That old saw “an ounce of prevention is worth a pound of cure”, is applicable to the problem of bright LCD displays causing prints that are too dark, but at the moment that ounce is one ounce of gold. The ounce in this case is an LCD display that is not too bright, that can be adjusted, calibrated and profiled to match the range of values in a print. The one brand that currently has that capability as delivered is Eizo with their CG/CE ColorEdge displays. I’ll soon be receiving their least costly, the CG222w that has a list price just under $1,500 for test and review. I realize few of my readers want to spend that much, or can afford to, even for a display that does not cause the prints too dark problem.
So I am also into exploring less costly alternatives like the calibration and profiling software Color Eyes Display Pro (http://www.integrated-color.com), and some more modest priced displays that will adjust, calibrate and profile to a brightness level that matches the level that was usual with old CRT monitors, which did not cause a prints too dark problem. But on a broader and deeper level, the fact so many who have communicated with me about their bright LCD displays and dark prints don’t want to spend much on displays, I find troubling. To even be in this position digital photographers at least have to have a significant investment in a computer and printer, but I believe most put a display low on their spending priority list of what is essential to do good digital photography; when the display, its quality and performance, should be at the top of their list of priorities. Almost any personal computer today with a decent video card and a medium allotment of RAM will do Photoshop and most other digital photo applications and processing. So to be pragmatic and realistic it is not unreasonable or too extravagant to spend as much or more on a display than on the computer running it. Yes, that makes sense if you consider everything you do with digital images on a computer is controlled by what you see on-screen, so the better you can see all there is to see in your images on-screen, is the only way to be sure they are as good as you expect when finished or printed.
Getting back to the practical problem at hand, a search for displays that will function at the same brightness level as a CRT is my next task. A couple of years ago when I was testing and reporting on a lot of different LCD displays, I attempted adjusting, calibrating and profiling some at a brightness of 90.0 CD/m2 with the colorimeters and software available at that time, and although some would adjust that low the screen reproduction quality and color performance that resulted was unacceptable compared to running the displays at 120.0 CD/m2. One of those I bought, a LaCie 320 (which is a re-labelled NEC 2090 UXi). And recently I obtained an X-Rite i1Photo, a comprehensive color management suite of software and hardware to calibrate and profile displays, scanners, printers, the whole gamut. I used it to re-adjust, calibrate and profile my LaCie 320 at the lower brightness aim point of 90.0 CD/m2. I was surprised, it was successful, and I believe it was in part because the i1 Photo software adjusts the display through the DVI connector using DDC to interface with the display’s firmware, while in the past most of my attempts were made using the display’s control buttons and OSD to make the adjustments. I then tried to do the same thing with the ColorEyes Display Pro software and a Spyder3 colorimeter, again adjusting, calibrating and profiling the display through the DDC/DVI interface to match all the aim points I had chosen including a white luminance of 90.0 CD/m2. The result was an adjustment that provided as good reproduction quality and color performance as before at the higher brightness of 120.0 CD/m2, and a profile that checked out as good or better than the previous display profiles I had been using.
I put this to a practical test color correcting a set of photo image files, and then printing them, and voila, perfect print densities without any corrections of any kind, just using a standard color managed printing workflow. These test results do not just back up the points I made above, they substantiate the evidence that the cause of prints too dark is LCD displays that are too bright.
More of my writing on digital photography is available through my web site at: https://sites.google.com/site/davidbrooksfotografx/
Showing posts with label prints too dark. Show all posts
Showing posts with label prints too dark. Show all posts
Saturday, February 21, 2009
Tuesday, February 3, 2009
THE UNBEARABLE LIGHTNESS OF BEING
This rather enigmatic movie title came to mind this week after engaging in both live and virtual e-mail conversation about the problem of “prints too dark” I have reported and commented on recently. It seems those most likely to have a knowledgeable understanding of the technical cause of this phenomenon, and any possible solutions are the least likely to want to admit the problem even exists. And if photographer computer users are obtaining too dark prints it is because they are at fault: either their home/office environment in which they do their photo processing on a computer is non-standard or their illumination for print viewing is inadequate.
I can’t really fault the experts when called to the plate for ducking and weaving when thrown a wild pitch, and this “prints too dark” problem has blind-sided everyone in the business except some professionals using sophisticated and expensive tools and techniques beyond the pale of what most enthusiast photographers can or want to afford. So you cannot expect a solution from them when they have not themselves experienced the problem of prints too dark.
