Fuji/Canon/etc colors?

[ZoranC]

During browsing of Fuji X thread I observerd this look / rendering that seems to be common to many of samples that I have, not being able to pin down exact root behind it, started labeling in my brain as "Fuji colors".

After I have finished browsing my knee jerk response was "I want to buy Fuji X" but then I started thinking would that make any sense because I have started wondering how much of such looks is due to "look by certain hardware" and how much is due to manufacturer's choices of tone mapping, curves, etc during RAW development, similar to choices "film emulation" software does.

In turn I started wondering, if it's the later, can such look be recreated/reproduced during RAW development of files from any camera, whether A7R or some else, and if yes how to do it.

What are your thoughts / experiences with this?

 

[deejjjaaaa]

one of many approaches that can be used = sail2ithaki - How to make a "correction" profile

 

[ZoranC]

one of many approaches that can be used = sail2ithaki - How to make a "correction" profile

Thank you, but (please correct me if I am wrong) you are pointing to article on how to color calibrate your output, not how to get output from say A7R to match output from Fuji XT1. Two completely different things.

 

[ZoranC]

I am one of those shooters who prefers using a one-fits-all tool, in my case LR. Basically, I'm lazy (or, put kindly, interested in efficiency). I have certain camera/lens combos that produce the files which, coupled with my inventory of presets, deliver what I want by way of color, "look," etc. What I don't do is try to get different "looks" from the same camera/lens by varying workflow, developers, plug-ins, what-have-you. I probably should but, my time being limited, I prefer to shoot more and process less, I guess.

Thank you but topic is not "do you personally care" but "is my assessment correct and if yes how to go after it".

 

[Oren Grad]

...I have started wondering how much of such looks is due to "look by certain hardware" and how much is due to manufacturer's choices of tone mapping, curves, etc during RAW development, similar to choices "film emulation" software does.

In turn I started wondering, if it's the later, can such look be recreated/reproduced during RAW development of files from any camera, whether A7R or some else...

No, you can't necessarily map the color rendered by one camera to that rendered by another. The spectral transmission properties of the CFA determine which colors the camera will be able to differentiate, and these vary from one camera to another. If one camera cannot differentiate a given pair of colors while another can, no amount of correction applied to the files from the first camera will enable you to make them match the files from the second.

How much this theoretical principle will affect you in practice will depend on the colors of the subjects you photograph and of the lighting under which you photograph them, on how far apart the two cameras' CFA's are, on your own perception of subtle color distinctions, and on which distinctions you care about.

As the discussion of Sensitivity Metamerism Index on the DxOMark website puts it:

Digital processing permits changing color rendering at will, but whether the camera can or cannot exactly and accurately reproduce the scene colors is intrinsic to the sensor response and independent of the raw converter.

Color sensitivity - DxOMark

 

[Annna T]

No, you can't necessarily map the color rendered by one camera to that rendered by another. The spectral transmission properties of the CFA determine which colors the camera will be able to differentiate, and these vary from one camera to another. If one camera cannot differentiate a given pair of colors while another can, no amount of correction applied to the files from the first camera will enable you to make them match the files from the second.

How much this theoretical principle will affect you in practice will depend on the colors of the subjects you photograph and of the lighting under which you photograph them, on how far apart the two cameras' CFA's are, on your own perception of subtle color distinctions, and on which distinctions you care about.

As the discussion of Sensitivity Metamerism Index on the DxOMark website puts it:

Digital processing permits changing color rendering at will, but whether the camera can or cannot exactly and accurately reproduce the scene colors is intrinsic to the sensor response and independent of the raw converter.

Color sensitivity - DxOMark

Well put..

Add to that the fact that the color pixels are arranged very differently on the Fuji sensor than on others, so it is probably more difficult to equal the rendering of Fuji cameras than that of say Nikon cameras with the A7r.

BTW : different developing software will render different results.

Personally, I have often found that Adobe is adding too much yellow and their skies are too much cyan. At first I thought it was due to the body not being calibrated correctly. So I got the XRite color checker passport... But this didn't change anything (in fact, I read later that Adobe was collaborating with XRite for colors.. So no surprise). It may just be a question of cultural taste, because I have often heard Americans complaining that there were magenta cast in their skies.
Things aren't writen in stone. Before the last version of ACR/LR, i was happier with the way LR was rendering the A7r colors, provided I avoided the adobe standard preset. With the last version (5.4) they seem to have added more yellow to the camera presets too. I prefer the colors of DXO or C1, but on the other side it makes my workflow more complicated and I like the tone sliders of LR, so I'm constantly returning to LR too, in spite of my dislike for their colors.

