The weakest link in previewing printed images is the computer monitor. It can't replicate the same colours as printed matter (not yet anyway). It doesn't change colour balance according to time of day or year, or react to ambient lighting (so our eyes adjust to its colours), although there are some moves in this direction.
The Context Preview would completely ignore these limitations and exploit the characteristic of our eyes and brains that sees colours as being relative to their surroundings rather than absolute quantities.
You would hold up the preview card (with a calibration image on it) and the card, your desk and your hand would be photographed by a webcam that you install so that it looks at your work over your shoulder. Your computer would work out the ambient lighting and colour cast of the cheap dodgy webcam and would callibrate the image to be previewed accordinging.
Now for the clever bit. On screen the computer displays not only the image to be previewed but displays it as if it were printed on the card*, showing the hand holding it, with the desk as background.
This allows the brain to perceive the colours of the image in context and allows your mind to make any allowances for the monitor, giving you a virtually perfect preview no matter how green your monitor image.
*The maths of recognising the card and replacing it with an appropriately skewed image is not difficult.-- st3f, Oct 29 2002 ...or you could calibrate your monitor...-- phoenix, Oct 29 2002 That's where I originally thought this was going. Use the webcam to image a color, display it on the monitor and tweak the video driver or something until the monitor-displayed color matched the reference color.
I know nothing about photography, imaging, relative color perception, etc. so this might be a perfectly functional idea. I'll wait for other's to weigh in on the subject before I vote.-- half, Oct 29 2002 waugs: yep, that's it, in a nutshell. It's all about context. If the monitor has a flaw you will see the flaw in both the background and the preview. There's nothing stopping you correcting the problem in addition to putting the preview in context.
phoe: Done that several times. I used to have different settings for different times of day (balance between artificial and natural light). Calibration without context doesn't address the differences between colour spaces, though.
half: you can be my neutral observer.-- st3f, Oct 29 2002 If someone slips into your house and replaces all your lightbulbs with slightly yellow ones your eyes compensate. If someone then gives you an object you can tell them what colour it is by comparing it to the known colour of your surroundings and yourself (partiucularly your hands when you pick things up).
By introducing a context (your hand and a bit of desk) around the previewed object you can understand and compensate for the image on your monitor in the same way as viewing things in different lights doesn't stop you from accurately perceiving colour in the real world.
[later:] waugs: I've just understood what you mean. Scratch the above. (This is going to get complex).
There are three colour spaces here -- theoretical space (perfect), photographed room space and monitor space.
The preview exists in the thoretical space and we want to move it to the monitor space, giving it a real context from the room space.
The calibration card exists in both the theoretical space and the room space.
By examining the calibration card (which has a range of tones and colours on it) we can calculate a transformation from photographed room space to theoretical space and from there to any space you choose.
From here we can go in one of several directions. We can transform the preview image into photographed room space before inserting it into the camera shot. We can transform the room shot into theoretical space. Or, we can tranform both into a space that has been created by calibrating your monitor.
Let's say we use the theoretical space. The image of the hand holding the calibration card is moved onto theoretical space, the preview is inserted into the image, and the result is thrown at the monitor.
The context around the preview image will enable you to preview the image accurately. If you're not happy with the contrast, brightness, gamma and colour cast you can adjust them using the context of the preview as reference. For greater accuracy, switch back to the image of the calibration card before the preview is inserted.
Whatever distortion of colour space the monitor introduces, it acts equally on the preview image and surrounding context so the answer is, no, you would not see the flaw through the flaw.
After all that I hope I correctly understand what you mean. Whew...
Note: Keep the calibration card spotlessly clean and handle only by the edges. Any nasty ink splodges on the card would render this whole process useless.-- st3f, Oct 30 2002 [st3f] The calibration is supposed to be the context. If you balance your monitor and/or output to a particular environment, the environment becomes the context. Your images (or the colors therein) won't look the same when you view them under different lighting conditions.
