[Tig] color table: system gives name?

[Steve Hullfish]

I also brought up this point of context being everything and I  
believe it is crucial to the discussion. Would any of you do color  
correction in a room that was neon blue? No. It would affect your  
judgement of the colors on the screen.

Also, think about that classic color theory optical illusion of the  
"Rubik's Cube" with mulitple colors. There's a dull orange on the top  
than seems to "match" a bright orange that's on the "shadow" side of  
the cube. Yet if you block out the other colors, they're both the  
same dull orange, but because one is on the shaded side, your eye  
makes it seem brighter.

Here are a few versions:

http://boingboing.net/2008/02/08/color-tile-optical-i.html

So even WITHIN a single image, our sense of color can be so altered  
by context that describing them seems dangerous. Of course, we need  
methods to communicate creatively, but I don't know that the  
discussion can be anything other than with the context specific  
references of the two (or more) people involved. In other words,  
having all creative people learn specific terms for specific colors  
just can't work.

On Jun 23, 2008, at 4:32 AM, Richard Kirk wrote:
>
> On the other hand, context is so often everything. In candlelit  
> shots, flesh tones and blonde hair are orange. They are often a  
> bright out-of-film-gamut orange. But, because it is a face and not  
> an orange, it doesn't look like the color it is called. I guess we  
> want words that describe how the color appears to us. We have sRGB  
> values if we want to define the absolute color on a display with an  
> implied device white. This would mean a R=G=B grey which would be  
> 'grey' on a sensible sRGB definition could be blue-grey, steel  
> grey, charcoal grey, battleship grey, and all the different slight  
> shifts from white that we can see, but we cannot always measure.