Luminous Landscape Forum
Raw & Post Processing, Printing => Colour Management => Topic started by: Doug Gray on March 31, 2018, 01:52:34 am
-
Having commented in the past about the fairly large differences printer manages v Photoshop manages for color reproduction I thought an experiment showing how a Colorchecker prints differently between the two processes. The attached images demonstrate the rather extreme lightening and increase in saturation that occurs on my Canon 9500 II just letting the printer print v properly color managed process.
I started with a Colorchecker CGATS color set from BabelColor. Created a CC image from them and converted it to sRGB. The conversion changes the cyan patch reducing saturation a bit but not any of the others.
Then I made a tiff image of the printed colors showing on the right side of each patch the starting CC color and the left side the color printed. One set is printed with Printer Manages using the defaults in the Windows driver, the other was made using Photoshop Manages with a custom profile and Perceptual Intent. The paper used was Canon Pro Platinum.
What is most startling is that Printer Manages pushes saturation (check the cyan patch). It also lightens most all the print colors a huge amount. Some of the colors are beyond sRGB even though the original was sRGB.
This is not a result of manufacturing tolerance. It's almost certainly intentional. Assuming Canon did research that shows people like casual snapshots printed this way what does this tell us? Perhaps this can be explained by CIECAM02? I would expect something like this to be done using CIECAM02 converting from a luminance of 500 Lux to say, 150 Lux.
See this link:
https://sites.google.com/site/clifframes/ciecam02plugin
Anyone here consider the use of CIECAM02 or tweak images to accomplish these large shifts?
First image is CC and Printer Manages, second image is CC and Perceptual with Photoshop Manages.
-
Not surprised, even though I can't see the images. You have them in TIFF format rather than JPEG. For whatever reason, either this website or Firefox won't open TIFF images in this Forum board. On the subject matter itself, I have reviewed the performance of Printer Manages Color (PMC) in one or the other of my printer and paper reviews (e.g. see Printing Can be Fun and Easy) and also noticed that there are saturation differences, but I was surprised by how close PMC can come to Application Managed printing. Methodologically orthodox quantitative evaluation is rendered difficult because one does not have Absolute Rendering Intent with PMC.
-
And difference via the OS?
-
Not surprised, even though I can't see the images. You have them in TIFF format rather than JPEG. For whatever reason, either this website or Firefox won't open TIFF images in this Forum board.
quite right, I know you are on MacOS, Firefox on Windows doesn't show them either. It would be useful for posters to render things to JPEGs so that we don't have to open the image in another application.
EDIT: In my case, clicking on the image title prompts me to either save the image or open it in Windows Photo Viewer which I guess is OK as I clearly saw the differences.
-
I see the images fine using Mac OS and Safari.
-
I guess we've sort of shanghaied this thread, but one can Google for "firefox tif viewer" while in W10 Firefox, first item is free download/plug-in via Mozilla.org for tif viewing.
Richard Southworth
-
Thanks for the feedback on what browsers work on these files.
I've recently shifted to 16 bit, zip compression, Interleaved RGB, IBM byte order, tiff files because they use up the smallest space at the same time as being able to have high resolution (dEs from digitization of < .05). This really is only necessary for high precision targets which I sometimes need for certain purposes. Displays on Windows IE. At least versions from Win 7 or later. Tiff files with zip compression are known to be less compatible but they take up much less disk space.
Tiff's are not needed for this image (or most) so I have changed them and placed them on a single jpeg. Unlike the 16 bit tiffs, the Lab values at the bottom of the patches will only match colors to 1 or 2 dE but the very large color shifts are way beyond that.
It's fascinating that a good Perceptual print results in a slight darkening. This is completely normal because Perceptual scales paper white to L*==100 so the printed white patch (L*=96) will be lower. But to brighten the other patch colors so much requires very high compression on the highlights.
Compare the 6 neutral patches on the bottom rows to see how extreme this tone curve change is.
-
And difference via the OS?
That's a good question Andrew. I suspect OEMs would use the same driver defaults with the same model and different OSes for Printer Manages Color but don't know for sure. I do know there is variation between printer OEMs because my Epson 9800 does this too, but it is more subtle.
I am quite curious as to how this has evolved and what the OEMs have chosen. If anyone wishes, I can provide image targets of CC colors for scanning with my iSis and will provide a mailing address for those willing to print them. I will post the results. It would be interesting to see how the different OEMs have made their choices and how it's evolved.
-
I guess we've sort of shanghaied this thread, but one can Google for "firefox tif viewer" while in W10 Firefox, first item is free download/plug-in via Mozilla.org for tif viewing.
Richard Southworth
Thanks Richard
-
I guess we've sort of shanghaied this thread, but one can Google for "firefox tif viewer" while in W10 Firefox, first item is free download/plug-in via Mozilla.org for tif viewing.
Richard Southworth
+1 to what Mark just wrote. Downloaded the plug in and the Tif images are right there now!!
-
Yes same here.
-
Here's what I'm seeing on Mac OS, Epson 3880.
Printed my Gamut Test File (http://www.digitaldog.net/files/Gamut_Test_File_Flat.tif (http://www.digitaldog.net/files/Gamut_Test_File_Flat.tif)) using both Application Manages Color (Photoshop) and Printer Manages Color.
Printer Manages Color is pretty butt ugly! Over saturated and lots of banding in Granger Rainbow etc. I picked RelCol and Auto using Epson Luster paper.
There's no comparison between the two; the print without proper color management is pretty awful in comparison. I could make some scans but not sure it's worthwhile. Both synthetic imagery (Bills Balls, gradients) and actual images from raw to ProPhoto are dark and contrasty and over saturated. I can't imagine anyone preferring this output.
Friends don't let friends use Printer Manages Color with such data. Maybe converting to sRGB first (which I'd never do and would hose much of the nice saturated colors) may be better.
If you folks really want to see scans or a photo of each, let me know or download and make your own prints with this file.
-
Here's what I'm seeing on Mac OS, Epson 3880.
Printed my Gamut Test File (http://www.digitaldog.net/files/Gamut_Test_File_Flat.tif (http://www.digitaldog.net/files/Gamut_Test_File_Flat.tif)) using both Application Manages Color (Photoshop) and Printer Manages Color.
Printer Manages Color is pretty butt ugly! Over saturated and lots of banding in Granger Rainbow etc. I picked RelCol and Auto using Epson Luster paper.
There's no comparison between the two; the print without proper color management is pretty awful in comparison. I could make some scans but not sure it's worthwhile. Both synthetic imagery (Bills Balls, gradients) and actual images from raw to ProPhoto are dark and contrasty and over saturated. I can't imagine anyone preferring this output.
What you see doesn't surprise me given the colors in your test image. It's really good for stress testing custom or canned profiles using color management. No telling what might result with printer management. Windows will only print sRGB images using the printer driver defaults. Like you, I'm not sure it's worthwhile converting it to sRGB then printing. The Colorchecker only has one patch outside sRGB. Almost everyone has one so it seemed like a reasonable candidate to look at what stuff the OEMs were up to.
Friends don't let friends use Printer Manages Color with such data. Maybe converting to sRGB first (which I'd never do and would hose much of the nice saturated colors) may be better.
If you folks really want to see scans or a photo of each, let me know or download and make your own prints with this file.
Yeah, no kidding. What I'm trying to wrap my head around is why OEMs would make such changes so hyper saturated, and strongly lightened tone curve? Do they have studies that show some preference for this? I'm pretty sure their studies show most people that aren't doing their own color management are in sRGB or some, probably unmanaged approximation of sRGB. But that doesn't explain the saturation and tone curve changes. However, at first blush it is at least somewhat consistent with the alterations made by CIECAM02 with reduced viewing luminance. But if so their studies must indicate some large preference for lower luminous environments and that they are targeting those for the plug and play crowd.
