

That’s right and I totally agree.
What I like about Hasselblad is the constant color across the models. Leica (M) was always different, each version has a different look. Fuji seems to be similar.

And thanks for the explanation of the film simulation thing. I thought it is in the raw file.

 

[rdeloe]

So Basic Question:
Is there ANY settings in the GFX100S that would affect the “colors” of the RAW file created by the camera when taking a picture?

Back to your basic question, are you using a native or an adapted lens? I ask because this introduces another variable. I've noticed wildly different white balance when I use a native versus adapted lens.

Here's the GF 35-70mm at 35mm on the left, and the Schneider-Kreuznach APO-Digitar 35/5.6 L-88 on the right. I've balanced exposure, but that's it. Both are in Lightroom with the default profile "Adobe Color".The GF image has Temp 5050 and Tint +21. The SK lens is Temp 4850 and Tint +50. Notice the magenta tint in the fence boards.



In this next pair, the only change is I've adjusted the SK version to have the same Temp and Tint value as the GF version. They're now much closer. If I needed to match them perfectly, I could do it.



The light did not change, and white balance is on "Auto" in the camera, so this is entirely a white balance problem. I presume, but do not know, that Lightroom is paying attention to something that was written into the RAF made with the GF lens, and that whatever that is was not written into the RAF for the adapted lens (because the camera thinks it is shooting without a lens).

The interesting question now is, "Which one is 'correct'?" This takes us into the realm of visual perception. My monitor is calibrated, but it's an older monitor, and maybe my eyes and brain are not calibrated properly -- lots of evidence pointing in that direction.... Nonetheless, the versions that match my perceptions of this scene are the GF version with the Tint of the SK version, and the SK version with the Temp of the GF version. In other words, I think the GF version has too much green, and the SK version has too much blue. This is what they look like with Temp 5050 and Tint 50. There's more magenta in the fence boards than there should be, but in the overall picture, the range of greens matches what I see better.

 

[mristuccia]

I'm getting ready to quit for the evening but before I do I have one issue with C1.
I can't see that C1 is applying ANY of the ICC profiles to any of my images. There are 50-60 different ICC profiles already in C1 and not a single one affects the images. I even tried the "yellow affects" profile and the B/W profile but nothing is changing the image.

As far as the WB goes I set the WB in the camera to Outdoors then took a pic. I then did a WB on the grey patch and it then gave the image a pink tint. See the right side of the screen shot below.
View attachment 213269

If I didn't do a WB correction I would lean towards the original RAF file looking pretty good as-is but I just wish I can get the ICC profiles to do something. It's possible that the RAF file is so close to being perfect that I can't see the ICC changes but since none of the "boxed" profiles or my really old custom ones I made under different lighting are doing anything I am guessing that the new ICC isn't being implemented.

Anyway, I'm done for the evening and I won't be able to get back to any of this until late Sat night or Sunday.

Have a great weekend everyone!
Kevin

Honestly I don't get it.
I've made a stupid test with the macOs Digital Color Meeter directly on this forum page.

if I measure your grey area on the right part with the macOs color picker I measure a perfect neutral color (R = G = B):



Same if I measure on the wall (R = G = B):



On the contrary, the left image is greenish (G > R and B):



On my calibrated monitor the image on the right side looks pretty much neutral, as the color picker indicates.
The left one looks clearly greenish, as the color picker indicates as well.

Am I missing something here?
Are you using a calibrated monitor?

 

[Mexecutioner]

Left looks green on my calibrated computer monitor and even on my iPad. Maybe the canon file is the problem.

 

[KEVINS]

-The greenish tint is CORRECT for the actual painting despite what people think how "correct/neutral" the one on the right is perceived.
-The greenish tint is a bit strong but it is close enough that I can deal with it. I purchased Lumariver this morning to see what the ICC profile looks like compared to ColorChecker. It appears that Lumariver has a process for "Reproductions" whatever that is, so I will go through this to see what I get.

Back to your basic question, are you using a native or an adapted lens? I ask because this introduces another variable. I've noticed wildly different white balance when I use a native versus adapted lens.