I don't like playing with the camera calibration : when you begin touching something it may get you some colors right in your landscape pictures, but may disturb other colors in other pictures, for example for skin tones etc..

 

[fotografz]

I somewhat face this issue every time I process a wedding.

I may have up to 5 different cameras and a myriad of lenses in the mix. A Sony A900, or A99, or A7R, or my Leica S2, or a M9 (CCDs), and my assistant's Fuji, Oly or their Canon or Nikon DSLR.

The objective then becomes to TRY and mitigate the differences so there is some sort of over-all flow without obvious bumps from one image to the next.

Frankly, there are too many variables to be exacting about it. How the cameras initially render color, the differences in optical performance, the vast spread of lighting situations, etc.

The best I've been able to do is strike a compromise in the service of some consistency. I'm fairly sure that I'd fail at that if I chose one camera/lens system and tried to doggedly match it with the others.

The only really consistent ones of the bunch are the CCD Leicas. The S2 and M9 images differ very little, and it is easy to process them to be indistinguishable … any differences being attributable to shear resolution and DOF rending between 35mm and MFD, but identical in color and hue.

- Marc

 

[Tim]

If I look through my printed albums of film there is a similarity to most of the images.

As I page through my digital prints is a cacophony of colours and tones all over the place. You can see the change in cameras as you page through.

BTW: I have just ordered an A7. Reason is, I very much like the DR and look of the colours and tones of the Sony.
Its not that I don't like the Fuji, (they are probably the best jpging cameras ever), its that I wanted the sensor size of the A7 and the rendering was also very acceptable.

 

[philip_pj]

Everything affects color along each part of the imaging chain, and you add other parameters in choice of color space and software, gamuts of cameras, etc. I have never found it easy to match between models, especially if there has been a large jump in some aspects - like DR. Older CCD cameras were also quite different to today's CMOS chips.

At the end of it alI, you need to have enough experience and control of color management and processing and tone handling generally to get them close to each other, as all factors affect the others. You end up with your personal 'house style' with some variations for images from different source files, small but usually fairly noticeable, to the practitioner anyway.

Quite a bit can be controlled by choice of gear. I shoot an RX1 and a7r, and have used an a99 so these share very similar colour engines and gamuts and sensor response, noise patterning, and work up much the same in ACR. Some lenses - to risk starting an argument - make life easier, Zeiss and Leica have strong signatures usually common in each range, so if you use a few ZE or R lenses they add more common ground. In fact this is one reason to stick with your tried and known setup - to help color consistency, esp if it's very important to your work. More productive and efficient too.

 

[deejjjaaaa]

Thank you, but (please correct me if I am wrong) you are pointing to article on how to color calibrate your output, not how to get output from say A7R to match output from Fuji XT1.

you can create a correction profile to use in PS to get the colors closer to what you receive from the target camera...

 

[deejjjaaaa]

Everything affects color along each part of the imaging chain, and you add other parameters in choice of color space and software, gamuts of cameras, etc. I have never found it easy to match between models, especially if there has been a large jump in some aspects - like DR. Older CCD cameras were also quite different to today's CMOS chips.

At the end of it alI, you need to have enough experience and control of color management and processing and tone handling generally to get them close to each other, as all factors affect the others. You end up with your personal 'house style' with some variations for images from different source files, small but usually fairly noticeable, to the practitioner anyway.

Quite a bit can be controlled by choice of gear. I shoot an RX1 and a7r, and have used an a99 so these share very similar colour engines and gamuts and sensor response, noise patterning, and work up much the same in ACR. Some lenses - to risk starting an argument - make life easier, Zeiss and Leica have strong signatures usually common in each range, so if you use a few ZE or R lenses they add more common ground. In fact this is one reason to stick with your tried and known setup - to help color consistency, esp if it's very important to your work. More productive and efficient too.

a lot of text - where is a practical advise though about how to do it ?

 

[deejjjaaaa]

Add to that the fact that the color pixels are arranged very differently on the Fuji sensor than on others, so it is probably more difficult to equal the rendering of Fuji cameras than that of say Nikon cameras with the A7r.

dear, you kind of forgot that Fuji makes a regular bayer camera with the same colors as x-trans...

 

[hiepphotog]

dear, you kind of forgot that Fuji makes a regular bayer camera with the same colors as x-trans...