-
Is there a reason why you chose to use default printer settings with "Printer Manages Color"? Isn't this kind of hobbling the results for a bias toward "Photoshop Manages Colors"?
I don't own an expensive printer. I have to use "Printer Manages Color" on my $50 Epson "All In One" printing Epson Ultra Premium Glossy 8x10's, but I do have to use specific driver settings (not defaults) that renders a close enough match on a wide range of photographed scenes.
But you've proven your point for high end printers costing over $100 that "Photoshop Manages Color" is the way to go. I'm sure this thread will help a lot of folks who use this level of quality printers.
-
Is there a reason why you chose to use default printer settings with "Printer Manages Color"? Isn't this kind of hobbling the results for a bias toward "Photoshop Manages Colors"?
I don't own an expensive printer. I have to use "Printer Manages Color" on my $50 Epson "All In One" printing Epson Ultra Premium Glossy 8x10's, but I do have to use specific driver settings (not defaults) that renders a close enough match on a wide range of photographed scenes.
What I'm trying to understand is what the printer OEMs think the part of their market, probably pretty sizable, that doesn't use color management considers "good." One presumes the defaults are set to what most like since otherwise it's a sort of commercial suicide. It differs considerably from properly color managed processes and I'd like to know why. The only explanation that makes any sense at all, other than gross incompetence, is that they are somehow applying a CAM (Color Appearance Model) to the images.
-
I don't own an expensive printer. I have to use "Printer Manages Color" on my $50 Epson "All In One" printing Epson Ultra Premium Glossy 8x10's, but I do have to use specific driver settings (not defaults) that renders a close enough match on a wide range of photographed scenes.
Good to know, should a future discussion about color management (or the lack thereof) and printing come up here or elsewhere.
Further, your idea that you have to use Printer Manages Color makes no sense considering I also own an Epson All In One (XP-860) and as you may be able to see, I can and do select custom ICC profiles within Photoshop and other applications using that printer and can utilize Application Manages Color!
-
What I'm trying to understand is what the printer OEMs think the part of their market, probably pretty sizable, that doesn't use color management considers "good."
Yes, 'good enough' color. Like close enough. Perfect for the hobbyists (http://forum.luminous-landscape.com/index.php?topic=123799.msg1036108#msg1036108) here ;D
-
What I'm trying to understand is what the printer OEMs think the part of their market, probably pretty sizable, that doesn't use color management considers "good." One presumes the defaults are set to what most like since otherwise it's a sort of commercial suicide. It differs considerably from properly color managed processes and I'd like to know why. The only explanation that makes any sense at all, other than gross incompetence, is that they are somehow applying a CAM (Color Appearance Model) to the images.
Have you found this CAM takes into account the other printer setting options that do change the results as I described above with the Epson "All In One"?
All this is you just speculating on what a bunch of Canon engineers or marketers think what color should look like. Maybe you're misjudging the market these printers were aimed at especially considering we're now living in a time of content overload where the value of an image isn't worth the trouble you've outlined.
IOW nobody cares except technologists who expect split hair precision printing targets from a device that's not meant to be used as a precise tool. Close enough is close enough and move on. People have over thousands of images to cull through to print, frame and hang on their walls. They'll probably spend 3 seconds looking at them and not notice how off some of the color is to what? The original image? Why would they want to make those comparisons?
-
IOW nobody cares except technologists who expect split hair precision printing targets from a device that's not meant to be used as a precise tool.
Speaking for all others or speaking for all hobbyists?
People have over thousands of images to cull through to print, frame and hang on their walls.
People and professionals often differ. A good definition of a professional: you can't pay or convince them to produce anything but the highest quality possible.
Don't just learn the tricks of the trade, learn the trade. That's another difference between pro's and amateurs. ;)
A professional is someone who can do his best work when he doesn't feel like it.
-Alistair Cooke
-
Before anyone gets carried away with conclusions about the relative merits or demerits of Printer Manages Color (PMC), (which BTW is not a workflow I would use for my printing), please consider that not all versions of it are born equal. Epson engaged some high-end scientific talent to design it for the SC-P5000 printer and the results that I obtained using this printer would seem to justify that effort (please see my Printing Can Be Fun and Easy article). The sample on the right is PMC, the one on the left is Application managed, custom ICC profile. The PMC version is a bit over-saturated, and the gray ramp is a bit less neutral looking and a bit heavier in the quartertones than it should be; but grosso modo it is a result not to be scoffed at - as I said, not for my portfolio, but for many it would be satisfactory. It may perhaps be relevant to recall here that I'm on OSX, so I had and used the option to work it in ARGB(98).
-
Yes, 'good enough' color. Like close enough.
I think it goes beyond "good enough." The OEMs understand color management. If it were "good enough" there would be clustering around a proper color management result and, from what I've seen so far, there is a consistent skewing towards increased saturation and some degree of image lightening.
I've been reading Fairchild's "Color Appearance Models," 3rd edition and it goes into considerable detail about appearance changes due to factors such as surround including a lot of things CAMs don't currently account for such as the cognitive effect of "illuminant discounting" which occurs when viewing something tangible like a print but does not viewing a monitor. Fairchild has published other studies on this as well. My reading of Fairchild's text resonates with what I've seen. So it seems to me the color shifts OEMs use in their defaults are intentional and that each one tries to come up with whatever recipe pleases the largest numbers in some obscure way. Their secret, if ugly to us, sauce.
-
Before anyone gets carried away with conclusions about the relative merits or demerits of Printer Manages Color (PMC), (which BTW is not a workflow I would use for my printing), please consider that not all versions of it are born equal. Epson engaged some high-end scientific talent to design it for the SC-P5000 printer and the results that I obtained using this printer would seem to justify that effort (please see my Printing Can Be Fun and Easy article).
Don't have such a printer Mark, curious how my Gamut Test File appears to you. I wouldn’t be surprised that our mileage may vary depending on the driver which kind of makes PMC iffy.
-
So it seems to me the color shifts OEMs use in their defaults are intentional and that each one tries to come up with whatever recipe pleases the largest numbers in some obscure way. Their secret, if ugly to us, sauce.
Like a Perceptual Rendering intent. Or transparency film.
-
Just tried a similar test to the one previous but used my Printer Test File (http://www.digitaldog.net/files/2014PrinterTestFileFlat.tif.zip (http://www.digitaldog.net/files/2014PrinterTestFileFlat.tif.zip)) which is in Adobe RGB (1998) rather than ProPhoto RGB. Not as awful but pretty bad!
Here's a quick and dirty shot of the Gamut Test File from my first test, under the GTI booth, shot on an iPhone X (quick and dirty). Just awful. Maybe Mark I'll try this on a much newer printer and thus print driver, a P800. But based on two tests today, you'd have to put a gun to my head to use Printer Manages Color with Adobe RGB (1998) or ProPhoto RGB image data on a 3880.
-
Don't have such a printer Mark, curious how my Gamut Test File appears to you. I wouldn’t be surprised that our mileage may vary depending on the driver which kind of makes PMC iffy.
That could be an interesting sample to prepare Andrew, but right at this moment I'm heavily preoccupied with a pretty complicated piece of work so I can't put the time into it just now, but time permitting I may do so, in which case I'll of course report back. But yes, that was the point I'm making - results could vary depending on the printer model, the driver and the operating system, so there's no way one could assure anything like the inter-device output consistency that is a hallmark or an ICC colour managed workflow.
