Kostas writes: > hi all! here is the cituation I am facing... Hi Kostas.. > I need to make a printing profile (or an PS action) the will make my images properly printable (color, brightness etc) by a printing shop (not photo) plotter... got this bit > so I tried this... > I started with a test print made, and then changed my image to look like the print in terms of color balance etc...then I backsteped by inverting my adjustments, looking in my notes for the recorded values used. In other words I am trying to find a standard method of adjustments & record it as a PS action, that when "played" on my image, it will make it strange-looking in my screen, but will print 100% in the said plotter this confuses me a little. The way I read it, you want to take an image and print it on the plotter but I don't understand why you need to reverse what you're doing. The way I'd approach it would be to take a fairly simple uncluttered small image and add a greyscale and a series of colours on it to start you off.. much like a 'Shirley' print. I'd then take a blank image and copy a whole series of this image onto the blank and apply a diverse number of colour adjustments to each of the individual images, different for each one, and in an ordered manner. Lets say you start by leaving the first alone so it's an unmodified version, then you apply a -20 Magenta to the second image, -15 to the next until you have a sequence up to +20. The next sequence would alter the yellow/blue from -20 yellow to +20 yellow and the final, -20 cyan to +20 cyan to alter the cyan/red balance. Flatten this final image, save it as a high quality jpeg and have the printer print the one image made up of many then you evaluate what it produces. this is a 'ringaround' and something one does when first learning to print colour on RA4 colour paper. It's a good method to learning about colour balance and an effective method of determining what the 'correct' average filtration is for the RA4 paper, or in your case, the plotter. What you'll see is that most of the images look bloody awful but one or two most closely resemble the appropriate colours. From this you would note which adjustment most closely represents the image, then you need need to fine tune the colour adjustments to get the image right. Let's say the -5 magenta looked good but not perfect, you may want to try again with a -7, -6, -4, -3 and -2 ,having the plotter print another group or cluster of images to identify the best. Once you have this close to perfect colour wise, you may need to fine tune the contrast or gamma to get the final accurate version from there what you would do is note this colour variation & gamma adjustment, so whenever you send a print to this plotter in future you save your correct image, then apply the 'filtration factor' or profile to the image - that is the one to go to the plotter. > follow along please... > I made a brightness/contrast adjustment of -20/+10 to approxiamte the test print. reversing it to +20/-10 did not bring me back to my original (obviously, since increasing the contast killed detail in the dark areas)...no luck here> things were more weird in color balance tinkering... > a change of magenta of 0/-20/0 required a change of -20/+20/-20 to be reversed, but in another step in the very same image, 0/+20/0 was enough. this confused me. Don't try to 'reverse engineer' Kostas, it is a hard way to achieve the results you want. > and here it is that i am completely amiss...how do i reverse reliably and repeatably a change in more green added? roll the slider the other direction or tinker with the other sliders? if you ever make a change to an image and discover it is incorrect, you are better off undoing that change through the history menu and applying a more appropriate change. Any change to an image only builds on changes you've already made and some as you discovered with the contrast cannot be undone. Data (colour or detail) is effectively LOST with each incremental step. not good - it's a very bad habit that I've seen many students of photography make. It's a bit like making a copy of a negative or print, and then copying it again.. every step away from the original results in a loss of information and accuracy.. hope this helps a bit Karl