I am using the Fujifilm GF 80mm f/1.7 R WR. I stayed with a native lens to hopefully eliminate any color/quality issues.


Onward:
I had BETTER LUCK after last night so I'm back at it this morning.

1. The RAF files do NOT show a pink tint.
2. Doing a DNG conversion using Adobe DNG Converter also looks fine.
3. Doing a WB correction in Post on the RAF file in C1 is causing a pink shift in the image.
4. Doing a WB correction in Post on the DNG file in ACR is also causing a pink shift in the image. I don't know why the color shift is happening, but it is. Normally I don't perform a WB correction in ACR but I tried it last night and it also adds the pink tint.
5. I set a custom WB in the camera and the RAF looks fine.
6. After setting WB in camera I created custom ICC profiles for C1 and for PS and the pink color does not show up when implementing these ICC's.

*Moving forward I will eliminate the WB correction in post but will use the custom WB in camera instead. This is giving me files that I can then create custom ICC profiles that I can play with in PS and C1.

************************
**********
****
*

At this point I think this particular issue is resolved. =)
Thank You All for your time and assistance!

I can now concentrate on trying to get an accurate ICC profile which I will compare betwen Lumariver and ColorChecker.

Side Note: I am having issues assigning custom ICC to the Pixelshift images so I will spend time this week trying multiple processes to see if I can get it to work. If I have questions I will post a new thread, so feel free to run for the hills if that thread pops up...

Have a Great Weekend everyone and again, THANK YOU VERY MUCH for your time!

Kevin

 

[mristuccia]

I still don't get it.
As you seem to be so keen about ICC profiling, I take for granted that you are profiling the whole chain (camera + computer monitor + printer), otherwise you won't get out of the loop so easily.

As for now, the important thing is that you've (self) solved the issue.

 

[KEVINS]

As for now, the important thing is that you've (self) solved the issue.

It wasn't just me. Steve's effort behind the scene on comparing the RAF and RAW files in his above post pointed me quicker to trying the in-body WB suggestion. His effort also produced a pink hue in the Fuji files when doing a WB in C1.

ks

 

[lookbook]

... thank you for your help Mr. @Steve Hendrix

 

[Pieter 12]

A bit off topic, but a lot of color is perception. This is fun and informative:
\https://www.xrite.com/hue-test
Pantone used to have an online test too, but it seems to have disappeared.

 

[Whisp3r]

A bit off topic, but a lot of color is perception. This is fun and informative:
\https://www.xrite.com/hue-test
Pantone used to have an online test too, but it seems to have disappeared.

Booyah! First time in my life that I'm happy to score zero on a test

 

[rdeloe]

Booyah! First time in my life that I'm happy to score zero on a test

View attachment 213370

Interesting. I also got a perfect (zero) score, which I find a bit surprising. I only started working in colour a few years ago, after decades of exclusively working in black and white. I find colour is a lot harder.

 

[lookbook]

... I would have been better with grey tones!!! : )​

 

[KEVINS]

I agree, most (all?) people on this forum work with perception so anybody can question what others are seeing and there is no right/wrong answer.

I am working with an actual painting that I am wanting to reproduce it's color so there is very little perception, only the amount of tolerance that I am willing to accept in the color reproduction, and I am willing to accept some amount of tolerance as long as I feel it's still represents the painting.

Finding people that actually reproduce colors is difficult and finding those that are willing to help is even harder. Years ago I started down this path and chose to duplicate someones hardware to minimize the growing pains and it's worked really good - until I bought a IQ180 which was ultimately deemed faulting by those of us swapping RAW files. There was so much going on behind the scenes that I was tossing $$$$$ at a software/hardware/component/calibration problem b/c we all figured the P1 was perfect and something upstream was causing the issues. Wrong. The P1 was the problem.

The P1 is gone and in it's place is this GFX100S based on another person that photographs million dollar paintings. I thought about the new 100MP Hassy and asked him if there was any regrets with his GFX100S and he said no and that I would be happy with it, not to mention I would love the pixel shift option. (Now I am having issues assigning custom ICC profiles to pixel shifted images..what a mess!).