-
That could be an interesting sample to prepare Andrew, but right at this moment I'm heavily preoccupied with a pretty complicated piece of work so I can't put the time into it just now, but time permitting I may do so, in which case I'll of course report back. But yes, that was the point I'm making - results could vary depending on the printer model, the driver and the operating system, so there's no way one could assure anything like the inter-device output consistency that is a hallmark or an ICC colour managed workflow.
Not sure it's worth the ink and paper, let alone time. I'll play with the P800 maybe tomorrow. Maybe just one print. But IF indeed this is a crap shoot, based on the driver and age of printer etc, and other than for hobbyists only working in sRGB where close enough and good enough is their goal, I don't see the point for most of the audience here venturing down this rabbit hole. Even an OK canned ICC profile will produce better results than I'm seeing today. They can soft proof, they can pick a rendering intent. And they get better output.
-
I think it goes beyond "good enough." The OEMs understand color management. If it were "good enough" there would be clustering around a proper color management result and, from what I've seen so far, there is a consistent skewing towards increased saturation and some degree of image lightening.
I've been reading Fairchild's "Color Appearance Models," 3rd edition and it goes into considerable detail about appearance changes due to factors such as surround including a lot of things CAMs don't currently account for such as the cognitive effect of "illuminant discounting" which occurs when viewing something tangible like a print but does not viewing a monitor. Fairchild has published other studies on this as well. My reading of Fairchild's text resonates with what I've seen. So it seems to me the color shifts OEMs use in their defaults are intentional and that each one tries to come up with whatever recipe pleases the largest numbers in some obscure way. Their secret, if ugly to us, sauce.
I do agree with Mark's statement about different printers of the same brand not delivering equal results with Printer Manages Color which seems to be the case with more expensive printers.
What doesn't make sense and wasn't raised in this discussion is how a printer driver on Mac OS X can hand off ProPhotoRGB data from Photoshop and Mac OS Preview app with Printer Manages Color and get as close as it does with a cheap Epson "All In One".
I demonstrated it here on this 2013 Photo.net thread on this very subject where I posted my Epson settings.
https://www.photo.net/discuss/threads/how-to-do-color-managed-printing-to-epson-printers-that-dont-have-paper-profiles.478458/#post-5652198
Five years later I get a poster thanking me for those settings so it must've worked for him and I'm having to assume he wasn't using a cheap printer, but now I'm speculating.
-
Just tried a similar test to the one previous but used my Printer Test File (http://www.digitaldog.net/files/2014PrinterTestFileFlat.tif.zip (http://www.digitaldog.net/files/2014PrinterTestFileFlat.tif.zip)) which is in Adobe RGB (1998) rather than ProPhoto RGB. Not as awful but pretty bad!
Here's a quick and dirty shot of the Gamut Test File from my first test, under the GTI booth, shot on an iPhone X (quick and dirty). Just awful. Maybe Mark I'll try this on a much newer printer and thus print driver, a P800. But based on two tests today, you'd have to put a gun to my head to use Printer Manages Color with Adobe RGB (1998) or ProPhoto RGB image data on a 3880.
OK, any time you start messing with "Bill's Balls" you're asking for trouble! :-)
The P800 outcome could be better - I don't know.
-
I do agree with Mark's statement about different printers of the same brand not delivering equal results with Printer Manages Color which seems to be the case with more expensive printers.
That opinion doesn't jive with this statement:
I don't own an expensive printer. I have to use "Printer Manages Color" on my $50 Epson "All In One" printing Epson Ultra Premium Glossy 8x10's, but I do have to use specific driver settings (not defaults) that renders a close enough match on a wide range of photographed scenes.
but now I'm speculating.
Remove the word now. The statement is now close enough to accurate. :o
-
OK, any time you start messing with "Bill's Balls" you're asking for trouble! :-)
He's got some big balls to mess with! ::) And they were not designed for printing but they do show some interesting results when used through differing profiles and print paths. The Granger Rainbow is I believe Kosher for printing and also show some interesting results. But in the tests I did today, the print area with actual images look pretty awful without an ICC profile print path.
-
Not sure it's worth the ink and paper, let alone time. I'll play with the P800 maybe tomorrow. Maybe just one print. But IF indeed this is a crap shoot, based on the driver and age of printer etc, and other than for hobbyists only working in sRGB where close enough and good enough is their goal, I don't see the point for most of the audience here venturing down this rabbit hole. Even an OK canned ICC profile will produce better results than I'm seeing today. They can soft proof, they can pick a rendering intent. And they get better output.
Well, yes. This raises of course the logic of who uses what workflow with which equipment. I have a hard time imaging that any one buying a P800 or a P5000 are the kind of people who would be totally happy with Printer Manages Color no matter how improved it has become, especially if they are on Windows OS. But for people buying inexpensive non-professional printers, which can nonetheless make very "nice" prints, PMC can be their option of choice because they may be casual printers, may not want to be bothered going down the learning curve of Application managed printing, and indeed may not even be using Photoshop or Lr again because of the learning curve. Even people using a P800 or P5000 may occasionally want to quickly zip out something "kind of acceptable", so PMC is there as a convenience. A mainstay - no way. Consider, however, the fact that Epson put a considerable effort into improving PMC tells me that their knowledge of market justifies it - recall they see the whole customer base in great detail - we don't.
-
I need to clarify in case their is some confusion. I think printing Photos I've worked on in Photoshop using Printer Manages Colors produces bad results. To me.
I want to understand what and why the OEMs are doing this so I've been characterizing the printers I have and shaking my head. I suppose if random cell phone snaps in sRGB are just printed w/o color management it's entirely possible that many, or most people would find the images appealing and perhaps better than properly color managed printing using only perceptual intent and no other adjustments. So I propose that IECCAM02 is likely one of the culprits. Anyone have any experience with it? I linked to a plugin for it earlier.
We are not the typical printers. We have refined processes, utilize soft proofing, recognize some of the differences between a monitor and an image in a different environment and make prints based on that experience. Most users do not.
-
Well, yes. This raises of course the logic of who uses what workflow with which equipment. I have a hard time imaging that any one buying a P800 or a P5000 are the kind of people who would be totally happy with Printer Manages Color no matter how improved it has become, especially if they are on Windows OS. But for people buying inexpensive non-professional printers, which can nonetheless make very "nice" prints, PMC can be their option of choice because they may be casual printers, may not want to be bothered going down the learning curve of Application managed printing, and indeed may not even be using Photoshop or Lr again because of the learning curve. Even people using a P800 or P5000 may occasionally want to quickly zip out something "kind of acceptable", so PMC is there as a convenience. A mainstay - no way. Consider, however, the fact that Epson put a considerable effort into improving PMC tells me that their knowledge of market justifies it - recall they see the whole customer base in great detail - we don't.
Let's make sure we specify here that PMC can stand for Photoshop Manages Color AND Printer Manages Color. I had to correct myself referring to this acronym in my previous posts.
If the user's primary goal is to serve the paper (craft papers with matte and satin finishes for interior decorating purposes) over optimizing the dynamics of the image (requiring glossy or less absorbent paper) custom ICC profiles and Photoshop Managing Color would be required.
For me Printer Manages Color works best and only best on glossy paper made by the printer manufacturer in my case Epson Ultra Premium Glossy. Matte papers (including Epson's) don't really serve the image's dynamics. IT'S NOT GOOD!
At this point for me all I care about are archival 8x10 prints that serve the image's dynamics with the least amount of time and cost invested. $50 All In One is fast enough but I don't believe inexpensive per print considering the cost of ink. And after about ten full color 8x10's on Ultra Premium Glossy, my cheap Epson printer requires head alignment and cleaning. And pizza wheel marks are an issue which I've given up on remedying.
