What color temperature produces a white patch at Lab 95,0,-2?

[disbellj]

This is theory, so please hear me out. It may be rubbish, but I believe all good things started as theory.

I figure since GRACoL2006_Coated1v2, FOGRA39, and now FOGRA47 all share the same Lab values for paper white, that there is only one color temperature (say between 5000K and 6500K) that will produce a white patch with the closest Lab values to the standard paper white (now for coated #1 AND uncoated #4).

I figure once that color temperature is known that produces the closest Lab values to the target paper white (Lab 95,0,-2) would be the one everyone aiming at these standards would use. Kind of a standard color temperature that actually makes the monitor match the proof or print, instead of it being "6500K" (which is what I'm told by X-Rite to aim to when using Eye-One Match 3 software on my new ACD 30", which does not produce as good of a match as I'd like). So that's what got me to thinking.

I'd like anybody's input on this. Is it possible to get one color temperature that produces this Lab value (or close enough we can definitely say that's the color temperature we would want to aim to)? If yes, then is it feasible that once the monitor is set up to this white point and recommended Gamma and Luminance, then we could soft-proof both GRACoL 7 (or FOGRA39, or FOGRA47) without checking 'Simulate Paper White", but still check "Simulate Ink Black"?

Thank you all,

Don

[kyle]

You can't translate a color temperature into lab because lab values are relative to the illuminant. The reference illuminant is 100/0/0 in lab. Color temperature is also an extreme simplification of a color. You can have completely different colors translate to the same color temperature with any given method, because at best you are condensing two dimensional information into one. It is generally best to have the white of your monitor match the white of your environment because of color constancy. In a room with lighting that is fairly blue, a color stimulus that is "neutral" will appear yellowish. Matching the monitor to the environment allows for more accurate assessment.

[TerryWyse]

It's tricky on a display Don to get a precise "paper white color temperature" since the *visual* appearance of the display's white point is so tied to the color temperature in Kelvin, the luminance of the display and the ambient light/environment.

About the best you can do is put the paper sample in a viewing booth and start iterating different color temperatures until you get a visual match to the paper as you perceive it in the light booth. As a very ROUGH guide, I'd start with a display luminance of 120-150 cd/m2 and then go with color temps from about 5500-6000 K. As you go with luminance levels >150 cd/m2, you can probably drop the color temperature. If your luminance level is <120 cd/m2, say for a CRT display, you'll probably find you'll need to go higher than 6000 K. Right now I'm working on an Eizo CG211 set to 140 cd/m2 @ 5700 K and it looks about right to my eyes.

It's notable that the ISO standards are fairly vague with respect to displays. I think they give guidlines but no specific numbers. I do recall that they recommend that the display luminance be no greater than 160 cd/m2 but I could be wrong on that.

With regards to displays, there ARE no absolutes!

:-)

Regards,
Terry

[GFYE]

Originally Posted by TerryWyse

With regards to displays, there ARE no absolutes!

Amen to that!

[roel78]

Maybe you want to take a look at the UDACT-software. The Ugra Display and Certification Tool offers the possibility of certifying calibrated displays for soft-proofing.
http://www.ugra.ch/ugra-display-anal...n-tool-2.phtml

It recommends a colortemperatur between 5600K and 6000K for papertype 1, luminance at least 120 cd/m2 and gradation at gamma L*.

Here is a report of my Eizo CG221: Report20100118 full.pdf

The softproof for coated paper is very good. But when I look at uncoated paper then the softproof seems to have a bit 'less power' then the printed version. Has anyone the same experience?

[gordo]

Don't forget that the eye/brain auto white balances. That's why a piece of white paper still looks white under a very wide range of lighting conditions (i.e. degrees Kelvin). Also the the eye/brain doesn't have a color memory, as soon as you look away from the monitor to a hard copy, it auto white balances making color comparisons difficult -some would even say impossible.

best, gordon p

[rich apollo]

I match the white point of the monitor to the (controlled) lighting under which I'm viewing the hardcopy, be that a viewing booth or the ambient light of the room (with appropriate bulbs). Typically, I'm around 5300K-5500K

I don't think you can get away from the Simulate Paper White, because of the effect of the profile's white point on the rendering of colors. For example, if you kick out two proofs - one using a relative colorimetric rendering intent and another using an absolute colorimetric rendering intent - the colors are rendered differently. In some colors I measured >10 deltaE.

[disbellj]

Thank you all for your replies.

I just know that I was told by X-Rite to use 6500K (Terry, you and I have talked, so I know you know the details, but these details are for others to read to know why I'm asking).

6500K doesn't match good enough to me. It looks good to my partner, but I guess I am critical on color matching.

Anyways, I remember the temperature being around 5300K when I used my other monitor (before it wouldn't calibrate due to lack of brightness).

I guess there is no color temperature that produces "closest to" Lab 95,0,-2. I guess I figured since color temperature is defined and measurable, looking for a specific Lab color (Lab color is device independent) could be done. Guess I'm wrong.

Thanks again guys!

Regards,

Don

[disbellj]

Originally Posted by roel78

Maybe you want to take a look at the UDACT-software. The Ugra Display and Certification Tool offers the possibility of certifying calibrated displays for soft-proofing.
http://www.ugra.ch/ugra-display-anal...n-tool-2.phtml

It recommends a colortemperatur between 5600K and 6000K for papertype 1, luminance at least 120 cd/m2 and gradation at gamma L*.

Here is a report of my Eizo CG221: Attachment 1073

The softproof for coated paper is very good. But when I look at uncoated paper then the softproof seems to have a bit 'less power' then the printed version. Has anyone the same experience?

roel78,

See in the UGRA report, how L of 94.92 landed at Kelvin 5522? This is basically what I was talking about. So I would try to aim (theoretically) at around 5500K to get around L 94-95. But I see it's not that easy.

[kyle]

You can't really set a monitor to have an L value. Paper can have an L value because you can measure how much light it reflects (with 100 being perfect reflection). A monitor is emitting light, and L values of particular colors are relative to its maximum "white," which is always 100, regardless of how bright or dim it is.