p.2 #1 · Color cast with Red River paper Arctic Polar Luster and Epson 3880
Between this and your other thread, there's really no need to be rude and intolerant, esp. here @ FM which is one of the more civil places.
Phrasing and intent are sometimes hard to discern in forums.
That said, they're both essentially everyday, throwaway papers. It's possible that each one requires slightly different tone/color curve/layers (ala softproofing) to achieve optimum results. I started w/ the Epson, went to the RR for a couple of boxes and am now on Ilford Pearl for a while (and testing out the Canson Satin)
p.2 #2 · Color cast with Red River paper Arctic Polar Luster and Epson 3880
Have you tried printing through Photoshop rather than Qimage? As I said before, I get as good or better results with RR papers than Epson using the standard RR profiles.
If you let printer manage colors the selection of paper profiles is limited to the brand specific papers embedded in the printer drivers. Not surprisingly you'll usually get good results if using printer brand papers with printer managing color because the "canned" profiles while generic are optimized for the papers.
There is a good bit of science that goes into the formulation of inks and paper coatings. The inks of one brand paper will usually not work as well in another brand's printer. Third party papers will usually come with a data sheet showing which of the canned profiles will work best for that paper for users who can only print with printer managing color because they lack an application like Photoshop that can manage the process. The recommendations for driver profile to use varies by brand of printed because each manufacturer formulates the inks and paper coatings differently. As with printer brand-to-printer brand miss matches it's reasonable to expect some third party papers will perform with one brand's ink formulations than the other.
Printing with Photoshop or another application managing color allows the use of third party profiles supplied with the third party printer and stored on the computer. As the link above suggests when a third party printer is used results are generally better when the printing is managed by the application using the matching third party profile. In theory that should produce a correctly color balanced print of a test file containing an R=G=B grayscale and color reference objects. In practice the results will only be as accurate as the third party profile being used.
Here where the problem is a perceived magenta cast the key word is "perceived". For example if you took a file known to be R=G=B in all the steps of gray scale embedded in the file and it was printed perfectly neutral on the print if you took that perfect print and compared it to the image on monitor with a green bias and your brain trusts the monitor as your frame of reference then the print will be PERCEIVED as having a magenta bias RELATIVE TO THE TRUSTED MONITOR BASELINE. The problem with that type of visual comparison is that human color perception adapts based on expectations. IF the file values in grays are R=G=B and you think your monitor is correctly calibrated then your brain will see the grays as neutral even if you monitor actually has a bias. I'm not saying for certain that's the cause here, but it is one possible explanation.
What you need to determine whether or not the paper profile is working correctly is some objective means to measure whether the CYMK inks are in fact reproducing a gray scale in a test file neutrally. One way to do that would be to print a color balanced (R=G=B in grays) photo of MacBeth target using the paper profile and application managing color then simply lay the actual MacBeth target next to the print under daylight. If an inaccurate profile is causing a magenta or other color cast in the print it will obvious by comparison with the original target.
If the print and original target match each other but print and target don't match monitor it will tell you the problem is the calibration of the monitor, not the printer profile.
If the print and target don't match and the print has the same bias you are seeing by comparison on your monitor then it will tell your monitor is probably well calibrated but the profile for the paper isn't a good match to that particular batch of paper. Batch variances are just a reality of any manufactured product. Trying a different batch number lot of the same paper would be one way to diagnose that problem. It the variation may also be due to the way the application is interpreting and applying the profile to manage the printer.
It is also very likely that the bias is the result of a combination of these factors: judging color on a monitor that isn't perfectly calibrated against a print produced with a generic third-party profile which isn't a perfect match to the batches of ink and paper you are using. A more foolproof way to diagnose the problem would be to print a profiling target on the printer / paper and have a competently created custom profile created for the batch of ink and paper being using.
The simpler and more immediate solution if you have the same consistent magenta cast in the print vs. your monitor image is to create a separate copy of the file for printing and apply a - magenta correction to it before printing
p.2 #4 · Color cast with Red River paper Arctic Polar Luster and Epson 3880
"A more foolproof way to diagnose the problem would be to print a profiling target on the printer / paper and have a competently created custom profile created for the batch of ink and paper being using. "
This is actually the ONLY way to insure good results all the time. Having printed on more types of paper than I can remember, they all have one thing in common. I made a GREAT custom profile for each one and they ALL print great. That makes the choice between papers come down to personal preference for the surface, texture, weight, etc, rather than brand a looking worse than brand b because of a profile of unknown origin.