In the last couple of weeks since one of my e-mails on this “prints too dark” was redistributed to a large audience I’ve received a lot of feedback and commentary describing a great diversity of perspectives on the problem. However the consensus definitely puts the blame on too bright LCD’s as the primary cause of getting prints that are too dark. That makes sense to me especially considering that the “prints too dark” problem has grown since and proportionally to the adoption of LCD’s for computer display purposes. Almost as much consensus exists for the idea that calibrating and profiling an LCD display to an aim point of a white luminance measurement and setting of 120.0 CD/m2 reduces the dark print problem considerably. A small contingent of responses from primarily color management consultants suggests the problem is the ambient lighting in the environment in which the computer display is used. Considering the problem seems to be greatest the brighter the LCD display is, this really seems unlikely unless we were talking about laptop users, and that is not the case in terms of the reports of problems on file, nearly all involved desktop displays.
The good news is that I just had an opportunity to test some software that is promising and pointing to a way to correct for the too dark print problem. It is at an early stage of development so there is still much that needs to be done to result in an effective solution. But what I have seen so far suggests that this “prints too dark problem” can be resolved. Keep your fingers crossed, all of you, as I will, hoping that soon this plight of digital photo printing will be lifted and we can look to better days ahead.
E-Mail: goofotografx@gmail.com
I can’t really fault the experts when called to the plate for ducking and weaving when thrown a wild pitch, and this “prints too dark” problem has blind-sided everyone in the business except some professionals using sophisticated and expensive tools and techniques beyond the pale of what most enthusiast photographers can or want to afford. So you cannot expect a solution from them when they have not themselves experienced the problem of prints too dark.
In the last couple of weeks since one of my e-mails on this “prints too dark” was redistributed to a large audience I’ve received a lot of feedback and commentary describing a great diversity of perspectives on the problem. However the consensus definitely puts the blame on too bright LCD’s as the primary cause of getting prints that are too dark. That makes sense to me especially considering that the “prints too dark” problem has grown since and proportionally to the adoption of LCD’s for computer display purposes. Almost as much consensus exists for the idea that calibrating and profiling an LCD display to an aim point of a white luminance measurement and setting of 120.0 CD/m2 reduces the dark print problem considerably. A small contingent of responses from primarily color management consultants suggests the problem is the ambient lighting in the environment in which the computer display is used. Considering the problem seems to be greatest the brighter the LCD display is, this really seems unlikely unless we were talking about laptop users, and that is not the case in terms of the reports of problems on file, nearly all involved desktop displays.
The good news is that I just had an opportunity to test some software that is promising and pointing to a way to correct for the too dark print problem. It is at an early stage of development so there is still much that needs to be done to result in an effective solution. But what I have seen so far suggests that this “prints too dark problem” can be resolved. Keep your fingers crossed, all of you, as I will, hoping that soon this plight of digital photo printing will be lifted and we can look to better days ahead.
E-Mail: goofotografx@gmail.com
Tuesday, January 13, 2009
TOO MUCH INFORMATION AND TOO FEW ANSWERS
The “prints too dark” problem inadvertently involves almost every aspect of computer image processing, displays, printers and their drivers, color management, and of course image editing. Yet it is just a single issue of prints that don’t match the screen in density/brightness. The explanation of why the problem seems insoluble is that all of these separate but involved functions are the domains of companies which compete with one another in a marketplace that discourages standards. For instance the driver interface for each brand of printer is different, as are the on-screen controls of each brand of LCD display. And of course the people who run these companies and the technical people who design the products do not see reality from the same perspective as users, who of course are an even more diverse population.
However, the “prints too dark” problem is a rather universal shared frustration that reduces individual satisfaction with the digital photo experience. So where do we look for answers and solutions? I have been exploring that with so far little success and a lot of time and effort consumed. Which made me think about what a better digital photo reality would look like, that would not have this problem caused by a differences in the brightness of LCD displays compared to prints on white paper.
That this is a problem is because each brand and model of LCD display tries to be brighter and contrastier than the competition because consumer think they want brighter and contrastier, yet the adjustment controls on every LCD display do not provide users with an exact method to achieve a known, objective brightness and contrast for a specific use like editing and processing digital photos. Those specific performance factors are largely agreed on; and they are a display color temperature of 6500K, a gamma of 2.2, a white luminance of 120.o CD/m2. So if LCD display makers would accept those aim points as standard and provide in the display controls a single option “PHOTO” that would preset the display to output at those levels, everyone would have a similar and dependable way to use their computer’s display to do digital photography more effectively and reliably in terms of results. Of course if LCD display makers offered a standard PHOTO preset, they would also need to provide a display profile file that accurately reflected the display’s characteristics in that PHOTO mode.
