In a message dated 3/15/06 2:49:06 PM, xun911@... writes: > I disagree. The rainbow is very unforgiving I don't think what everything > the rainbow reveals should be taken literally but if you know how to use it, > it's wonderful tool to detect problems. > > Well, given that Tom has posted, as he calls it, a sharp chainsaw in the files section, perhaps a bit of description is in order. Granger Rainbows are cross-color gradients in flat 2 dimensional image files. Since all the colors have one or two of the RGB channels maxxed out, they represent full color saturations. The gradients are basicly the "skin" of the RGB color space they are displayed in. They show nothing about the inside of the colorspace (where photographic colors live) but only define the extreme exterior. I like to describe a 3d graph of an ICC profile as being like a bronze statue of a tornado; yes, it shows a familiar form, but its all about the surface of the tornado, when the important stuff is happening way inside. Granger rainbows take this superficial focus on the skindeep to a new dimension, by focusing our attention on how things look at the surface of a profile only. Few if any of these surface colors will ever be reached by photographic content; many cannot be reached by the monitor they are shown on, or the printer they are pri nted on. They are in the land of vector art and artificial colors. With a Granger rainbow as the active window, do a Custom Proof Setup command in Photoshop, and choose your monitor profile. Then select the gamut limit command, and watch the entire rainbow disappear... its all out of gamut. Do the same with a matte printer profile, and you'll lose nearly as much. A gloss profile has marginally more or the colors in gamut. I often see composite test images, where half the page is devoted to synthetic gradients and Granger rainbows... odd, since the goal is to print good photo color, and there is almost no overlap between these synthetic colors, and photo color, or for that matter, printer color. So yes, there is a purpose to Granger rainbows, but it is nearly as abstract as the rainbows themselves, having mostly to do with out of gamut colors, gamut limits, relations between workingspace gamut lines and printer gamut lines, etc... C. David Tobie Product Technology Manager ColorVision Business Unit Datacolor Inc. CDTobie@... www.colorvision.com >
Message
Re: [colorvision_group] Granger Chart - For evaluations
2006-03-15 by CDTobie@aol.com
Attachments
- No local attachments were found for this message.