As you say: "If someone slips into your house and replaces all your lightbulbs with slightly yellow ones your eyes compensate. If someone then gives you an object you can tell them what colour it is by comparing it to the known colour of your surroundings and yourself (partiucularly your hands when you pick things up)." But what if the room you're in is one you've never been in before? You have no point of refence - especially if someone is handing you an image they calibrated ('contexted') in a different environment under different lighting conditions.
On the other hand if every monitor in the world (and every output device) were calibrated identically then everyone is using the same reference/context. This also eliminates vagarities in eyesight between people. I think you'll find it's easier and more reliable to remove the green tint from your monitor than to train your brain to ignore it.-- phoenix, Oct 30 2002 I was going to ask how you calibrate the calibration card but phoenix has beaten me to it. Perhaps the answer is to eliminate colour from the equation altogether and preview everything in grey-scale.-- DrBob, Oct 30 2002 Phoe: I'll admit I was a bit cavaler with the lack of calibration of the monitor. Removing a green tinge will help no matter what proofing system you use.
I'm not saying that callibrating monitors is a bad thing, just that even an expensive, well calibrated calibrated monitor provides a poor proof. If you can see how bad a proof it is providing, and in what ways it is bad, you are in a better position to understand the image it is showing.
DrB: I dunno. I just copied the one I was given.
waugs: Yes, the image will be shown with all the monitor's deficiencies, but so will the image of your hand and the area surrounding the preview.
Because you know what your hand looks like and what your desk looks like you can recognise any lack of truth in the image as such, and have a better understanding of how the printed article will look.-- st3f, Oct 30 2002 One major difficulty here is that although human visual perception tends to be concentrated in roughly 'red', 'green', and 'blue' cones, visible light represents a continuous spectrum. It is possible for several different light sources to appear as indistinguishable "white light" when shown on a suitable white surface, but to have some objects which appear identical under some such lighting conditions appear different under others.
Unless your camera can identify the full spectrographic characteristics of its subject, preview accuracy will necessarily be limitted.-- supercat, Oct 30 2002 I must be missing something. Why would you want to see colors more accurately? Seems rather pointless, if not nonsensical.
Now invent a camera or film that can register fluorescent colors and you're onto something.-- General Washington, Oct 30 2002 GW: It's to cut down on all the really expensive (and often unnecessary) proof prints that are generally run through on the slow road to production.
Re a camera that can record fluorsecent colours: Film manufacturers keep adding layers to their films to widen the spectrum of sensitivity (and fill is some of the less sensitive blanks). My guess is that the additional cost of adding several fluorescent responsive layers to an emulsion combined with the lowered contrast of a thicker film will kill this one soon after brainstorming. Digital cameras on the other hand...
...actually with digital cameras you'd be trading cost and resolution. Specific receptors could be built into the CCD to detect specific wavelengths that the other receptors do not detect well and that are important to you.
Trouble is, once you've done this you're going to have to add fluorescent inks to your printer and fluorescent, um fluorescing pixels to your display...
...or you could just gradually increase the frequency response of each of the photoreceptors in a camera or emulsion layers in film so that they get closer to human vision. I hope and assume that this is gradually happening with technological advancement.
Solution with existing technology. If you want to print an image that has a fluorescent colour in it either: prepare the artwork as separations with a layer for the fluorescent ink or separate existing artwork into more than CMYKX where X is the spot colour you want to recreate. It may take a while to get the balance right but you create pictures with colours that lie outside the CMYK gamut using this method.-- st3f, Oct 30 2002 You guys are funny.-- bristolz, Oct 30 2002 <blink>. Ah, reality.-- st3f, Oct 30 2002 Er, fluorescent colors don't have a signature wavelength... it's just that films and digital cameras register intensity as more important than wavelength, whereas with eyes it's the other way around.
And.
There's some weird misapprehensions here. You don't need fluorescent ink to reproduce a fluorescent image, just adjust the palette.-- General Washington, Oct 30 2002 random, halfbakery