However that still leaves a part of the problem unresolved, and that is that an LCD display ideally adjusted, calibrated and profiled, is still brighter in luminance range than what a print on white paper can be, and that differential will result in prints that are 10 to 15% too dark relative to the screen image.
There are two ways to make photo prints. One is to let the printer driver adjust the image values to internal standards determined by the driver to achieve a good print, more or less as the printer designer sees what should be an ideal print image. The other is to have the image editing application from which a photo is being printed control color, which is referred to as a color managed workflow, which is a precise method to match the print output colors to what is seen on-screen. Unfortunately, the printer paper profiles used only contain information to provide color matching, the profile and color managed workflow does not match screen image brightness to print density because there is no density midpoint information in a profile to tell the print driver how light or dark the print should be.
Ideally then the color management system needs an upgrade to include brightness range midpoint matching correction. But such an overhaul of the entire color management system would be a massive undertaking and require possibly years and far too much cost to implement.
An alternative then must be in the hands of the printer companies to add an output print density adjustment for color matched printing in the printer driver dialog. Or color managed photo imaging application software needs to include a way to adjust the midpoint image brightness of the file information that is sent to a printer so a correct print image density is achieved. Implementing this may be most easily achieved with a plugin utility similar but simpler than Photoshop’s output Transfer Function.
How will those who have the means to facilitate a fix for “prints too dark” will respond to the problem remains to be seen. I have done just about all I can in my role as a gadfly to define and describe the problem as I have come to understand it and convey that information by all the means available to me. I am open to discussing the problem and possible solutions, but to be candid, I want a fix I can offer to users who communicate their frustrations to me, as well as a rest from writing about this same thing that has dominated my attention for weeks now. Anyone with a problem or a solution is welcome to contact me at goofotografx@gmail.com, or if more background is needed on the general subject my Digital Darkroom Resource CD is available and is described at: https://sites.google.com/site/davidbrooksfotografx/
However, the “prints too dark” problem is a rather universal shared frustration that reduces individual satisfaction with the digital photo experience. So where do we look for answers and solutions? I have been exploring that with so far little success and a lot of time and effort consumed. Which made me think about what a better digital photo reality would look like, that would not have this problem caused by a differences in the brightness of LCD displays compared to prints on white paper.
That this is a problem is because each brand and model of LCD display tries to be brighter and contrastier than the competition because consumer think they want brighter and contrastier, yet the adjustment controls on every LCD display do not provide users with an exact method to achieve a known, objective brightness and contrast for a specific use like editing and processing digital photos. Those specific performance factors are largely agreed on; and they are a display color temperature of 6500K, a gamma of 2.2, a white luminance of 120.o CD/m2. So if LCD display makers would accept those aim points as standard and provide in the display controls a single option “PHOTO” that would preset the display to output at those levels, everyone would have a similar and dependable way to use their computer’s display to do digital photography more effectively and reliably in terms of results. Of course if LCD display makers offered a standard PHOTO preset, they would also need to provide a display profile file that accurately reflected the display’s characteristics in that PHOTO mode.
However that still leaves a part of the problem unresolved, and that is that an LCD display ideally adjusted, calibrated and profiled, is still brighter in luminance range than what a print on white paper can be, and that differential will result in prints that are 10 to 15% too dark relative to the screen image.
There are two ways to make photo prints. One is to let the printer driver adjust the image values to internal standards determined by the driver to achieve a good print, more or less as the printer designer sees what should be an ideal print image. The other is to have the image editing application from which a photo is being printed control color, which is referred to as a color managed workflow, which is a precise method to match the print output colors to what is seen on-screen. Unfortunately, the printer paper profiles used only contain information to provide color matching, the profile and color managed workflow does not match screen image brightness to print density because there is no density midpoint information in a profile to tell the print driver how light or dark the print should be.
Ideally then the color management system needs an upgrade to include brightness range midpoint matching correction. But such an overhaul of the entire color management system would be a massive undertaking and require possibly years and far too much cost to implement.
An alternative then must be in the hands of the printer companies to add an output print density adjustment for color matched printing in the printer driver dialog. Or color managed photo imaging application software needs to include a way to adjust the midpoint image brightness of the file information that is sent to a printer so a correct print image density is achieved. Implementing this may be most easily achieved with a plugin utility similar but simpler than Photoshop’s output Transfer Function.
