R5 colour science and skin tones.

Feb 15, 2020
505
347
Here are a couple of quick test shots that demonstrate the most extreme difference I have been able to find between EOS R and R5 in terms of colour (magenta cast).

Left image is EOS R set to 'adobe color' (f1.2, 1/200 sec, ISO 100)
Right image is R5 set to 'adobe color' (f1.2 1/200 sec, ISO 100)

Lighting is Dedolight DLED7 set at 2800k from camera left and Litepanels Gemini 2x1 set at 10,000k from directly in front.

Another interesting finding is that the R5 files are looking quite a bit darker than the EOS R files at the same setting when viewing in Adobe Camera Raw as a DNG. When opening the R5 files in DPP they are much brighter. In the below example the lights are set to the same power and at the same distance, but in the EOS R image I actually decreased exposure by 0.3EV to make them a little more comparable. Found that a bit strange.

When using the lights set at 4000kelvin the difference between the two cameras still exists but it is a bit more subtle...
 

Attachments

  • magenta_test_03.jpg
    magenta_test_03.jpg
    352.8 KB · Views: 105
  • magenta_test_04.jpg
    magenta_test_04.jpg
    404.6 KB · Views: 94

Viggo

EOS R5
Dec 13, 2010
4,737
1,428
Did you just trust the Kelvin setting on your lights, or is this measured, then how? Or sampled from gray card etc?
 
Feb 15, 2020
505
347
Did you just trust the Kelvin setting on your lights, or is this measured, then how? Or sampled from gray card etc?
In those particular examples the Litepanels Gemini was set to 10,000 kelvin. I doubt it is very accurate at that setting and unfortunately I don't own a colour meter to get a bettet idea of what is going on. I wanted to push the colour temperature difference to the extreme to see how the different sensors would react to a more challenging situation.

When the Gemini is set to 4000k it is accurate to within +-50k and has a 0 DUV measurement (no green or magenta cast) so I trust it at that particular setting. I set the R5 to +5 green in camera when using the light at 4000k and the magenta cast was mostly gone. It also rendered my 5500k Broncolor studio flash quite neutral with the same +5 green setting. So I'm inclined to just leave the camera set to +5 green permanently.

I will be doing some more tests soon to see what is going on. I'm eagerly awaiting full camera raw support from Adobe to see if that makes a difference too..
 
  • Like
Reactions: Viggo
Feb 15, 2020
505
347
Here is a test shot with the front light set to 4000k (Gemini) and the side light set to 2800k (Dedolight). The camera was set at 4000k with a +5 green value

Second image is using natural window light with the same in camera +5 green value and set to 5000k
 

Attachments

  • magenta_test_05.jpg
    magenta_test_05.jpg
    337.5 KB · Views: 87
  • magenta_test_06.jpg
    magenta_test_06.jpg
    224.7 KB · Views: 128
Last edited:
  • Like
Reactions: Viggo

SecureGSM

2 x 5D IV
Feb 26, 2017
2,376
1,246
In those particular examples the Litepanels Gemini was set to 10,000 kelvin. I doubt it is very accurate at that setting and unfortunately I don't own a colour meter to get a bettet idea of what is going on. I wanted to push the colour temperature difference to the extreme to see how the different sensors would react to a more challenging situation.

When the Gemini is set to 4000k it is accurate to within +-50k and has a 0 DUV measurement (no green or magenta cast) so I trust it at that particular setting. I set the R5 to +5 green in camera when using the light at 4000k and the magenta cast was mostly gone. It also rendered my 5500k Broncolor studio flash quite neutral with the same +5 green setting. So I'm inclined to just leave the camera set to +5 green permanently.

I will be doing some more tests soon to see what is going on. I'm eagerly awaiting full camera raw support from Adobe to see if that makes a difference too..
Shoot a quick message to Canon Australia. Let them know. This should have been quite easy to correct via a firmware update. I think.
 
Feb 15, 2020
505
347
Shoot a quick message to Canon Australia. Let them know. This should have been quite easy to correct via a firmware update. I think.
I might contact them, thanks for the suggestion. I'm not very confident in it being addressed in a firmware update as it would be quite a major change to the camera. My own testing and all of the sample images I've seen so far all seem to have a noticeable magenta bias. It seems that for some reason Canon has reverted to it's older colour science... quite different to the 5Div and EOS R which I loved for colour!
 
