In a slightly different context, if I am trying to learn how to do something, and I am led to a video how-to, I want to know whether I can find something I can read. Many other people, if they are led to a textual explanation, want to know if they can find a video. IMO, those are different learning styles.
3 Likes
Is there any reason to keep the temperature number when invalid? Instead of trying to find a way to say the number is meaningless, maybe the best is just to remove it and replace it by a “(non blackbody illuminant)” text.
2 Likes
I think that we all can agree that the results appear to be correct and are trustworthy … but … for any ‘seasoned’ photographer who has long term experience with traditional scene temperature colors the modules output can be baffling.
The temperature indicated appears to be incorrect … the word ‘invalid’ confirms that a problem exists , and the color patch appears to be totally irrelevant to the scene.
If one reads the manual, carefully, then maybe the process will make sense but unfortunately none of the 3 above items will be useful in a practical way.
The expected output after clicking on the eye-dropper should be believable even to those who are not color specialists.
Not sure if nobody will die if you wrongly compute the air friction forces, but that’s probably another story 
Let’s keep it invalid or maybe rename that to “meaningless number” 
I think the whole notion of color temperature in post processing is unhelpful. The light of the scene is long gone, and all one is left with is the encoding captured through a set of bandpass filters. Channel-multiplier white balance is far more appropriate to consider, especially when determined from a patch in the scene that is supposed to render “white”.
So, may be the label should be hard-coded "(look at something else…) ?? 
4 Likes
I had a quick look at what learning styles is all about, and right away it smelt like junk. I looked up a couple of the critiques and the problem is that there is no evidence that “learning styles” is some sort of innate trait.
I rather think its like the Myers-Briggs personality testing - thought provoking but not a description of the real world. Myers-Briggs is still in widespread use, mainly due to heavy marketing and if “learning styles” is widely talked about I suspect someone is making money out of it somehow.
2 Likes
I agree and I don’t dispute that it would seem there is no evidence for efficacy of either imposing a certain style or allowing self selection of a particular style on individuals that would lead to enhanced outcomes. However if people believe they are visual learners then as I read somewhere allowing that choice makes them “feel” more comfortable and engaged you can get a bigger “buy-in”…likely why it sticks around as you can see how that could be leveraged for marketing…
Seconded. It’s wrong and stupid, but everyone is used to that crappy way of encoding a color.
3 Likes
I think it ought to be helpful, at least for daylight shots, as a starting point for tweaking. However I think there is a little anomaly which someone could perhaps explain. The same temperature gives different results depending on whether you’re scene-referred or “basic”. So the former is with white balance set at “camera reference” and in colour calibration you set say 6000K for an overcast day. Now switch c.c. off and set 6000K in white balance. The image goes somewhat cooler. Should it not stay the same?
No it’s not. The number is wrong all the time. I got D55 light bulbs 96 CRI, they register at 4000 K because, after input color profile and BS RAW white balance, the resulting illuminant in pipeline is no longer the scene one. You would need to calibrate input color profile and RAW WB on D65 reference spectrum to get a proper reading later.
So WB is just… the color of the illuminant. Lch, Ych, RGB, XYZ, whatever but no K.
2 Likes
Unplanckian
I’m only half joking.
Also I think its important to note that the module was also added to allow for multiple instances later in the pipeline if warranted and maskable. So there are many functions the module could have. If I have a scene and the light on the rock is just too yellow or too warm or what ever I just draw a box on the rock to get the opponent color then go to the hue chroma and take the chroma down to zero and then dial in chroma back in until the light is now where I want it…in fact I really ignore the K reading all the time and mostly switch to the hue chroma sliders. If for some reason I don’t like the calculated WB that is the way I tweak it…
What about matched vs unmatched, ie from what Aurelien said then the value either matches a valid model used by CC for calculation or it does not??
How about something like “tinted” or “not-neutral”. White points that are not on the illuminant axis are some non-zero distance from it, and hence have a tint in either the green/yellow or purple/red direction - i.e. they do not look neutral.
5 Likes
Of the opponent color?
Neutral has too many different meanings, but tinted is probably the most accurate non-frightening word.
3 Likes
non-planckian sounds ok (not a planckian radiator)
1 Like
As much as I dislike the whole thing, I’ll suggest this: “non-illuminant” @gwgill reminded me that there’s a thing called an illuminant axis in the chromaticity plots, and anything that plots off that axis is not one of those…
The scene was lit by a “non-illuminant”? It didn’t illuminate the scene?
Illuminant is whatever light source is lighting the scene. That axis is called planckian locus, aka any tristimulus that results from a black body radiator emission.