How will those who have the means to facilitate a fix for “prints too dark” will respond to the problem remains to be seen. I have done just about all I can in my role as a gadfly to define and describe the problem as I have come to understand it and convey that information by all the means available to me. I am open to discussing the problem and possible solutions, but to be candid, I want a fix I can offer to users who communicate their frustrations to me, as well as a rest from writing about this same thing that has dominated my attention for weeks now. Anyone with a problem or a solution is welcome to contact me at goofotografx@gmail.com, or if more background is needed on the general subject my Digital Darkroom Resource CD is available and is described at: https://sites.google.com/site/davidbrooksfotografx/
Saturday, January 3, 2009
THE PERCEPTION FACTOR
Have you ever blown through a red light in your car, and as soon as you got through the intersection you realized what you had just done? Other than worrying about whether a cop saw what you did, you may have realized that your eyes saw the red light, but your mind did not register the perception and respond as usual so you could stop and wait for the light to change. What this kind of incident illustrates is that human vision is made up of two distinct functions, what our eyes see and what our mind perceives. As well as a third factor memory, which allows us to not pay conscious attention to everything familiar our eyes see in the course of daily activities, otherwise we would never get anything accomplished if we had to consciously deal with everything in our vision familiar or not.
So what does this have to do with photography and any current digital problems like “prints too dark”? Back in the 70’s an intense interest in Ansel Adams Zone system was appended by Adam’s stress on “visualization” as a way to exercise some control between perception and a predictable photographic result. Around the same time Richard D. Zakia at RIT published a book of articles and scientific papers called Perception and Photography that was rather widely read providing a more modern understanding of human vision and how it can be more successfully employed as part of the process of making effective photographs and controlling the performance of the media. Thirty some year later perception has a somewhat different photographic role and even more important imaging function with photographs that are the result of what is done visually on-screen to edit and manipulate the attributes of a photograph in applications like Photoshop. Which brings us to the current problem of a mismatch in density or image brightness between what is seen on-screen and sometimes in a very different, too dark, print result.
That there is a problem today is due, in large part, to the introduction of very much brighter LCD displays compared to the CRT’s monitors that preceded them. What is perceived on screen does not match what results in a print in density, even though in a color managed workflow the colors seen on screen can be accurately reproduced in a print quite reliably. It should be noted here that digital camera JPEG images adjusted by a camera’s post exposure processing, if just opened in a computer and printed without editing, will very likely NOT suffer the “prints too dark” problem. So the only sure way the LCD brightness problem will result in print darkness is if the image is edited by PERCEPTUAL adjustment for brightness on-screen with an LCD and the file is then saved and printed.
Sometimes just reading the data and coming to conclusions on the basis of the fact an LCD display, even when optimally calibrated and profiled to a brightness luminance of 120.0 CD/m2, compared to the lower range of a typical CRT monitor with an average white point luminance of 90.0 CD/m2 (which did not induce “prints too dark”), is not totally convincing nor entirely complete evidence. So, fortunately for this research, I had a very good Sony Trinitron 520 Pro model 21” CRT stored for the last 3 years and after only a couple of years of use. And it just happens my Mac Pro will run two different kinds of displays in mirrored mode with both having the same signal image on-screen, as well as both being individually and independently calibrated and profiled. So after some digging in my closet I dragged out that 100 pound monster and set it up beside my NEC Multisync 2090 UXi LCD display. Both screens happen to be the same measured size, but once turned on and warmed up they do look very, very different and it is not just the NEC LCD display is a bit brighter. There really is not that much brightness difference as the Sony 520 CRT produces a maximum white point density on the high side at 97.0 CD/m2 compared to my LCD sitting next to it at 120.0 CD/m2.
Even though both were calibrated and profiled to the same aim points of a color temperature of 6500K and a gamma of 2.2, there was a difference in perceived contrast that looked higher in addition to greater brightness of the of the LCD, but what was really surprising was after precise calibration and profiling the neutral gray of the two displays looked very different, with the Sony CRT appearing to be warmer and less neutral than the NEC LCD. In other words they measure the same but look different, which may be due to a difference in light frequency and modulation because of the distinctly different light sources involved. The bottom line is that not just brightness but every visually apparent factor will affect and influence perceptual image editing adjustment between the two kinds of display, and the resulting images will reflect some of that difference. I think in my case when I switched to LCD’s from CRT’s I was consciously aware a perceptual accommodation was needed and I altered my image editing accordingly, because my print output was pretty consistent. However, I do my own custom printer profiles and I did make some print density adjustments in the printer profiles I produced in the las 2-3 years I have been using LCD displays.
