Fuji/Canon: rendering of natural colours Part 1 Blue & Lilac

The weather has improved recently, the sun is shining and the flower season is upon us again (at least here in Germany). A few weeks ago when photographing blue bells I noticed a discrepancy between my Fuji and Canon images but so far did not pursue this. Today, I decided to take a slightly more thorough approach and photograph a selection of natural colours using both a grey card to set the white balance with and a set of iPhone images for comparison's sake and because I have found the iPhone, whatever else its many failings, usually gets the colours right and it allows me to easily check the colours against the real thing. Obviously this is no scientific comparison (who has a spectrocolorimeter at home?). Nevertheless, I still hope it may be of interest as a real world example of how both cameras deal with natural colours (yes I know the lenses may also have an impact) and it might even generate an interesting discussion.

Since this approach requires 5 images per colour I have split them up into different parts (no idea how many parts are allowed).

Red is missing! My one and only deep red rose is out of reach, but there will be others soon.

Cross-posted on Canon R.

METHODOLOGY

—All flowers photographed one after another on a sunny morning so that the light conditions were very similar indeed

—Raw only (except for iPhone)

—A grey card was used to set the white balance for the images taken with both cameras.

—No further processing in LR (apart from importing the images) except for setting the white point using the grey card and the occasional exposure, shadow and highlight adjustment

-Colour names somewhat arbitrary

CONCLUSIONS

1. There isn't much of a difference between the Fuji and the Canon images in terms of colour

2. The biggest problems both cameras have are with blue, purple and pink. Pink can be largely corrected using the grey card, blue and purple are best left alone because using a grey card shifts the colour towards purple in the case of blue and lilac in the case of purple.

3. If images are sometimes too cool out of camera, they tend to be too warm when corrected for white balance

4. White balance correction does not always make the colours more realistic, especially in blue/purple area

BLUE


iPhone image corresponds (by and large) with reality


Canon no white balance correction: slight purple shift


Canon with white balance correction: distinct purple shift


Fuji no white balance correction: distinct purple shift (same plant, different flower)


Fuji with white balance correction: not that different from out of camera image



LILAC

Both cameras were very similar and both were too cool (violet) out of camera.


iPhone: realistic


Canon no white balance correction: too cool (violet)


Canon with white balance correction: more realistic


Fuji no white balance correction: too cool (violet)


Fuji with white balance correction

Emile15 said:

METHODOLOGY

—All flowers photographed one after another on a sunny morning so that the light conditions were very similar indeed

—Raw only (except for iPhone)

—A grey card was used to set the white balance for the images taken with both cameras.

—No further processing in LR (apart from importing the images) except for setting the white point using the grey card and the occasional exposure, shadow and highlight adjustment

Click to expand...

If you are interested in colour, you need to create your own .dcp profiles for use in ACR/Lightroom.

Create a profile for each camera and lens combination (different lenses can produce different colour shifts that Auto WB doesn't always equalise).

Relying on the Adobe profile for each camera is 50% of what you are testing here, not only the camera's sensor.

I suggest you invest in a X-Rite ColorChecker Passport - and I also use Adobe's 'DNG Profile Editor' to actually generate the .dcp - as it does a good job there. Making a .dcp profile doesn't take long and then you are free from the skewed nature of Adobe's built-in profiles.

Thanks. Useful advice of which I am aware. My interest is in this exposition is simply to present what one ends up with when using either camera with or without white balance in LR. I guess it serves primarily to raise my own awareness of what influence colour renderings. The next step might indeed be the approach you suggest.

Emile,

There's some things that you can do to get a better test.

First, are you using a UV filter? UV was a real problem with film rendering natural colors of flowers. I've read conflicting information on whether the is a UV filter in front of the X-Trans sensor. With film UV would expose the blue layer.

I suggest you get a WhiBal card, which is a calibrated target. That's what I use. I take a few seconds to set a custom white balance. It really makes a difference when shooting in shade with the subject illuminated with skylight. AWB doesn't come close.

Do a test with the WhiBal custom white balance and look at JPEGs rendered by the film simulations. Standard should be the most neutral. The others are skewed on purpose.

As someone mentioned, you can get the X-Rite Color Passport if you want true neutrality. The software analyzes the target card and creates a custom profile for the camera and lighting conditions. I'd still start with a custom white balance. I use the WhiBal and X-Rite together, but normally I just use the WhiBal because I want the look of the film simulations.




Fuji X-H1, Pro Neg, Cloudy WB




Fuji X-H1, Pro Neg, Auto WB (R0, B0)




Fuji X-H1, Pro Neg, Custom White Balance




Fuji X-H1, Pro Neg, Custom White Balance, WhiBal target for CWB

The illuminant has changed for the rhododendrons hasn't it? It is mainly skylight not sun judging by the shadows,

Yes, that makes that part of the test not useful. An improvement of test technique, plus more data would make these comparisons more helpful.

Hi Doug, I did use a grey card to set the white balance for the shots. In fact in some of the later series you can see the card.

No, the rhodo images on the Canon and the Fuji were taken within a minute or two of each other; only the iPhone pictures were taken slightly later.

" —Raw only (except for iPhone) "

I see a fault here, RAW has no color. Color is the result of rendering (by the software you use).

So you are evaluating neither Fuji nor Canon. For that you should use SOOC images or canon/fuji proprietary software. In the case of canon that will be DPP4 (don't know what software fuji offers).

Otherwise you are comparing Apple colors to Adobe colors.