At least I have a physical hard copy of my images at a reasonable viewing size.
-
I suppose if random cell phone snaps in sRGB are just printed w/o color management it's entirely possible that many, or most people would find the images appealing and perhaps better than properly color managed printing using only perceptual intent and no other adjustments.
Well I believe sRGB is the expected route for this print mode due to Dave P. of Adobe stating on Windows everything gets funneled into sRGB as I believe we've discussed.
Anyway, I converted my Gamut Test File to sRGB then made a print that's the same as my first test with ProPhoto RGB. In the ColorSync area of PMC, I set it to Epson Color Controls, then Color Mode to No Color Management as that's just one option a hobbyist might select but shouldn’t. It looks pretty bad too but not as bad as using ProPhoto RGB.
Next I set the Epson Color Controls to Epson Standard (sRGB). That print looks decent but not anywhere as good as the full ICC Profile path from ProPhoto RGB.
So there are several possible settings using PMC and two are awful, one's OK (close enough?).
Best of the lot can't compare to the ICC path especially when viewing gradients and Bills Balls; lots of banding there, lack of shape in the balls, far less color saturation.
-
Let's make sure we specify here that PMC can stand for Photoshop Manages Color AND Printer Manages Color.
You're confused again. We (some here) are not. That's again an issue when you speak for anyone other than yourself here.
There are two modes:
Printer Manages Color.
Application Manages Color.
Applications have names. They may show up, they may not. But when so set, they manage the color!
Go into Lightroom assuming you own it. You will see "Managed By Printer". And no Tim, there's no LMC for Lightroom, or PMC for Preview, or AMC for Acrobat. Those are names of applications.
I had to correct myself referring to this acronym in my previous posts.
Par for the course! Because PMC isn't an acronym for Photoshop Manages Color even if that application starts with the letter P!
Assumptions allow the best in life to pass you by. -John Sales
-
For me Printer Manages Color works best and only best on glossy paper made by the printer manufacturer in my case Epson Ultra Premium Glossy. Matte papers (including Epson's) don't really serve the image's dynamics. IT'S NOT GOOD!
That opinion doesn't jive with how you've told us you are forced to print (you have no options if we are to believe your first text):
I don't own an expensive printer. I have to use "Printer Manages Color" on my $50 Epson "All In One" printing Epson Ultra Premium Glossy 8x10's, but I do have to use specific driver settings (not defaults) that renders a close enough match on a wide range of photographed scenes.
Please make up your minds as to how something you have no option over works better than something you state you have no option to use. Or not; probably a better tactic in these kinds of technical discussions of color and imaging....
-
I have reviewed the performance of Printer Manages Color (PMC)...
Before anyone gets carried away with conclusions about the relative merits or demerits of Printer Manages Color (PMC)....
Even when correctly spelled out clearly, some find it foggy. I've edited your text above with formatting to focus lurkers (and a non lurker) that the correct acronym was indeed provided. Long before the confused arrived to post here (the second post!). :P
-
Well I believe sRGB is the expected route for this print mode due to Dave P. of Adobe stating on Windows everything gets funneled into sRGB as I believe we've discussed.
And, as I've pointed out to Dave P., it doesn't since I can correctly print an image in ProPhoto RGB, using a custom profile and any Intent using only Printer Manages Color from my Epson 9800. However, it's also clearly dependent on the printer driver code since I cannot do that on my Canon 9500 II. It's not a Windows thing but the printer driver code and that's written by/for the OEMs.
Anyway, I converted my Gamut Test File to sRGB then made a print that's the same as my first test with ProPhoto RGB. In the ColorSync area of PMC, I set it to Epson Color Controls, then Color Mode to No Color Management as that's just one option a hobbyist might select but shouldn’t. It looks pretty bad too but not as bad as using ProPhoto RGB.
Next I set the Epson Color Controls to Epson Standard (sRGB). That print looks decent but not anywhere as good as the full ICC Profile path from ProPhoto RGB.
That makes sense.
So there are several possible settings using PMC and two are awful, one's OK (close enough?).
Best of the lot can't compare to the ICC path especially when viewing gradients and Bills Balls; lots of banding there, lack of shape in the balls, far less color saturation.
Agree to the extent I can replicate.
-
And, as I've pointed out to Dave P., it doesn't since I can correctly print an image in ProPhoto RGB, using a custom profile and any Intent using only Printer Manages Color from my Epson 9800.
Yes, I know there's some controversy here. So you're picking that profile rather than Automatic (if an option) on Windows? Seems the Auto (default) on Mac isn't very good. Any chance you have an Epson to try the same test?
-
Yes, I know there's some controversy here. So you're picking that profile rather than Automatic (if an option) on Windows? Seems the Auto (default) on Mac isn't very good. Any chance you have an Epson to try the same test?
That was the Epson 9800 that allows printing using ICC profiles. To print a ProPhoto image you have to select the ProPhoto profile as source in the driver, then select the printer profile and intent also in the driver. At least on my Win 10 x64. Can't tell any difference between a print made that way and the usual Photoshop managing color. It prints all the way to the printer's gamut edge. Option is not available on the Canon 9500 II. ICC can be selected but no way to select a printer profile or input RGB colorspace.
-
That was the Epson 9800 that allows printing using ICC profiles. To print a ProPhoto image you have to select the ProPhoto profile as source in the driver, then select the printer profile and intent also in the driver. At least on my Win 10 x64.
If it isn't a hassle, love to see a screen capture as the bit about selecting ProPhoto as the source isn't AFAIK an option on Mac. Which could be one of the issues.
Can't tell any difference between a print made that way and the usual Photoshop managing color. It prints all the way to the printer's gamut edge
Wild! That seems different than what is happening on the Mac. But as at least three of us posting here know, over the years, printing with Epson drivers/Photoshop on the Mac has been often buggy and problematic.
-
If it isn't a hassle, love to see a screen capture as the bit about selecting ProPhoto as the source isn't AFAIK an option on Mac. Which could be one of the issues. Wild! That seems different than what is happening on the Mac. But as at least three of us posting here know, over the years, printing with Epson drivers/Photoshop on the Mac has been often buggy and problematic.
Here you go:
I selected the wrong profile for the Epson (a canon 9500 profile). I meant to select a 9800 one.
-
Here you go:
I selected the wrong profile for the Epson (a canon 9500 profile). I meant to select a 9800 one.
Thanks! Nothing like that on my Mac.
-
What I'm trying to understand is what the printer OEMs think the part of their market, probably pretty sizable, that doesn't use color management considers "good." One presumes the defaults are set to what most like since otherwise it's a sort of commercial suicide. It differs considerably from properly color managed processes and I'd like to know why. The only explanation that makes any sense at all, other than gross incompetence, is that they are somehow applying a CAM (Color Appearance Model) to the images.
If what the TV manufactures do is anything to go by, then accuracy is not high on their list. Being eye-catching, attention grabbing, eyeball-popping, hyper-real, dramatically contrasting and retina searing is likely to be closer to the mark! The first thing anyone after accuracy does with a modern TV is turn off all the "show room mode" gimmicks. The second thing would be to calibrate it to meet Rec.709 and similar relevant standards.
Judging by the type of example prints used to demonstrate even high end large format printers, I don't think the art of selling your printer to the average consumer is any different. And I rather doubt that a color appearance model has much to do with it either - my impression is that the traditional mode of operation amongst Asian printer manufacturers is extensive hand tuning to make it "look good", although there have been a series of technical reports over the years on methods to tweak images to "optimize" the appeal of memory colors etc., so it's certainly possible that this type of thing has been applied as well. Setting up a printing system to be technically accurate seems to be a rather foreign concept, except perhaps as a base for optimizing a prints appeal.
