3 ms·
Is there a more advanced standard colourspace that we could be using?
by mavhc 4y ago
Is there a more advanced standard colourspace that we could be using?
- exo762 4y agoJust guessing, but that seems to imply two possibilities. One containing colors that can't be implemented - no such pigments / ink / chemicals exist. And other one describing actual inks and their behavior - which sounds like what Pantone is doing.
- mavhc 4y agoI think I'd start with: a list of how much of each range of frequencies of light are reflected, a list of light that's emitted, how matte/mirrored the surface is
- brudgers 4y agoYes, Pantone describes inks using samples of actual products because it is in the business of selling ink. Familiarity is the big design advantage of Pantone color. I look at the samples, pick a color, and that’s what I will get. Whether or not my monitor is calibrated or whether or not I am using a computer at all. From a business perspective, if you’re logo is Reflex Blue, then any reputable print shop can print your leave-behinds and any decent sign shop can make your tradeshow display.
- PaulHoule 4y agoThere are wide color gamut RGB spaces such as https://en.wikipedia.org/wiki/ProPhoto_RGB_color_space https://en.wikipedia.org/wiki/ProPhoto_RGB_color_space that do a better job of modelling many forms of color vision than conventional RGB spaces. The response of the eye under ordinary situations is modeled well with three stimulus values, but real colors are on a spectrum. Consider https://en.wikipedia.org/wiki/Dolby_3D https://en.wikipedia.org/wiki/Dolby_3D which uses a total of six colors and high precision filters such that there is a red that passes the left filter and a red that passes the right filter. The two reds look almost the same without the filters, but with the filters one becomes completely invisible. In situations where light is interacting with something other than your eye (filters, a prism) the physics become vastly more complex. Speaking of 3D images lately I've been frustrated with wide gamut displays because I make images that are viewed through these glasses https://en.wikipedia.org/wiki/Anaglyph_3D https://en.wikipedia.org/wiki/Anaglyph_3D Namely the red channel goes to the left eye and the green and blue go to the left eye. (0, 180, 0) in sRGB is converted by the application (Photoshop, Web Browser) to something like (16, 176, 15) before sending to a wide gamut model. Adding a little bit of red and blue to the wide gamut green makes it look less saturated and more like the sRGB green, but it's bad for anaglyph stereograms because the extra red light causes an unwanted ghost image to appear in the wrong channel.
- patmorgan23 4y agoI think it's less about 'color space' and more about modeling material properties.
- acomms 4y agoThis is about "printing" isn't it? I guess 3D modelling should have spectral qualities - but Pantone is made for spot colouring in print and paint.
- mavhc 4y agoExcept the parent talked about reflections and other frequencies of light. Do we need a next gen colourspace (materialspace?)
- BashiBazouk 4y agoIt use to be and still is to a certain degree but Pantone has been trying to move digital and make consistent colors when printing 4 color. All the old books have ink formulas for spot colors which have limited uses these days. The new books ditch the formulas and give RGB, CMYK and html for each swatch.