Although it would resolve the prints too dark problem, would I want to go back to CRT monitors if they were still made? Not on your life, because LCD’s provide a more neutral gray, greater internal light and color contrast to reveal tone differences and easier editing, and at least my LCD’s are much sharper than any CRT I had used in the past. So let’s hope for a print brightness fix and keep the LCD’s!
So what does this have to do with photography and any current digital problems like “prints too dark”? Back in the 70’s an intense interest in Ansel Adams Zone system was appended by Adam’s stress on “visualization” as a way to exercise some control between perception and a predictable photographic result. Around the same time Richard D. Zakia at RIT published a book of articles and scientific papers called Perception and Photography that was rather widely read providing a more modern understanding of human vision and how it can be more successfully employed as part of the process of making effective photographs and controlling the performance of the media. Thirty some year later perception has a somewhat different photographic role and even more important imaging function with photographs that are the result of what is done visually on-screen to edit and manipulate the attributes of a photograph in applications like Photoshop. Which brings us to the current problem of a mismatch in density or image brightness between what is seen on-screen and sometimes in a very different, too dark, print result.
That there is a problem today is due, in large part, to the introduction of very much brighter LCD displays compared to the CRT’s monitors that preceded them. What is perceived on screen does not match what results in a print in density, even though in a color managed workflow the colors seen on screen can be accurately reproduced in a print quite reliably. It should be noted here that digital camera JPEG images adjusted by a camera’s post exposure processing, if just opened in a computer and printed without editing, will very likely NOT suffer the “prints too dark” problem. So the only sure way the LCD brightness problem will result in print darkness is if the image is edited by PERCEPTUAL adjustment for brightness on-screen with an LCD and the file is then saved and printed.
Sometimes just reading the data and coming to conclusions on the basis of the fact an LCD display, even when optimally calibrated and profiled to a brightness luminance of 120.0 CD/m2, compared to the lower range of a typical CRT monitor with an average white point luminance of 90.0 CD/m2 (which did not induce “prints too dark”), is not totally convincing nor entirely complete evidence. So, fortunately for this research, I had a very good Sony Trinitron 520 Pro model 21” CRT stored for the last 3 years and after only a couple of years of use. And it just happens my Mac Pro will run two different kinds of displays in mirrored mode with both having the same signal image on-screen, as well as both being individually and independently calibrated and profiled. So after some digging in my closet I dragged out that 100 pound monster and set it up beside my NEC Multisync 2090 UXi LCD display. Both screens happen to be the same measured size, but once turned on and warmed up they do look very, very different and it is not just the NEC LCD display is a bit brighter. There really is not that much brightness difference as the Sony 520 CRT produces a maximum white point density on the high side at 97.0 CD/m2 compared to my LCD sitting next to it at 120.0 CD/m2.
Even though both were calibrated and profiled to the same aim points of a color temperature of 6500K and a gamma of 2.2, there was a difference in perceived contrast that looked higher in addition to greater brightness of the of the LCD, but what was really surprising was after precise calibration and profiling the neutral gray of the two displays looked very different, with the Sony CRT appearing to be warmer and less neutral than the NEC LCD. In other words they measure the same but look different, which may be due to a difference in light frequency and modulation because of the distinctly different light sources involved. The bottom line is that not just brightness but every visually apparent factor will affect and influence perceptual image editing adjustment between the two kinds of display, and the resulting images will reflect some of that difference. I think in my case when I switched to LCD’s from CRT’s I was consciously aware a perceptual accommodation was needed and I altered my image editing accordingly, because my print output was pretty consistent. However, I do my own custom printer profiles and I did make some print density adjustments in the printer profiles I produced in the las 2-3 years I have been using LCD displays.
Although it would resolve the prints too dark problem, would I want to go back to CRT monitors if they were still made? Not on your life, because LCD’s provide a more neutral gray, greater internal light and color contrast to reveal tone differences and easier editing, and at least my LCD’s are much sharper than any CRT I had used in the past. So let’s hope for a print brightness fix and keep the LCD’s!
Tuesday, December 23, 2008
THE INNOCENT MONKEY WRENCH
Not finished, nor ever expect to anytime soon, going back and spending a good part of the weekend reading “prints too dark” complaints and commentary on digital photography forums. It was no trouble finding plenty of examples posted on popular digital photography web sites. What was surprising was the diversity of situations described involving the problem of getting too dark prints, leading to a great variety of speculation as to what was causing the darkness of the prints produced, as well as just as wide an expanse of suggestions of why there is a problem and what fixes might be applied.