  • Like
Reactions: SecureGSM

stevelee

FT-QL
CR Pro
Jul 6, 2017
1,979
735
Davidson, NC
Reds are defiantly an issue
Autocorrect sometimes makes you say things you didn't Nintendo. But in this case, it might be appropriate.

I take it that the moral of your story is that when photographing well-lit expanses of color, particularly red, underexposing a bit can help preserve gradations in that color. Is that about it? And one can use the in-camera histogram as a rough guide, realizing that it is a representation of the JPEG the camera would do on its own?

I would have never thought of that. So it is useful to have one more thing to keep in mind in adjusting exposure and making compromises. In the Porsche picture, there need be little concern for preserving shadow detail in the dark corners of the background. In situations where more issues are in question at the same time, bracketing can be in order, I'd expect.
 
  • Like
Reactions: analoggrotto

stevelee

FT-QL
CR Pro
Jul 6, 2017
1,979
735
Davidson, NC
I'm not sure I follow all the ins and outs here. I mean, if the default setting is +16 Magenta, why is it a surprise to have a magenta cast? Wouldn't this say more about the conversion software than about the camera per se? Wouldn't just changing the default in Lightroom take care of the problem, or am I missing something?
 

analoggrotto

EOS RP
Aug 27, 2016
279
161
It looks like a big reason for the red clipping in the Porsche photo is because it has been shot in direct sunlight. The official porsche picture is under studio lights which would no doubt be much softer than direct sun and would have allowed for those gradations in the paint.

After receiving my R5 and taking a few portraits, I am less concerned abou the magenta cast I had noticed previously. After converting the images with the Adobe DNG converter and using the "adobe color" profile, moving the green/magenta slider to 0 or +5 seems to have removed the colour cast completely (default starting point was +14 magenta).

I'll be doing some more testing in mixed lighting conditions to get a better feel for the camera... will let you know how I get on.

Another reason could be that I boosted the reds in LR, switched to PS for some background trickerly, reverted to LR and boosted the reds even further. I really should go back to the source files and figure out what was going on. All I know is that it was quite satisfying in the end and I dont even really like Porsche.
 

Attachments

  • X18_7908-Pano-Edit.jpg
    X18_7908-Pano-Edit.jpg
    305.6 KB · Views: 60
Feb 15, 2020
505
347
I'm not sure I follow all the ins and outs here. I mean, if the default setting is +16 Magenta, why is it a surprise to have a magenta cast? Wouldn't this say more about the conversion software than about the camera per se? Wouldn't just changing the default in Lightroom take care of the problem, or am I missing something?
On the EOS R, the default in Adobe Camera Raw is around +17 magenta and those files have no magenta cast. I am finding on some R5 files I am having to go -10 to sometimes -20 magenta to correct for the magenta cast. There is a massive difference between R raw files and R5 raw files in this sense.

I should also point out that this magenta cast is also present when using the canon DPP4 software which by default is set to 0 +- green/magenta. Using DPP4 I am having to go to +5 green and sometimes as much as +10 green (maximum setting) to correct the colours.

I have no idea why Canon decided to change the colour science so much... but the EOS R is much more to my liking in all of my tests so far...
 
Feb 15, 2020
505
347
Lots of people complained that the 5D4/R images were producing skins tones that were too green. You can’t please everyone when it comes to colours!
That's very true, it is subjective after all. Personally I loved the colours of the 5d4 and EOS R and was hoping they would continue into the R5..
 

Act444

EOS R
May 4, 2011
1,129
201
Lots of people complained that the 5D4/R images were producing skins tones that were too green. You can’t please everyone when it comes to colours!

If I’m honest, I think a red tint is easier to deal with on skin than a green one. Red can be worked with, but green just looks....unnatural. IMO. That’s a change I lamented going from the 5D3 to the 5D4. Of course, no color cast at all would be ideal, but if one had to be picked...

It’s the blandness that I dislike though, not so much the tint. It seems to be challenging at best to replicate older colors, at least in DPP.
 
  • Like
Reactions: Chris.Chapterten
<-- start Taboola -->