(But to be fair, is this much different to the development of photographic film reproduction ? The whole process wasn't optimized for accuracy, but for reproducing images the way people like to remember them, by increasing contrast and saturation etc.)
-
Speaking for all others or speaking for all hobbyists? People and professionals often differ. A good definition of a professional: you can't pay or convince them to produce anything but the highest quality possible.
Don't just learn the tricks of the trade, learn the trade. That's another difference between pro's and amateurs. ;)
A professional is someone who can do his best work when he doesn't feel like it.
-Alistair Cooke
Please, an amateur is someone well versed in the art who chooses not to make a living using that knowledge. I am an amateur of photography and proud of it.
-
Please, an amateur is someone well versed in the art who chooses not to make a living using that knowledge. I am an amateur of photography and proud of it.
Some absolutely are and strive for as equal quality/knowledge. Some (one) by example earlier; not so much if at all by admission (a lack of being well versed/knowledgeable).
-
Some absolutely are and strive for as equal quality/knowledge. Some (one) by example earlier; not so much if at all by admission (a lack of being well versed/knowledgeable).
Agreed. 8)
-
Tested the P800 today. As a side note, haven’t used it in many months, did a nozzle check, all clean!
First, using PMC with what *appears* to be a default in Mac when selected is equally awful as the 3880. That is, setting the ColorSync Radio button to be on and profile set to Automatic. Maybe this automatic selection is broken and maybe I need to manually select "other profile" but the Epson installed profiles don't show up in the list. I could select a custom profile. None the less, this setting doesn't work!
What is interesting is what the Epson driver tells the user in Advanced Color Settings as seen below. The data provided to the user is updated in the newer driver (outlined in red).
Next I set the radio button from ColorSync to EPSON Color Controls. In the main printer dialog (Basic) there are two possible options for Color Mode: EPSON Standard (sRGB) or Adobe RGB (1998). I picked Adobe RGB (1998). The print is MUCH better! Acceptable without any of the ugliness I provided in my shots off the iPhone. How does it compare using a custom ICC profile and a full color managed path? Not as good and not close (enough) or otherwise for me. First, the ICC profile path provides greater saturation and better tonal distribution. Somewhat expected from ProPhoto RGB source data. There's more banding in PMC seen on gradients and Bill's Balls! Skin tones are better from the ICC Profile (man in boat); too magenta on PMC. The dark but saturated colors in the cloth image is easier to see (again, better tonal range or what one might call DR) than PMC. But PMC when properly set here doesn't suck at all! But for anyone who wants the best possible print quality, I'd be hard pressed to recommend PMC over a good ICC profile. Something most of us here who've actually tested this should know. But overall, worth half a dozen piece of paper to come to a conclusion void of assumptions. ;)
-
A comparison study of test print snapshots from Andrew against a commercial Epson Surecolor. Comments are noted in the image attachment.
-
That's pretty awful! What "commercial Epson Surecolor" did that an what were the settings?
-
Wonder how images that aren't cranked to 11 in the saturation department print. None of those test images I would want hanging on my wall, but I do see that printing is a performance no matter how ugly the image.
I don't understand why I don't get those garish colors and crappy color transitions on my $50 Epson "All In One".
-
I don't understand why I don't get those garish colors and crappy color transitions on my $50 Epson "All In One".
So much to not understand.
-
That's pretty awful! What "commercial Epson Surecolor" did that an what were the settings?
I think this iPhone photo I made a couple years ago from a test of an Epson "S Series" machine (from an array of 6) about 1 year old at the time using only 4/C S Ultrachrome GS2 ink (solvent.) This series is now also available as a "photo" setup with more inks.
Is it understood that the two upper panels in my posted image upload is Andrew's examples, and the bottom panel is my example of the commercial machine?
-
I think this iPhone photo I made a couple years ago from a test of an Epson "S Series" machine (from an array of 6) about 1 year old at the time using only 4/C S Ultrachrome GS2 ink (solvent.) This series is now also available as a "photo" setup with more inks.
Is it understood that the two upper panels in my posted image upload is Andrew's examples, and the bottom panel is my example of the commercial machine?
Ah, I misunderstood.
Looks like the full set of "Bill's Balls" on the bottom. It turns out that Bill made the ball images intending them to be used in "printer space," not Lab or ProPhoto, as a way to visually check the smoothness of the printer driver. They should be printed with color management off just like one prints a profile target. Abrupt banding that shows up in device space can't be fixed by profiling and rapid, but not abrupt, changes at any point over the radius indicates that the profile may need a higher number of patches to track these colors.
When printed in device driver RGB space they are, by definition, in gamut. So a good printer design won't produce sharp banding on any ball.
However, they can and will look different on different printers but that's what profiles correct quite well.
-
Wonder how images that aren't cranked to 11 in the saturation department print. None of those test images I would want hanging on my wall, but I do see that printing is a performance no matter how ugly the image.
I don't understand why I don't get those garish colors and crappy color transitions on my $50 Epson "All In One".
What do tests of the 2 files look like from your $50 Epson "All In One"?
-
Looks like the full set of "Bill's Balls" on the bottom. It turns out that Bill made the ball images intending them to be used in "printer space," not Lab or ProPhoto, as a way to visually check the smoothness of the printer driver.
Yes, and that's what you see in the labeled "v No color mgmt but calibrated v" panel and under that is a single row of "color managed" balls. In this machine's environment, "no color mgmt" is otherwise known as "All Profiles OFF" in the RIP. The machine requires upstream RIP settings however to control ink restrictions and at least some sort of calibration curve. From looking at Andrew's first panel, it seems to me his machine is completely out of sorts when using whatever settings used at the time of his test. It certainly doesn't not appear to be PMC.
-
Yes, and that's what you see in the labeled "v No color mgmt but calibrated v" panel and under that is a single row of "color managed" balls. In this machine's environment, "no color mgmt" is otherwise known as "All Profiles OFF" in the RIP. The machine requires upstream RIP settings however to control ink restrictions and at least some sort of calibration curve. From looking at Andrew's first panel, it seems to me his machine is completely out of sorts when using whatever settings used at the time of his test. It certainly doesn't not appear to be PMC.
Yours look pretty reasonable. At least as far as one can see from the snapshot. They vary quite a lot printer to printer. Good idea to check them with a soft proof. For images in device space, Photoshop->View Proof-> and select preserve numbers and the printer profile. Should match what's printed.
Andrew's test file is in ProPhoto and is designed to stress profiles. It's good at evaluating profiles handling of highly saturated colors, many of which are at the ProPhoto extremes. Images often have colors that are outside what can be printed so handling these smoothly (pleasingly) can be useful. I use several images for checking profiles. Andrew's stress test and the classic Kodak PhotoDisc is the opposite of Andrew's stress test having few colors that push printers but a nice selection of faces and odd's and ends. The third one is in Lab space I synthesized in Matlab. It's a set of circles with L* at 5 unit intervals and a*, b* coordinates that go from neutral in the center to about 120 in saturation. It's identified some rather odd anomalies in I1P Perceptual Intent tables. I do a quick soft proof on that but don't often print them. When I do it's to check smoothness for in-gamut colors.
-
What do tests of the 2 files look like from your $50 Epson "All In One"?
Oh, I'm sure they'll look like shit with Printer Manages Color.