A small sample of suggested diagnosis indicated the problem does come from a display that is too bright. Another small group suggested the too dark print is caused by selecting the wrong profile for printing. Of those who identified an LCD display being too bright, a few suggested calibrating and profiling the display while many more suggested just reducing the adjustment of the brightness control. To the average digital photo enthusiast this diversity of the way the problem of prints too dark is experienced and described, as well as the assortment and sometimes weirdly fanciful fixes offered, surely does not indicate a clear and simple understanding of why prints are too dark or what to do about it.
There is a reason for this, and that is the most likely cause may not have any apparent association in time or action with the experience of the problem of dark prints, and there may be other problems involved beyond the primary cause, obscuring and complicating the problem; in other words a user is getting prints too dark and is having another issue with printing that is concurrent in time but functionally not connected to why the prints are dark. The other confusing aspect of this is that the cause of too dark prints in most cases only shows up when prints are made from particular image file sources and using particular printing options - sometimes.
Instead of making this “prints too dark” issue even more mysterious and perplexing let me explain. The primary cause of dark prints is usually a bright LCD display, not directly but indirectly and not printing all kinds of image files, just some and the only when certain methods or workflow options are chosen. In other words LCD displays are too bright, from a little brighter than a CRT monitor to a lot brighter. This fact does not directly cause prints to be output too dark as there is no direct link, or causal path between the display and printer. But LCD brightness does affect on-screen perception of images, and if a photo image for instance is opened in an image editing application and then adjusted for brightness and the file saved, the midpoint setting that results from that brightness adjustment is related to the scale of brightness values of the display. The range of values of an LCD display is greater than the range of corresponding density values in a print so the midpoint setting of the file adjusted for LCD screen brightness is “misplaced” to print a correct or equivalent print density. Think of it this way: take two straws, one longer than the other, then put a pen mark at the midpoint of each straw, equal distance from each end. Then put the straws on end, with one end on a table-top, and the straws side by side. You will see the midpoint marks on each straw will be at different heights from the table top - it is a mismatch of midpoints, as it is a mismatch of midpoints between an LCD display on-screen brightness adjustment of an image and what the midpoint needs to be to obtain a correct, perceptually equivalent, lightness adjustment of print output.
This mismatch of density midpoints never occurred with CRT monitors because the white point luminance was only around 90.0 CD/m2 so the range of brightness values of CRT on-screen images was about the same as the range of density values that can be reproduced by a photo printer. An ideally calibrated and profiled LCD display will have a white point luminance of 120.0 CD/2, which is 25% greater than the range of densities than can be reproduced by a print. While at default manufacturer settings LCD displays can produce white point luminance readings of as much as 300.0 CD/m2 and greater.
Some may say, I have an LCD display and my prints aren’t too dark. That may because they are printing digital camera JPEG files directly without any image editing, or it may be they are printing from files made using a CRT for color correction and editing stored on a CD, or files off the internet, even screen capture prints will reproduce with densities closely matched to what you see on-screen. But then if someone says they do color correct and edit their images on-screen, save the files and then print them and they are not too dark. Well, again they would not be too dark if printed using the “printer driver manages color” setting. Most photo capable printers if the easy, automatic print workflow is chosen, actually color correct, adjust and optimize the file information received automatically as part of the print processing; and if the midpoint is misplaced by using an LCD display for editing, the printer driver corrects for it. But that easy automatic “printer manages color” setting may not be ideal, especially if you want matched color with what is on screen or if the subject of the image is unusual, like a winter, snow covered scene. It is usually only when a photographer is using an application like Photoshop, LightRoom or Aperture that is color managed and selects to print with the application managing color, and printing to a specific printer output profile, that too dark prints result.
LCD displays have many advantages over the old CRT’s they replaced, but their brightness threw an innocent monkey wrench into the works of a color managed print workflow, and maybe a few other unlucky users making prints as well.
A small sample of suggested diagnosis indicated the problem does come from a display that is too bright. Another small group suggested the too dark print is caused by selecting the wrong profile for printing. Of those who identified an LCD display being too bright, a few suggested calibrating and profiling the display while many more suggested just reducing the adjustment of the brightness control. To the average digital photo enthusiast this diversity of the way the problem of prints too dark is experienced and described, as well as the assortment and sometimes weirdly fanciful fixes offered, surely does not indicate a clear and simple understanding of why prints are too dark or what to do about it.