So here we all are on a Memorial Day weekend in the US and I just spent the afternoon feeding downed tree branches into the chipper and now it's time to research the ICC issue wit the pixel shifted files..

Have a great weekend All!
ks

 

[Alan]

The only thing I can add here is to remind about metameric failure. Some (many?) paints / materials will look different under different light sources. Things like UV coated / uncoated flash tubes, low quality LEDs, etc will throw wrenches in color matching efforts. Good luck!

 

[Pieter 12]

I documented paintings for years with large format (8x10 and 4x5) transparency film, as well as with 35mm. There can be enough difference between how colors look with transmitted light (projectors, light tables, monitors) and reflected light—prints and paintings. Also, some pigments reflect differently depending on the spectrum of the light source. Add to that surface texture and gloss/reflections and the introduction of polarizing filters on the light sources and lens, and it can be quite difficult to accurately reproduce some artwork.

 

[Pieter 12]

Thinking some more about the photographing and reproduction of artwork, it is almost always a losing proposition if you are seeking total accuracy. First of all, anything posted on the internet is limited in gamut and will suffer, sometimes immensely. If the purpose is for reproduction in a book or catalog, the printing process itself is limited and colors can be affected by the paper stock, number of inks on press, varnishes and coatings. In the end for many pieces of art, nothing does justice to the real thing.

 

[KEVINS]

Thinking some more about the photographing and reproduction of artwork, it is almost always a losing proposition if you are seeking total accuracy. First of all, anything posted on the internet is limited in gamut and will suffer, sometimes immensely. If the purpose is for reproduction in a book or catalog, the printing process itself is limited and colors can be affected by the paper stock, number of inks on press, varnishes and coatings. In the end for many pieces of art, nothing does justice to the real thing.

Yep, I agree 100%! Which is why I mentioned my acceptance of some amount of tolerance within the images, and my 5DSr produces images within the tolerance. Are they exact for every possible situation? Nope, but perfectly acceptable for the images intent.

I'm one of the few artist in the world that spends a month or more sanding/polishing the canvas before painting. Among other things this helps remove glare coming off the weave/strands of the canvas which affects the overall color in a photo and in person. Removing all the tiny little light reflections off the strands produces colors that are more saturated in photos compared to looking at the painting in real life where we are viewing "watered down" colors from the microscopic light glare. Art reproduction is an art in itself and it's a struggle for those that go down this path...

I plan to take more pics tonight using what I learned then see if I can get the Fuji images to an acceptable image quality. I was fighting getting a custom ICC applied last night so I want to start over with new photos and a clear head.

ks

 

[Pieter 12]

Yep, I agree 100%! Which is why I mentioned my acceptance of some amount of tolerance within the images, and my 5DSr produces images within the tolerance. Are they exact for every possible situation? Nope, but perfectly acceptable for the images intent.

I'm one of the few artist in the world that spends a month or more sanding/polishing the canvas before painting. Among other things this helps remove glare coming off the weave/strands of the canvas which affects the overall color in a photo and in person. Removing all the tiny little light reflections off the strands produces colors that are more saturated in photos compared to looking at the painting in real life where we are viewing "watered down" colors from the microscopic light glare. Art reproduction is an art in itself and it's a struggle for those that go down this path...

I plan to take more pics tonight using what I learned then see if I can get the Fuji images to an acceptable image quality. I was fighting getting a custom ICC applied last night so I want to start over with new photos and a clear head.

ks

Completely off-topic, but have you ever tried painting on the flat side of masonite that has been gessoed and sanded multiple times, the last with 000 wet sandpaper? Different feel than canvas (no bounce) but a super-smooth surface to work on.

 

[KEVINS]

Completely off-topic, but have you ever tried painting on the flat side of masonite that has been gessoed and sanded multiple times, the last with 000 wet sandpaper? Different feel than canvas (no bounce) but a super-smooth surface to work on.

I have not. I've thought about it but I kinda like the bounce of canvas, not to mention it's lighter to move around when they get large and fully framed. My better pieces get a double frame which makes them look great but they can be a bear to move around and hang.

ks