You all have to have all of your ducks in a row, or you're just spinning your wheels, trying to reinvent those wheels. It's a waste of time and money and leads to false assumptions regarding one brand over another. Since the goal is always to have the best screen to print match, every variably has to be dealt with. A good monitor with great calibration and profiling are step one. Proper viewing conditions are mandatory. The right output profile for the media is crucial - and you just can't rely on the manufacturer's ones. Sometimes they're just okay and often times they suck. The difference between an okay profile and a great one is enough to change your mind about any given paper.
p.2 #5 · Color cast with Red River paper Arctic Polar Luster and Epson 3880
From the Abobe tutorials linked above:
"Monitors display images using light, whereas desktop printers reproduce images using inks, dyes, or pigments. For this reason, a desktop printer cant reproduce all the colors displayed on a monitor. However, by incorporating certain procedures (such as a color management system) into your workflow, you can achieve predictable results when printing your images to a desktop printer. "
I share that view regarding goals. My goal isn't to match print to screen since I never view both that the same time and I understand an exactly match isn't possible given the monitor and printer I own in part because I have an iMac monitor.
The reason I don't spend a ton of money on a high end monitor that will better match the printer gamut is because I also understand that when I capture a full range image with neutral color balance there's very little intervention needed on my part to make the image on the screen and the print as good as they can possibly look within the physical limitations of their respective gamuts.
In other words if you start with a good file and don't screw with it based on what you think is in it, judged from the monitor image, it will, if printed with an accurate press profile, produce a print that look like the scene you saw in person. When profile based color management works as designed which as Peter says requires custom profiling of the printer for each printer paper / ink set the results, as Adobe says, are predictably good from within the frame of reference your equipment provides. If you start with a file with full tonal range and neutral color balance you will get a print with a full tonal range and neutral color balance.
Lining up the ducks isn't much use unless you can figure out what in the flock is happening. You can't control what you can't measure and "getting your ducks in a row" requires starting with references in test image you can monitor throughout the workflow. http://super.nova.org/TP/DuckButts.jpg
People who get confused and mess up the color work flow in ways they can't figure out typically don't use control targets at capture they can measure and correct color to in post processing. Without a target as a reference baseline they start fiddling with the color based on their perception of the monitor image, which may or may not be perfectly calibrated. Even on a perfectly calibrated monitor staring at a brightly colored image while editing will skew color perception to the point you and start with a file that has perfect color balance and introduce a color imbalance when editing.
p.2 #7 · Color cast with Red River paper Arctic Polar Luster and Epson 3880
All I have to say is that virtually everyone who manipulates images and makes prints does so primarily by what they see on their monitors - everyone except Chuck, that is. No matter. The reality is that no one leaves their images "as is" - how they came out of their cameras. We manipulate not only because we can, but because we can make better, more effective images that way. For those old enough to remember printing a black and white neg, how many of you ever made a "straight" print with no burning or dodging? That's what I thought. We need to see reliably what's on our file, and be able to make good decisions about making improvement - improvements beyond simple white and black point settings.
Everyone has their own comfort level of what they can afford in terms or achieving good monitor and print calibration, and I don't expect everyone to go as far as I do, but it's good to know what's available if you want to improve your game. It's also good to know that when all the pieces fall into place, this color management stuff works amazingly well, and even given the differences between emitted light from monitors and reflected light from prints, you really can achieve what a lot of people would term, a match. And even in the areas where there are small problems in displaying colors accurately, they're usually not of a nature that alters your overall perception of the print, plus with experience, you develop an automatic internal translation in much the same way we used to going from Polaroid to Ektachrome.
Doug - No music this Christmas. My gf played the KCET Christmas eve show with the zydeco band she's a part of but all I've been listening to is recorded. Muddy Waters' first Alan Lomax field recordings from '41 and '42 is pretty cool.
p.2 #8 · Color cast with Red River paper Arctic Polar Luster and Epson 3880
Peter you continue to misunderstand and misrepresent what I suggest. All I'm saying is that if an image is captured with correct color balance there shouldn't be any need to change it based on what is seen on the monitor because profile managed color will fit it with correct color balance to the gamuts of the working space. monitor, and printer in a way that looks "normal" relative to the scene as perceived by eye.
I do a lot of manipulation based on what I see on the monitor, but I don't rely solely on the monitor to as my "trusted baseline" to judge color because I know my brain adapts my perception of color and what I think I see might not be what is actually there. I use gray cards in test shots as a reference, set custom WB off that card, then open the test image with that card as the first step in PP in evaluating the image, objectively, on the monitor.
Here's an example of my workflow. After setting lights and custom WB off a plain gray card I have the subject hold this target for a test shot....
Out of camera the RAW file has R=G=B eyedropper values in all the B&W patches on the target. I can verify this with the eye dropper tool. I have an object I know to be neutral the target an a way to verify it the eye dropper readings.
What you don't seem to take in to account is how image content can affect color perception when looking at it on a monitor. Our eyes, looking at bright saturated colors in an image on a monitor will adapt in a way similar to how we think a shirt is white regardless if it is daylight or tungsten illumination. So even if your monitor cost $4,000 and is calibrated perfectly your brain isn't. The color content in the image will affect you perception. In short you can't trust what is seen on a monitor completely not because if its gamut or it's state of calibration BUT BECAUSE THE BRAIN ADAPTS IT'S PERCEPTION BASED ON THE CONTENT ON IT.