There is a reason for this, and that is the most likely cause may not have any apparent association in time or action with the experience of the problem of dark prints, and there may be other problems involved beyond the primary cause, obscuring and complicating the problem; in other words a user is getting prints too dark and is having another issue with printing that is concurrent in time but functionally not connected to why the prints are dark. The other confusing aspect of this is that the cause of too dark prints in most cases only shows up when prints are made from particular image file sources and using particular printing options - sometimes.
Instead of making this “prints too dark” issue even more mysterious and perplexing let me explain. The primary cause of dark prints is usually a bright LCD display, not directly but indirectly and not printing all kinds of image files, just some and the only when certain methods or workflow options are chosen. In other words LCD displays are too bright, from a little brighter than a CRT monitor to a lot brighter. This fact does not directly cause prints to be output too dark as there is no direct link, or causal path between the display and printer. But LCD brightness does affect on-screen perception of images, and if a photo image for instance is opened in an image editing application and then adjusted for brightness and the file saved, the midpoint setting that results from that brightness adjustment is related to the scale of brightness values of the display. The range of values of an LCD display is greater than the range of corresponding density values in a print so the midpoint setting of the file adjusted for LCD screen brightness is “misplaced” to print a correct or equivalent print density. Think of it this way: take two straws, one longer than the other, then put a pen mark at the midpoint of each straw, equal distance from each end. Then put the straws on end, with one end on a table-top, and the straws side by side. You will see the midpoint marks on each straw will be at different heights from the table top - it is a mismatch of midpoints, as it is a mismatch of midpoints between an LCD display on-screen brightness adjustment of an image and what the midpoint needs to be to obtain a correct, perceptually equivalent, lightness adjustment of print output.
This mismatch of density midpoints never occurred with CRT monitors because the white point luminance was only around 90.0 CD/m2 so the range of brightness values of CRT on-screen images was about the same as the range of density values that can be reproduced by a photo printer. An ideally calibrated and profiled LCD display will have a white point luminance of 120.0 CD/2, which is 25% greater than the range of densities than can be reproduced by a print. While at default manufacturer settings LCD displays can produce white point luminance readings of as much as 300.0 CD/m2 and greater.
Some may say, I have an LCD display and my prints aren’t too dark. That may because they are printing digital camera JPEG files directly without any image editing, or it may be they are printing from files made using a CRT for color correction and editing stored on a CD, or files off the internet, even screen capture prints will reproduce with densities closely matched to what you see on-screen. But then if someone says they do color correct and edit their images on-screen, save the files and then print them and they are not too dark. Well, again they would not be too dark if printed using the “printer driver manages color” setting. Most photo capable printers if the easy, automatic print workflow is chosen, actually color correct, adjust and optimize the file information received automatically as part of the print processing; and if the midpoint is misplaced by using an LCD display for editing, the printer driver corrects for it. But that easy automatic “printer manages color” setting may not be ideal, especially if you want matched color with what is on screen or if the subject of the image is unusual, like a winter, snow covered scene. It is usually only when a photographer is using an application like Photoshop, LightRoom or Aperture that is color managed and selects to print with the application managing color, and printing to a specific printer output profile, that too dark prints result.
LCD displays have many advantages over the old CRT’s they replaced, but their brightness threw an innocent monkey wrench into the works of a color managed print workflow, and maybe a few other unlucky users making prints as well.
Friday, December 19, 2008
“PRINTS TOO DARK”, now I need your help!
First to bring you up to date. Since my workflow article ran in the December issue of Shutterbug, word has gotten around and back to me providing lots of information for a better picture of the problem. I have done more testing, which concluded even for a well color managed system like my own with LCD’s there is some darkening in print results because of color correcting and editing with an LCD with screen brightness set exactly at a luminance of 120.0 CD/m2. I confirmed this by opening some finished scanned image in Photoshop, files done when I had CRT monitors installed and then stored on CD’s that have been printed in the past. These image files look fine in terms of density on my LCD screen, and print as they did in the past achieving the same print density the screen appearance would suggest to expect. I have to assume the reason is that the image brightness midpoint setting in Levels made using a CRT monitor with a white point luminance of 90.0 CD/m2 (which was not changed for this current and test and printing), and the correct brightness of the print is because the brightness range of the CRT closely matched the density range of a high quality inkjet print.
