or Connect
AVS › AVS Forum › Display Devices › Display Calibration › D65: 6500K or 6504K?
New Posts  All Forums:Forum Nav:

D65: 6500K or 6504K?

post #1 of 14
Thread Starter 
So I read the Wikipedia page on D65, which discusses why 6504K is the "correct" temperature versus 6500K. I believe it, because I've read similar things elsewhere, but I don't understand why.

I have some specific questions on the subject, but I welcome any and all information on the subject.
  • When physicists revised the Planckian locus, did the entire xyY coordinate system shift?
  • If so, did the XYZ and CIELAB coordinate systems shift as well?
  • If you were to recalculate the SPD for D65 (link), would you use 6500 or 6504? Why?
  • Is Correlated Color Temperature (CCT) different than the color temperature?
post #2 of 14
D65 is a spectral source, so it's definition isn't a CCT(yes color temp and CCT are the same thing). It's defined by the reletive intesity of light between wavelegnths 300-830nm (http://www.cie.co.at/index.php/LEFTM...hp?i_ca_id=298).

We then use a color matching function to turn that spectral response into XYZ data. The XYZ data is then used to derive x,y data (the big Y is actually the value from XYZ).

CIELAB a=0, b=0 is set to the white point of the given color system, so it really is uneffected by any of this. LAB does this because anytime your white point changes your brain adapts an maps that color to white.

Hope that answers that.
post #3 of 14
With respect to the difference in the 6500K and 6504K numbers: the additional 4K is a result of a refinement in the value of Planck's Constant over the years. It is now 0.06% better known than when first derived. It is updated every four years by the Committee on Data for Science and Technology (CODATA).


Larry
post #4 of 14
Thread Starter 
So, D65 didn't move, but if you were to derive the CCT for D65, that moved?
post #5 of 14
No, the CCT did not move. Its position in the coordinate system changed and the definition of D65 was changed in accordance. In other words, what was once defined by the coordinates on the 6500K locus now falls on the D6504K locus. It is the same. Only the "name" has changed.

I don't understand why it seems to be so important to you. Such a small change is totally meaningless in context here. It is only important to physicists working in quantum mechanics. If you're interested in learning more about the constant and its relationship to black-body radiation, I doubt that you'll find it here.


Larry
post #6 of 14
Thread Starter 
The CCT did not move, but it's suddenly 6504 not 6500? That doesn't make sense to me.

And as to why I'm interested, I love the math behind calibration.
post #7 of 14
Okay. But since my instructor days are well in the past, I'll do something that I dislike -- link to Wikipedia.

http://en.wikipedia.org/wiki/Planckian_locus

and

http://en.wikipedia.org/wiki/Correla...or_temperature


Vary Planck's constant (odd concept that) and see the results for yourself. The math is simple but understanding the physics behind it might take you a bit of time.


EDIT:

I'll give you a big hint: Let temperature be a function of a variable times a constant. Now keeping the variable the same and making a change to the constant, you see that even though the initial conditions are the same, the resulting temperature value is different -- even though it has not really changed. Clear, now?




Larry
post #8 of 14
The bottom line is... 6504 is the right number... 6500 is shorthand. Like if the speed limit was 65.2 mph, people would naturally use 65 and we'd all know what they were talking about. Nobody can see the difference between 6500 and 6504 so it's really not important in that context. It's a .06% difference. Color Temp is a pretty terrible "spec" for color anyway. It barely accounts for the amount of green in an image and is massively dominated by the balance between red and blue. You could have a ton of green in the images and it could still measure 6500K. d65 is the standard we calibrate to. If you hit d65, the color temp will be 6500K or (6504K). But you can have 6500/6504K color temp and be WAY off the d65 point.
post #9 of 14
I agree that there is no perceptible difference or should there be any but 6500K is not "shorthand" for 6504K. There is a true physical reason for the difference and it has to do with a more precise determination of Planck's constant. If you want to discuss the physics of the situation, I will be glad to do so in PMs. Some of us are not dummies, you know.