That adaptive process when looking at a monitor is the reason it is important to create a "neutral zone" around the monitor in the room environment. Looking way from the monitor at a wall that is painted in a known neutral shade is for your color perception like sending a fighter to a neutral corner, allowing it to recalibrate to a known neutral value it trust as being the "correct" color.
Including a gray card and other reference targets in a test shot has the same effect. The screen shot above is from the Camera Profile screen of ACR. Abobe, when adding cameras to ACR, also creates profiles which simulate the styles in the camera. Styles alter color balance selectively but without changing the R=G=B balance in neutral tones. For example below is a RAW file of a MacBeth target shot with "neutral" style then rendered in different ones in PP. In that case I was using DPP but ACR works the same:
You can see how while the colors shift around in hue and saturation between styles the application manipulating the color knows what content is neutral and keeps it that way. What is the "correct" or most accurate color? None of them, but that really doesn't matter. What matters is finding the color rendering which works best for whatever is on the screen. I like you and everyone else to that based on what is seen on the monitor, but I do so from A BASELINE I CAN TRUST AND MEASURE the gray card in the test shot.
Color is subjective and the style that looks great for a landscape will not be the best choice for a portrait... http://super.nova.org/TP/ColorChart_LS.jpg
I will try and compare different adjustments until by comparison one-to-another, I arrive and the selective color balance I find works best for that particular subject's skin tone and clothing. I do my color adjustments in the Camera Calibration /Styles screen in ACR not with with the Temperature / Tint sliders.
As you understand (but others might not) the difference between adjusting on the Camera Profile screen vs. just moving the Temperature / Tint sliders on the main screen is what happens in the neutral tones. Temperature will shift the neutral tones yellow <>blue and Tint will bias them green <> magenta. That is to say if you were to shoot an image with Custom WB and include the same card in a test shot it would have R=G=B values on the card when first opened, but if the Temp slider is moved the RGB values on the card and neutral image content will shift.
By using the selective color controls on the Camera Profile screen on the RAW file I can make a skin tone warmer or cooler to better match what I see by eye on the person's face but the camera didn't capture accurately without change the tone of their white sweater that the camera did record accurately.
Of course there are instances where I will want to introduce a warm / cool basis to the neutrals with a global adjustment with the Temperature slider to convey mood, and also shooting situations such and the greenish bias of light in a forest where I will need to make an overall global adjustment with green <> magenta slider. But I will make those decision after making my initial evaluation / adjustment of color selectively from the neutral baseline. After I adjust my test image to my satisfaction I will apply the adjustment meta-data to the other RAW files in the batch eliminating the need to correct each one individually and making the adjustments and appearance more uniform across all the images in the batch.
It's not the workflow you use perhaps but one that worked for me quite well for 30 years making color separations from color images and managing printing plants. I simply adapted those methods to the process control of color in my photography. In printing a control target is printed in the border of every press sheet to enable the pressman to measure ink density, trapping, slur, dot gain and a host of other variables both by eye and via instrumentation. Without the target and the objective measurements of the densitometer what was actually happening on the press would be an guess.
Because human vision adapts everything we see is a subjective judgement, not an objective one. Our eyes will adapt perception of the scene in person and on a monitor similarly. I use reference targets in my images, when possible, as my trusted baseline for color, not just my monitor. By setting Custom WB off the gray card I know the color out of the camera should be R=G=B on the card and overall content. When I open the file to start editing it I can both measure the card to confirm the camera did the custom WB correctly and verify my monitor calibration because if it is off too far the card and image will not look neutral. Throughout the process of adjusting contrast and color on the image by having the reference gray card and MacBeth chart in the image I can more easily see from an objective baseline what the is actually causing the changes I see in the image.
So Peter you are correct when you say I do not base color "primarily" by what I see on monitor. My primary baseline for judging color are process control baselines I can measure, starting with Custom WB and a gray card as my primary trusted standard and then adjusting my images from that trusted baseline based on what is seen from that known "neutral corner" in making the subjective judgement of how to change what the camera captured. In other words I use the monitor just like you do, but don't make it my primary standard because I understand how my perception of its content is subject to adaptive interpretation by my brain.
I share your view that the profile based color management is quite remarkable. In fact so much so that I trust that when I capture an image well in the camera that technology will take care of managing to produce good looking screen and prints without the need for much manipulation by me. That is after all what is is designed to to, no? Any manipulation needed is a matter of compensating for the inherent technical limitations of the capture process (such as applying USM) and artistic interpretation of what was actually in front of and captured by the lens.
p.2 #10 · Color cast with Red River paper Arctic Polar Luster and Epson 3880
I use RR Arctic Polar Luster as my main paper. The results match WHCC, Millers and other Pro labs. I calibrate the monitor and printer with the X-Rite ColorMunki Photo.