Do you still see the same so-called "Fuji color" with the X-A1? Otherwise, it's a moot point.

 

[ZoranC]

No, you can't necessarily map the color rendered by one camera to that rendered by another. The spectral transmission properties of the CFA determine which colors the camera will be able to differentiate, and these vary from one camera to another. If one camera cannot differentiate a given pair of colors while another can, no amount of correction applied to the files from the first camera will enable you to make them match the files from the second.

If for a second we ignore differentiation ability question is "why not"? There is obviously a way to "color correct" both camera A and B to let's call it "reality camera". Then task of matching A to B would be converting A to "reality" and then applying "from B to reality" transformation in reverse.

 

[ZoranC]

you can create a correction profile to use in PS to get the colors closer to what you receive from the target camera...

How exactly, please?

 

[deejjjaaaa]

Do you still see the same so-called "Fuji color" with the X-A1? Otherwise, it's a moot point.

you do

 

[deejjjaaaa]

How exactly, please?

read the instructions and think

 

[Oren Grad]

There is obviously a way to "color correct" both camera A and B to let's call it "reality camera".

No, there isn't. To borrow a bit more from the DxOMark discussion of Sensitivity Metamerism Index (emphasis added):

The underlying physics is that a sensor can distinguish exactly the same colors as the average human eye, if and only if the spectral responses of the sensor can be obtained by a linear combination of the eye cone responses. These conditions are called Luther-Ives conditions, and in practice, these never occur. There are objects that a sensor sees as having certain colors, while the eye sees the same objects differently, and the reverse is also true.

Similarly, you cannot be sure of being able to correct two different cameras to match each other, unless one camera's RGB color response at the level of the CFA can be expressed as a linear combination of the component RGB responses of the other.

That doesn't mean you can't ever get a match that's "good enough"; depending on what you're looking for, it might well be possible. But strictly speaking, in general, you cannot count on being able to achieve an exact match.

 

[philip_pj]

On Fuji colors and always at the risk of giving offence, I find them far less appealing than what Sony is giving us these days. Quite poor separation of tones with each hue area - so for example greens all tend to look the same across the range of luminance, a form of posterisation. Subjects all seem to have a 'look at me' quality to them, as a result!

Many spend years learning PS to pick up ways of separating color tones in Fuji's excellent E6 emulsions using PS's many color controls, so it's ironic that the company is gathering plaudits for its very similarly afflicted digital output.

Sony's world is much more distinguished and photorealistic in color than the garish yellow/green hybrid tones that populate Fuji's world view so fully. Their color is quite unsophisticated, and for nature shots, often cartoonish and Disney-like. There is more subtlety in a Ridley Scott sci-fi movie!

It is also very popular so I hasten to add that color preference is very personal, as we all see differently, and there are good reasons to believe this applies especially to color perception. I would never post this on the Fuji forum of course, it would start a bun fight.

On Fuji generally, why so many B&Ws if the color is so popular for users? They are running at maybe 60-65% in the hundreds of images I checked just now. OK, this is all for Sony users only.

It's somewhat easier in PS to mimic a primitive color response - so going from Sony to Fuji is not as hard as trying to recreate subtle lifelike tones from the narrow gamut of Fuji's output. The best control is Selective Color (selcol) which lets you mix final color after you see how PS renders overall tones in a PSD or TIFF file - and by the look of so many images in the X thread, you need to go heavy on the Hue/Saturation (huesat) control and narrow range contrast (can do with luminance masks). Some work in LAB, others don't like what it does to the files.

It's one case where working in sRGB is probably better, as you want to *reduce* color gamut by clipping, not expand it, as other makers are trying to do - they want to show more colors, not fewer! A monitor with good color response, calibrated well, is a big help too.

 

[Guy Mancuso]

For me I'll buy cameras for the color they provide and what looks good to my eye. One reason is post processing for me is not where I'm making my money. pros make money shooting in general spending hours in post is something I try to avoid on the whole. Sure tweaking out a red bias or yellow bias coming off the sensor you want to build a preset or calibrate that stuff out but if I want a Fuji look which is more yellowish or greenish you maybe able to get close but never accurate to it . I'm not a big fan if Fuji color but that's me. To me the best cam to look more CCD sensor is Sony. I'm coming from MF and it seems the sony comes fairly close to MF CCD. MF has better tonal range though. Guess my point here is I'm trying not to work my butt off in post. Just buy what looks good to your eye and software that matches your vision. Software matters a lot