6500K and the xy coordinates where it crossed the Planckian locus of a black-body initially was defined by a committee as D65. Subsequently, a more precise value of Planck's constant was determined and the locus shifted because of this. The committee kept the xy coordinates of where the initial color temperature crossed the locus the same. But this meant that the color temperature was actually 6504K for D65. It's that simple.


Larry
post #10 of 14
Thread Starter 
Quote:
Originally Posted by LarryInRI View Post

6500K and the xy coordinates where it crossed the Planckian locus of a black-body initially was defined by a committee as D65. Subsequently, a more precise value of Planck's constant was determined and the locus shifted because of this. The committee kept the xy coordinates of where the initial color temperature crossed the locus the same. But this meant that the color temperature was actually 6504K for D65. It's that simple.

Exactly what I was looking for. Thanks so much.

To answer my own question, when recalculating the SPD, you would use the old value of 6500, because the equations for CCT->SPD reflect the old Planckian locus.
post #11 of 14
Quote:
Originally Posted by LarryInRI View Post

I agree that there is no perceptible difference or should there be any but 6500K is not "shorthand" for 6504K. There is a true physical reason for the difference and it has to do with a more precise determination of Planck's constant. If you want to discuss the physics of the situation, I will be glad to do so in PMs. Some of us are not dummies, you know.

6500K and the xy coordinates where it crossed the Planckian locus of a black-body initially was defined by a committee as D65. Subsequently, a more precise value of Planck's constant was determined and the locus shifted because of this. The committee kept the xy coordinates of where the initial color temperature crossed the locus the same. But this meant that the color temperature was actually 6504K for D65. It's that simple.

Larry

Whatever. 6504 doesn't matter a WHIT when discussing calibration or color temp unless you are a color scientist working in a lab. It doesn't matter AT ALL why 6504 is the right number. It is the right number and everybody who knows ANYTHING about calibration understands 6504 is the right number, but we all still talk about 6500 BECAUSE the "4" is so far below the threshold of human visual perception compared to "0" that it DOESN'T MATTER if we say 6500 or 6504 when talking about home theater. If I was in a lab working somewhere that it MATTERED whether you were talking about 6500 or 6504, I'd use 6504. If I'm talking to someone about home theater, 6500 is close enough. And I thought I was anal.

FURTHERMORE - I've never seen a video display that could EVER produce a 6504 measurement result two times in a row, even if the measurements are made seconds apart. Even when measuring with a $55,000 labratory grade spectroradiometer in a temperature and lighting controlled lab environment with stabilized and "perfect" power so the power line can't influence the TV/monitor in any way.

AND ON TOP OF THAT - meters used for home theater calibration, even expensive ones in the range of $7000-$25,000 can't produce a 6504 reading two times in a row from a calibrated (non-video) source that is constantly 6504. And the variation you get will be several times the difference between 6504 and 6500.

AND THAT MEANS - the difference between 6504 and 6500 is considerably smaller than the (short- and long-term) tolerances of video displays (including projectors) and meters typically used for home theater calibration.

Bottom line... in the real (home theater) world, the difference between 6500 and 6504 is MEANINGLESS. If you are a color scientist working in a lab where it matters, use 6504. If you are a home theater enthusiast... meh... use what's easier to remember. When we calibrate we (should) focus on how to get each grayscale step closer to d65. Color Temp is meaningless to a calibrator who is focused on the right thing (d65).

I guess I am still pretty anal.


post #12 of 14
^^^^ Well said!
post #13 of 14
Ah come on DOug tell us how you really feel about this.lol.
post #14 of 14
Doug,

Why you even responded to my post with that "meaningless" stuff is a mystery to me. Changing the subject is a standard way of saying "I don't understand what you are talking about."

Of course, without question, there is no visible difference between 6500K and 6504K. I even said so myself. I was simply explaining the scientific cause of the difference between the numbers. But since you dismissed me with an arrogant "whatever." I will return it to your non-contextual response -- whatever.


Larry
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration
AVS › AVS Forum › Display Devices › Display Calibration › D65: 6500K or 6504K?