I'm having to assume of course because I have no interest in printing garish colors in images like Bill's Balls or the over cranked green boat image I wouldn't hang on a wall. Also I no longer have a working Epson "All In One". I now get my prints on a Fuji Frontier at Walmart.
It's clear advertising and graphics pro's need to use custom profiles and Photoshop Manages Color to insure Bill's Balls smooth gradient transitions in electric eyeball popping color, but that's not the market intended for a $50 printer. You want to turn printing into a horse race that's fine. You win. I concede.
How about do a comparison print of a picture of an animal or landscape with realistic looking colors?
-
Oh, I'm sure they'll look like shit with Printer Manages Color.
Wrong again; can’t you read? They don’t look too bad when PMC (figured out what that means?) setting is one way, but not another. But not anywhere as good as AMC with a good ICC profile.
You should not be sure of anything. That I am thus far sure of based on your comments here.
I'm having to assume of course
A nasty habit.
-
Oh, I'm sure they'll look like shit with Printer Manages Color.
If you've only imagined it, you haven’t' experienced it. I however have experienced it.
Below is a close up of some of Bills Balls, printed using PMC on the all in one, Epson XP-860 on Luster paper. Doesn't look like crap as seen when using the WRONG PMC settings (default) on Mac. And you can't make that mistake on this all in one. There's only one set of settings (No ColorSync Radio button, no Automatic, no Epson sRGB etc).
Once again we see that Tim, with his undefined All In One isn't testing anything before assuming and posting here. Sad! But consistent. He tells us he can't using ACM on such a printer (I've proven that wrong). He assumes Epson all in one would print awful from the Gamut Test File he hasn't taken the time to download or print (sad again).
It's clear advertising and graphics pro's need to use custom profiles and Photoshop Manages Color to insure Bill's Balls smooth gradient transitions in electric eyeball popping color, but that's not the market intended for a $50 printer.
Well at least one advertising and graphic pro has proven that wrong! Another quick and dirty iPhone shot under the GTI booth. Apples to Apples view of the various output.
I never assume. It is a shocking habit -- destructive to the logical faculty. -Sir Arthur Conan Doyle
-
What do tests of the 2 files look like from your $50 Epson "All In One"?
An excellent question that will not be answered (or answered incorrectly) from one poster, but you can see on my Epson all-in-one, not bad! As seen below, these are the only settings for this Epson when using PMC; simple and it works pretty well. One can of course use a custom profile if desired, despite what another person has told us.
-
I never assume. It is a shocking habit -- destructive to the logical faculty. -Sir Arthur Conan Doyle
As ever, there's an xkcd for that.
Jeremy
-
OK Jeremy, that does it - this has now gone far beyond my paygrade :-), so never too old to learn something "new", I just had to look it up: XKCD (http://www.explainxkcd.com/wiki/index.php/Main_Page); never too late to learn in my old age, and it ain't even colour management, thank goodness.
-
Wow, they even have one that comes close to colour management, wouldn't ya know:
Visual Field (http://www.explainxkcd.com/wiki/index.php/File:visual_field.png)
-
As ever, there's an xkcd for that.
Jeremy
I'm assuming you think this is good moderating, right?
-
Get a life Tim - may be refreshing to know that moderators can multi-task. There comes a point when it's nice to take a break from counting how many angels can dance on the proverbial head of a pin.
-
I'm assuming you think this is good moderating, right?
Assuming seems to be a professional obsession of yours. Why not ask? Yes, it's a fine post from Jeremy compared to those that are composed of misinformation and misunderstanding, parceled off to attempt an impression of understanding the topic of color and printing.
Like Mark, I didn't know what xkcd was and I learned something new (and correct) from his post. You should try that tactic someday..... :P
-
There comes a point when it's nice to take a break from counting how many angels can dance on the proverbial head of a pin.
And so you think ridiculing someone who participates as you do in counting these angels dancing on the head of pin is a an honorable thing?
Get a life you say, Mark? How about you and everyone here come up with something I can laugh at. You all suck at humor. Why is that?
-
I can laugh at myself Tim, not laughing at anyone else. We all need to "take a pill" once in a while!
I think that's the moderator's message - let's not take ourselves so seriously as we often have a tendency of doing. Would avoid a lot of "interpersonal issues" all round.
-
I think that's the moderator's message - let's not take ourselves so seriously as we often have a tendency of doing.
Seems the one person taking himself very seriously (but few others are taking him seriously) is.... No need to name names here. ;D
BTW, there also comes a point when it's nice to take a break from counting how many angels can dance on the proverbial head of an ICC profile.
-
How about you and everyone here come up with something I can laugh at.
Lots of options here:
https://www.wayfair.com/decor-pillows/sb0/mirrors-c416716.html?refid=GX88041546988.Mirror%7Ee&position=1t1&network=g&pcrid=88041546988&device=c&targetid=aud-42686225422kwd-15049040&channel=GoogleIntent&gclid=Cj0KCQjwqYfWBRDPARIsABjQRYxT6GF2AeuKaPWMKPyDMS4QwJEgvsOOczlI_3y_DGCrC5sUjPOqWF8aApZrEALw_wcB
-
BTW, there also comes a point when it's nice to take a break from counting how many angels can dance on the proverbial head of an ICC profile.
Ah, you haven't seen anything yet - just wait.. :-) (Hors d'oeuvre: the number is over 4600; the main course will be served anon).
-
OK Jeremy, that does it - this has now gone far beyond my paygrade :-), so never too old to learn something "new", I just had to look it up: XKCD (http://www.explainxkcd.com/wiki/index.php/Main_Page); never too late to learn in my old age, and it ain't even colour management, thank goodness.
The first xkcd cartoon I ever saw was this one, which comes close. The author has written a couple of books, too. "What if?" is very amusing.
Jeremy
-
I'm assuming you think this is good moderating, right?
It's taking part, not moderating. I'm still a member of the forum. There's no conflict.
Jeremy
-
The first xkcd cartoon I ever saw was this one, which comes close. The author has written a couple of books, too. "What if?" is very amusing.
Jeremy
That's a riot! Thanks.
-
You all suck at humor. Why is that?
I think probably you mean that our senses of humour don't match yours, just as in another thread it's clear that our concept of what constitutes good art doesn't match yours.
De gustibus non est disputandum.
Jeremy
-
One never needs their humor as much a when they argue with a fool.
-Chinese Proverb
-
I think probably you mean that our senses of humour don't match yours, just as in another thread it's clear that our concept of what constitutes good art doesn't match yours.
De gustibus non est disputandum.
Jeremy
Do your moderator's duty and lock this thread. OP question has be asked and answered. Everything is now spiraling out of control!!!!!
-
Do your moderator's duty and lock this thread. OP question has be asked and answered. Everything is now spiraling out of control!!!!!
I was hoping to get some informed comments about why manufacturers are doing this and, specifically, if it can be ascribed to Color Appearance Models, which suggest increased saturation and a brightening of the tone curve will produce a printer image that appears, after a time to adapt, to be closest to what was on the computer monitor.
There seems to be agreement that Printer manages generally changes these, possibly because it produces more pleasing prints, but it seems to me this is exactly what Perceptual Intent is supposed to do but it's changes are much smaller in my experience.
-
Do your moderator's duty and lock this thread. OP question has be asked and answered. Everything is now spiraling out of control!!!!!
