WYSIWYG - Do we really get it?

[Eivind Røhne]

Speaking of Phocus wysiwyg... Discovered something odd today. Actually, I've noticed it from the very beginning when I started using FlexColor, but I've never really thought about it.

I've attached an image which has a background that was set to blow out to pure white in the studio. I know, it's not art, only meant to illustrate the point... 

According to FlexColor, when hovering the mouse around the picture, the white background reads as 255/255/255. But as you may notice, the white area in the main window, underneath the image actually looks brighter. That's strange... What could be brighter than 255/255/255...? So I took a screendump and opened it in Photoshop. Took a new reading, and gueass what?! The image area that FC claims to be pure white (white studio background) is actually 252/254/253.........! And the main window area underneath the image is 255/255/255. What happened to wysiwyg? That actually means that all images opened in Photoshop will look a little bit brighter than in FC. Haven't dared to test what FC reads as pure black.........

I've got one question: Why?

[robertpoll]

I'd tend to agree with Derek that this is related to monitor calibration. I guess 252/254/253 has been mapped to white in your calibration process, so if the calibration is correct and this really is white then the 255,255,255 area below (that isn't part of the image, and is part of the application so won't be colour managed) is slightly off white (slightly green, and a little bit of blue, right?). If you think about it, it would be quite a coincidence that equal amounts of R, G and B came out as white on any real monitor so white will always look less bright than 255,255,255.

It does seem slightly odd that one of the channels in the white value isn't 255 though I suppose but maybe that's related to rounding in the process somewhere, or maybe the calibration chooses to set the highest channel to 254. Might be interesting to see what happens if you recalibrate.

Presumably when you export to a tif then the area comes out at 255,255,255?

...rob

[robertpoll]

That said, interestingly I've not been able to make this happen on my system. The explanation does assume that the screengrab process isn't colour managed, but I believe that's correct. Eivind, can you let me know exactly what you're doing to get the screengrab into PS - alt/printscr then paste directly into PS, or are you going via another app and saving a file.

...rob

[Eivind Røhne]

That said, interestingly I've not been able to make this happen on my system. The explanation does assume that the screengrab process isn't colour managed, but I believe that's correct. Eivind, can you let me know exactly what you're doing to get the screengrab into PS - alt/printscr then paste directly into PS, or are you going via another app and saving a file.

...rob

Made a screengrab with CTRL+PrintScreen and pasted into a new Photoshop document.

[Eivind Røhne]

Do you have any monitor calibrations going on?

Maybe this is a silly question, but how can monitor calibration affect this at all? I thought the FC or PS eyedroppers read the actual picturedata sent to the display, not the color my eye sees... If I do a completely wacky calibration so the screen goes all blueish or greenish or whatever, or maybe fiddle with the control buttons on a monitor to get all wrong colors, the picture data read in FC or PS would still be the same... Wouldn't they? And the image background and the white of the window receives the same data to be display, don't they, both to be diplayed as pure white (255/255/255)...?

[Eivind Røhne]

Sorry, I just got back. Nothing to do with the monitor controls on the outside of the monitor; it has to do with color mappings in the graphics driver or separate software if you calibrate your monitor. It could also have to do with color spaces and what FC and PC are doing with mapping each color. This is just a thought, for all I know it could be something else. Let us know if you find anything.

Sorry Derek, but I still don't get it... Maybe we norwegians should do what the bears do during winter time. Hibernate...

I have an Eizo CG210 with Eizo's ColorNavigator software, calibrated to 6500 Kelvin etc... If I calibrate my display to, lets say 5000 Kelvin; whites on screen will look more yellow to the human eye than 6500K. Right?! But if I measure the image in FC by moving the mouse over the image, the colors should read the same, no matter if I calibrated to 5000K or 6500K...?! A monitor calibration don't change the image data/color, only how the human eye sees it on screen. Right?!

[NickT]

It's true that monitor calibration does not affect data, but a screen grab  creates NEW data, and will have your monitor's profile embedded. You don't say what you have Photoshop set to do with mis-matched profiles.

What happens when you save out this image from flexcolour and honour the embedded profile? i'm betting you'll be seeing 255,255,255 again

Nick-T

[Eivind Røhne]

It's true that monitor calibration does not affect data, but a screen grab  creates NEW data, and will have your monitor's profile embedded. You don't say what you have Photoshop set to do with mis-matched profiles.

What happens when you save out this image from flexcolour and honour the embedded profile? i'm betting you'll be seeing 255,255,255 again

Nick-T

Hi Nick,

When the file is saved from FC and opened in PS, it's pure white as it should be. The screen grab was meant for upload illustration here only, because the problem is that I can also see this difference between the image-pure-white and the window pure-white on screen while I'm working in FC. So to me it seems that FC doesn't give me the right color(s). If I have FC open (v4.8.6 btw) with the same image open in PS at the same time, I would expect to see the same image, but I can see a difference... The whites are a little bit whiter etc...

Countdown to hibernation commencing...