Am i reading this Color chart right?

the 7a3 will have a pink rose in the color where as a 7c4 will be more in the amber color range?
This is looking at Color bins for XPL U5 7A3 and the XPG2 Q2 7C4.
What does the U5 & Q2 mean??

Chart:

Yes. The color bins are big enough to fit a truck in, though.

The other numbers are Efficiency ("Brightness") Bins, the lumen output at a given current and temperature. See the Datasheets for more info.

Really a separate issue but I wonder... What color bins are considered "ideal" by manufacturers or consumers? Does the tint need to precisely match the black body curve, or is there a tolerance range?

I know some members in the flashlight section have expressed their preference that their light be just a little rose color tinted.

So this chart is just slapped together. Not NASA approved

Mr. Nobody said:

So this chart is just slapped together. Not NASA approved

Click to expand...

It looks to use the ANSI or CIE color space (With the white swirl through a color space, ROYGBIV going counterclockwise around the perimeter). But the real gamut of the color space is cropped out, because 'White' LEDs had best not be too far off the white-ish center.

Cree made their color bins pretty large to simplify their binning structure. I get 'best' color results in the 4A to 4C space, because I like creamy neutral white. But I prefer to mix 6000K CCT and 4000K CCT in the kitchen, and take things down 1000K in relaxing areas.

Anders Hoveland said:

Really a separate issue but I wonder... What color bins are considered "ideal" by manufacturers or consumers? Does the tint need to precisely match the black body curve, or is there a tolerance range?

I know some members in the flashlight section have expressed their preference that their light be just a little rose color tinted.

Click to expand...

Why would you "wonder" other than to hear yourself type. I have told you I would say at last 5 times on likely 5 different threads that people prefer light just below the black body curve and have provided links to research that shows this.

AnAppleSnail said:

But I prefer to mix 6000K CCT and 4000K CCT in the kitchen, and take things down 1000K in relaxing areas.

Click to expand...

Which would make the tint slightly magenta, when you combine a higher and a lower color temperature source together.

If we draw a line and then find the midpoint between two points on a black body curve (in a CIE color coordinate graph), the midpoint will be a little off from the color coordinates of the correlated color temperature for that point.

This effect becomes insignificant at higher color temperatures (for example, if you combined 4000K with 9000K) but is more pronounced when a lower color temperature is involved (2700K with 5000K for example).

Anders Hoveland said:

If we draw a line and then find the midpoint between two points on a black body curve (in a CIE color coordinate graph), the midpoint will be a little off from the color coordinates of the correlated color temperature for that point.

Which would make the tint slightly magenta, when you combine a higher and a lower color temperature source together.

Click to expand...

Merging averages is a funny business, and a great time to talk about my favorite statistical fallacy, Simpson's Paradox. In short, if we have a true statement about two or more subgroups of a population, and combine the data without care for an unmeasured variable, we will make a false statement about the whole population or measure. So it is possible to show that, with all of its six departments slightly in favor of women, Berkeley selects against women, or that given six age categories of smoker and nonsmoker deaths, that smokers live longer. Still, there are tools available to look at the whole population or data - Modeling or survey.

I just plain wouldn't know HOW to model-average different spectra together, except to get a Spectral Power Distribution for both sources, superimpose them, and then evaluate the locus in the color space I could set that output in. Otherwise, I would have to build the light, and test it, and decide if it is adequate. I have done this previously with Cree XR-Es and gotten light I liked. It works with LED strips and fluorescent lamps, too.