I absolutely disagree! Locking threads (deleting them) should be a last case scenario; someone yelling fire in a crowed theater. I'm not sure it is even a moderators duty to lock a thread at least without asking the OP (Doug in this case) who can lock it anytime he wants. IF you find this is out of order, you can stop receiving notifications. Further, speaking for the OP and suggesting his question is answered has little merit IMHO. Plus, we should give Tim an opportunity at the very least to answer the question asked by Stephen in post #55 (http://forum.luminous-landscape.com/index.php?topic=124034.msg1036363#msg1036363). Doubt that will happen but give him the opportunity. More critically, there is large potential for future posts on topic and based on sound testing and understanding of this topic from others.
IF you like forums where posts are locked and deleted, LuLa isn't your place; the DP Review forums are a perfect location for this kind of censorship!
“Censorship is telling a man he can't have a steak just because a baby can't chew it.” ― Mark Twain
-
I don't see any reason to lock it either. It's gotten a bit testy at times, but there have been much worse. The subject matter, if we can revert to it, actually is quite interesting. We *may* have hit the end of the road, however, in terms of additional useful information to enlighten the subject because there is a whole lot about how and to what end these algorithms were designed that the manufacturers aren't about to share with us, notwithstanding one thing I do know about one of them - high-end scientific talent went into it. But this isn't a reason to lock the thread either. For all we know it could still attract some knowledgeable contributors and contributions, so I suggest just leaving it open.
-
Mark,
This is why I find it an intriguing topic. I've pretty much gone as far as possible with icc profiles and color management as have many others here. But there remains the somewhat little discussed issue that the prints we make, unadjusted for viewing differently than in a hard proof setup that closely matches out monitors, is perhaps not the ideal for some prints that will not be viewed in such a manner. It's clear that CIECAM02 is an attempt, though less than ideal, to accommodate this.
I really like the repeatability across similar media and different printers that ICC profiles provide, but would also like to see what can be done using a controlled CIECAM02 model to produce prints optimal for other lighting conditions. I suspect some of this has been done by OEMs when creating their "printer manages" mapping. And OEMs have considerable resources and smart color scientists at their disposal.
Having such a thing would bring the repeatability of ICC profiles together with the presumed "enhancement" that OEMs appear to think is better in some sense. This would provide techniques/tools that professionals could use to decouple the OEM secret sauce and allow them to more easily tailor their own works while achieving the consistency needed when switching or upgrading printers.
BTW, much of this line of thought is an outgrowth of things like Ctien's comments as well as from reading a text I discovered from one of Andrew's posts I read somewhere I've forgotten. Fairchild's "Color Appearance Models, 33rd edition - Wiley"
It's a most fascinating text and I've found it elucidates many of the observations made such as the fact that a white piece of paper on one's desk does not look at all like exactly that same "piece of paper" colorimetrically identical on a monitor. It's an experiment worth doing just to demonstrate how huge a change in appearance occurs when one "knows" a piece of paper is white. I had previously commented about this phenomena and Fairchild spells it out See pages 172-174 in the text.
-
It's clear that CIECAM02 is an attempt, though less than ideal, to accommodate this.
Yes, absolutely as current ICC color management isn't based on such complex color appearance. But what is? That's kind of the $64K question isn't it. I doubt the printer manufactures will tell us if truthfully. Just examine the post today about LR 7.3 and some marketing nonsense (IMHO) about their new profiles and what we are told they can produce:
http://forum.luminous-landscape.com/index.php?topic=124095.msg1036824#msg1036824
What we can do is what you've started here; make some tests and show the results. They don't appear so great to me! I haven't seen anything PMC on three Epson printers provided when ink hits the paper that compares to using an ICC profile workflow. Old fashion technology indeed. But the proof's are in the prints.
-
I haven't seen anything PMC on three Epson printers provided when ink hits the paper that compares to using an ICC profile workflow. Old fashion technology indeed. But the proof's are in the prints.
Generally I agree, but the gap is narrowing as the drivers improve, viz what I got out of the Epson SC-P5000 using the Atkinson test page. It was surprisingly similar, though still not quite as good as the AMC-ICC result, where saturation, colour balance and tonal gradations remain somewhat better.
-
Generally I agree, but the gap is narrowing as the drivers improve, viz what I got out of the Epson SC-P5000 using the Atkinson test page. It was surprisingly similar, though still not quite as good as the AMC-ICC result, where saturation, colour balance and tonal gradations remain somewhat better.
I think we're saying the same thing, even though we're looking at quite different printers: PMC ain't there yet when compared to an ICC profile.
Then there's the inability to control rendering intent, soft proof etc. PMC will never be there compared to an ICC profile workflow.
Better would be ICC profiles that take advantage of the color appearance modeling Doug speaks of.
It's good to hear the newer Epson is better than my P800/3880/All-in-one but I'm not trading them in for a newer printer because PMC there is better.
-
I think so.
-
There seems to be agreement that Printer manages generally changes these, possibly because it produces more pleasing prints, but it seems to me this is exactly what Perceptual Intent is supposed to do but it's changes are much smaller in my experience.
I'm not sure the sort of changes some people might apply to an image to make it attractive ("make it pop") fall into the typical expectation of "perceptual" or even "saturation". "Artistic" might be closer to the mark.
But I tend to see it as a clash of traditions - the ICC tradition is founded on measurements and color accuracy, with perceptual/saturation mappings based on cLUTs and the idea of gamut mapping, while the other tradition is based on hand tuned curves and other algorithmic or manual tweaks made on the basis of visual assessment or per channel type measurements such as density, with the explicit goal of making images "look good".
-
I'm not sure the sort of changes some people might apply to an image to make it attractive ("make it pop") fall into the typical expectation of "perceptual" or even "saturation". "Artistic" might be closer to the mark.
But I tend to see it as a clash of traditions - the ICC tradition is founded on measurements and color accuracy, with perceptual/saturation mappings based on cLUTs and the idea of gamut mapping, while the other tradition is based on hand tuned curves and other algorithmic or manual tweaks made on the basis of visual assessment or per channel type measurements such as density, with the explicit goal of making images "look good".
I quite agree! There is a difference between IECCAM02 and tweaking colors to make them "look good." The former's purpose is to make the image appearance look as close as possible to that on the monitor. The latter goes a step beyond.
My guess is that "looking good" wins in the marketing contest but it is subjective as can be and, frankly, I don't find printer manages colors to be all that good looking. OTOH, the devices I have are getting long in the tooth.
-
I quite agree! There is a difference between IECCAM02 and tweaking colors to make them "look good." The former's purpose is to make the image appearance look as close as possible to that on the monitor. The latter goes a step beyond.
My guess is that "looking good" wins in the marketing contest but it is subjective as can be and, frankly, I don't find printer manages colors to be all that good looking. OTOH, the devices I have are getting long in the tooth.
Whatever "secret sauce" is being discussed here could be made into a series of .icc profiles or maybe LUTs that could be applied to an image just before printing to get a similar result, or variety of results.
The goal here would be an "accurate" print, but maybe the "appearance" of accuracy :) or maybe something else entirely...
-
Whatever "secret sauce" is being discussed here could be made into a series of .icc profiles or maybe LUTs that could be applied to an image just before printing to get a similar result, or variety of results.
Or any of the massive numbers of sliders or editing tools in products used to edit images. Output specific edits!
-
I was hoping to get some informed comments about why manufacturers are doing this and, specifically, if it can be ascribed to Color Appearance Models, which suggest increased saturation and a brightening of the tone curve will produce a printer image that appears, after a time to adapt, to be closest to what was on the computer monitor.
In this instance, my informed comment is coming from past experience. Sometimes a product can be touted as “better, new, improved” on the trade show floor by trickery. Years ago certain machines were sold as “sharper than the competition” only because the default sharpening setting was always ON at it’s setting of zero and to turn it off one would need to make the setting of negative 3. Another trick was to mislead a prospect by using “lpi” as a specification of resolution when really the specification was “lines per image” and not the usual “lines per inch.” Too many tricks to count.
