On Mon, August 22, 2005 2:55 pm, Warren DeLano said: > > Hold on -- from the end-user's standpoint, the only thing that matters > is: "does the image I see on the screen match what comes out of the > printer". Who cares about the "printer's point of view"?!
The printer (a person running the press) does; not the little ink-jet thingy on the table :-) Printing a PyMol image to a local device (ink-jet or laser probably) needs RGB data not CMYK (anyone who has a printing device that takes CMYK directly and knows how to use it doesn't need any help, they've got all this stuff down pat) which it then converts to its own internal CMYK or CcMmYK definitions. Feeding it CMYK simply results in the printer driver converting first to RGB and back to it's own CMYK. > > In the real world, scientists do not have time or resources to mess with > color profiles and color-calibrated hardware. They simply want to go > from screen to hardcopy knowing their colors will not be grossly > distorted by their hardware, which is often consumer-grade, > non-calibrated equipment. Pretty much. > > Though CMYK is indeed device-dependent, there are some color regions > that, *in practice*, are more problematic than others when printing from > an RGB source. If you stick to safer regions of RGB space that are > pragmatically captured through PyMOL's "CMYK" space command, then you > will get closer to a WYSIWYG experience even in the absense of > calibration. Wouldn't it be better to rename this colour space as printRGB or something to more accurately reflect what it is: a reduced gamut RGB space that trys to avoid some of the colours that routinely produce unwnted colour shifts? > > On the other hand, blithely working in RGB space and then relying upon > automatic RGB->CMYK color translations in the *absense* of color > calibration for both display and printer almost always results in > unacceptably poor color quality, and that is the practical real-world > issue facing PyMOL users. But the journals all have professional printers (the people) who are quite capable of re-casting an RGB file into a CMYK file for their printing conditions. I'm not saying it isn't easier to do that with PyMol's reduced gamut space just that if you want or need a larger colour gamut there is nothing preventing you from using one as long as you have some tools to help you. > > So I would recast your advice as follows: Unless you have calibrated > hardware for both display and printer, then you are definitely not "far > better off getting the RGB image out of pymol to look just the way you > want it to". Instead, you are far better off avoiding areas of RGB > color space that are difficult or impossible to handle without > professional-grade color hardware, and that is the sole task PyMOL's > CMYK capability is designed to help you with. If somebody has gotten to the point where they are worrying about their own CMYK separations then telling them to just use a reduced gamut space to limit the conversion problems is less than helpful to my mind. They need to understand what is going on so they can make the decisions that are necessary if they are to get the best colour reproduction on the journal pages. Rich