To double check myself, I also opened a recent CD made since I began using LCD displays, and although not a big difference the prints from these images color corrected on an LCD screen, print a little darker. That they print darker confirms what many have complained about, but that my results are a modest darkening, is because my LCD is only 25% brighter than my CRT’s were, and that would only displace the midpoint setting by about half that, so the prints aren’t going to be dramatically darker. But if I were working with an LCD that is at manufacturer default brightness, as some users are, like with iMacs that have produced measured white luminance readings of 300.0 CD/m2, then the midpoint differential or displacement is much bigger, and prints would be consequently very disappointing in being much too dark.
Using some of these same recent LCD color corrected files that print 10-12% darker I identified, to confirm my contention in a recent blog that Soft Proofing does not predict print density, just color matching; I ran the image files through soft proofing and the screen result was a match in density with the original file on screen in PS and lighter than the print result.
This got me thinking, that I and many of you reading this using an Adobe application to print, once you get to the printer drive screen you have the option of saving the “print” you have set up as an Acrobat .PDF file. Being this .PDF file saved instead of making a print on paper has gone through the printer driver, shouldn’t the density of the resulting Acrobat image match the same difference, darker, if that’s what you would get in a paper print? What I am suggesting is that maybe rather than the standard internal Photoshop soft proofing, saving an image through the printer driver as an Acrobat .PDF file could be a better predictor of print output density, and allow photographers to preview print density and avoid wasting paper by making prints that are too dark.
My computer system for printing is rigorously color managed and the darkening of prints because of LCD brightness is slight, so the test Acrobat .PDF files I have produced seem to match paper print densities. But my dark print problem is minimal. So anyone who has a greater “prints too dark” problem, would you please try using the printer driver Save As PDF option and see if the .PDF file opened in Acrobat Reader matches a paper print density or the brightness of the original image file open on-screen in Photoshop, please, please! One note: to make an Acrobat .PDF file as a printer driver option, you need to set Photoshop Color Management in Print window Color Handling to “printer Manages Color”, and if the Source profile is Adobe RGB (1998) the Printer profile should be the same. If you do this experiment in using Acrobat. PDF to produce an image file viewed in Acrobat Reader, and it predicts print density then we can develop a workaround using it to get a paper print correction, and avoid wasting paper and ink on “too dark prints”. E-mail: goofotografx@gmail.com
To double check myself, I also opened a recent CD made since I began using LCD displays, and although not a big difference the prints from these images color corrected on an LCD screen, print a little darker. That they print darker confirms what many have complained about, but that my results are a modest darkening, is because my LCD is only 25% brighter than my CRT’s were, and that would only displace the midpoint setting by about half that, so the prints aren’t going to be dramatically darker. But if I were working with an LCD that is at manufacturer default brightness, as some users are, like with iMacs that have produced measured white luminance readings of 300.0 CD/m2, then the midpoint differential or displacement is much bigger, and prints would be consequently very disappointing in being much too dark.
Using some of these same recent LCD color corrected files that print 10-12% darker I identified, to confirm my contention in a recent blog that Soft Proofing does not predict print density, just color matching; I ran the image files through soft proofing and the screen result was a match in density with the original file on screen in PS and lighter than the print result.
This got me thinking, that I and many of you reading this using an Adobe application to print, once you get to the printer drive screen you have the option of saving the “print” you have set up as an Acrobat .PDF file. Being this .PDF file saved instead of making a print on paper has gone through the printer driver, shouldn’t the density of the resulting Acrobat image match the same difference, darker, if that’s what you would get in a paper print? What I am suggesting is that maybe rather than the standard internal Photoshop soft proofing, saving an image through the printer driver as an Acrobat .PDF file could be a better predictor of print output density, and allow photographers to preview print density and avoid wasting paper by making prints that are too dark.
My computer system for printing is rigorously color managed and the darkening of prints because of LCD brightness is slight, so the test Acrobat .PDF files I have produced seem to match paper print densities. But my dark print problem is minimal. So anyone who has a greater “prints too dark” problem, would you please try using the printer driver Save As PDF option and see if the .PDF file opened in Acrobat Reader matches a paper print density or the brightness of the original image file open on-screen in Photoshop, please, please! One note: to make an Acrobat .PDF file as a printer driver option, you need to set Photoshop Color Management in Print window Color Handling to “printer Manages Color”, and if the Source profile is Adobe RGB (1998) the Printer profile should be the same. If you do this experiment in using Acrobat. PDF to produce an image file viewed in Acrobat Reader, and it predicts print density then we can develop a workaround using it to get a paper print correction, and avoid wasting paper and ink on “too dark prints”. E-mail: goofotografx@gmail.com
Subscribe to:
Posts (Atom)