So, if a printer engineer / manufacture can create some way to impress a buyer, investor, accountant, a superior; that could explain a lot and it may not actually be better, new, or improved!
My contacts at Kodak, Fuji, and Epson have long retired and would not necessarily have any more technical information than we already have at this point. However, I’m walking distance to the Canon Experience Center showroom in Costa Mesa, CA., so if anyone can provide the name of contact, I would be happy to take the time to see them and ask questions. I know Canon has at least more than a couple of very savvy reps at trade shows whom I’ve met and could have insight , I just don’t have any of their names.
-
Do your moderator's duty and lock this thread. OP question has be asked and answered. Everything is now spiraling out of control!!!!!
Clearly our perceptions of the scope and extent of a "moderator's duty" differ. Even if it were so in this case (and the continuation of the thread would suggest that it is not), a thread will not be locked merely because I consider it to be complete, still less because anyone instructs me to do so.
The thread's originator has the ability to lock it if he considers that further contributions are not useful.
Jeremy
-
I really like the repeatability across similar media and different printers that ICC profiles provide, but would also like to see what can be done using a controlled CIECAM02 model to produce prints optimal for other lighting conditions.
ArgyllCMS gamut mapping (perceptual/saturation) uses CIECAM02 by default. Plenty of parameters to play with of you want :-)
-
Clearly our perceptions of the scope and extent of a "moderator's duty" differ. Even if it were so in this case (and the continuation of the thread would suggest that it is not), a thread will not be locked merely because I consider it to be complete, still less because anyone instructs me to do so.
The thread's originator has the ability to lock it if he considers that further contributions are not useful.
Jeremy
Yes, the perceptions are certainly different. I was a moderator on the Coffee Corner for a period of time when we were discussing climate change. I tried my best to keep things on track and several times sent private messages or even public ones to those who were veering off track. You might do well to use the same tools when you see personal attacks of any nature on threads as has happened on this thread. Civility is key to any on line discussion and it is the moderator's job to insure that takes place. Posting of cartoons in jest is no way to accomplish this goal.
Now a response might be to just ignore posts that one finds offensive but then one has to spend time separating the wheat from the chaff. Do you expect that it is the job of those who read the forum to calll out 'snarky' posts? There are individuals who often contribute valuable information but then have marked lapses in judgement. I'm just tired of seeing this type of behavior tolerated as it verges on bullying.
-
Yes, the perceptions are certainly different. I was a moderator on the Coffee Corner for a period of time ....
How long?
-
Yes, the perceptions are certainly different. I was a moderator on the Coffee Corner for a period of time when we were discussing climate change. I tried my best to keep things on track and several times sent private messages or even public ones to those who were veering off track. You might do well to use the same tools when you see personal attacks of any nature on threads as has happened on this thread. Civility is key to any on line discussion and it is the moderator's job to insure that takes place. Posting of cartoons in jest is no way to accomplish this goal.
Now a response might be to just ignore posts that one finds offensive but then one has to spend time separating the wheat from the chaff. Do you expect that it is the job of those who read the forum to calll out 'snarky' posts? There are individuals who often contribute valuable information but then have marked lapses in judgement. I'm just tired of seeing this type of behavior tolerated as it verges on bullying.
I am most grateful for your comments and your advice, which I shall treat with due respect.
Jeremy
-
In this instance, my informed comment is coming from past experience. Sometimes a product can be touted as “better, new, improved” on the trade show floor by trickery. Years ago certain machines were sold as “sharper than the competition” only because the default sharpening setting was always ON at it’s setting of zero and to turn it off one would need to make the setting of negative 3. Another trick was to mislead a prospect by using “lpi” as a specification of resolution when really the specification was “lines per image” and not the usual “lines per inch.” Too many tricks to count.
So, if a printer engineer / manufacture can create some way to impress a buyer, investor, accountant, a superior; that could explain a lot and it may not actually be better, new, or improved!
My contacts at Kodak, Fuji, and Epson have long retired and would not necessarily have any more technical information than we already have at this point. However, I’m walking distance to the Canon Experience Center showroom in Costa Mesa, CA., so if anyone can provide the name of contact, I would be happy to take the time to see them and ask questions. I know Canon has at least more than a couple of very savvy reps at trade shows whom I’ve met and could have insight , I just don’t have any of their names.
Hopefully those savvy Canon reps don't talk about the trickery part.
I would think the only valuable info as Doug asks concerning "looking good" printer algorithms would have to come from Canon engineers and scientists. From your experience have you found printer reps can understand and convey this level of complexity to make it understandable even to them self?
In my experience what I've found to be the most unforgiving stress tester of a printer is the photographer's taste in color. That's why we provide Soft Proofing so the technology can remind them they need to find another hobby.
-
In my experience what I've found to be the most unforgiving stress tester of a printer is the photographer's taste in color. That's why we provide Soft Proofing so the technology can remind them they need to find another hobby.
How do you soft proof with your undefined/unnamed all-in-one Epson (hobbyist) printer that you tell us cannot support AMC? That's Application Manages Color to help you along.
-
Hopefully those savvy Canon reps don't talk about the trickery part.
Some reps just have not yet learned as much poor practice.
From your experience have you found printer reps can understand and convey this level of complexity to make it understandable even to them self?
The reps I've spoken with, yes. Well qualified they are.
In my experience what I've found to be the most unforgiving stress tester of a printer is the photographer's taste in color. That's why we provide Soft Proofing so the technology can remind them they need to find another hobby.
Then again there are "artists."
-
In my experience what I've found to be the most unforgiving stress tester of a printer is the photographer's taste in color. That's why we provide Soft Proofing so the technology can remind them they need to find another hobby.
I would be interested to hear more about you technical experience with stress-testing printers. What methodologies have you used to do this work? Have you deployed objective testing procedures or have you witnessed how photographers' taste in colour defies their printers, and if the latter, which photographers may I ask? Or are you speaking for yourself stress testing your All-in-One with prints from Costco? I'd just like to get a handle on what kind of experience we're discussing here.
As for Canon product reps, scientists and engineers, I don't know what line of work you are in, but I'd hazard a guess that I have talked with more of them more often than you will probably encounter in many years and have found them to be highly professional; and as they should, they know more about this business than probably most of put together here.
-
I would be interested to hear more about you technical experience with stress-testing printers. What methodologies have you used to do this work?
Have you deployed objective testing procedures or have you witnessed how photographers' taste in colour defies their printers, and if the latter, which photographers may I ask?
Or are you speaking for yourself stress testing your All-in-One with prints from Costco?
Good questions! Along with those asked in post #55.
When someone says, That's a good question. You can be sure it's a lot better than the answer you're going to get. -Source Unknown
-
I would be interested to hear more about you technical experience with stress-testing printers.
The many photographic inkjet prints on a number of paper and metal surfaces I've inspected at my local art gallery tells me this technology is similar to handing scissors to children and telling them to go play on the freeway of color fidelity. Lots of clipped color detail in the name of "art". I doubt any of them had ever heard of the term "Soft Proofing", but from the garish colors I also doubt even if they did employ it, it wouldn't have made an improvement.
My "Printer Manages Color" prints off my Epson "All In One" and those printed from Walmart in sRGB were much better and didn't use Soft Proofing. But those were of scenes that didn't require cranking up the saturation on warm colors to depict for example a western scene of a wagon train with a team of horses at sunset. The most valued and expensive component of these gallery prints I saw was the amount of ink laid out.
So much for the "pin head dance" on this subject.
-
"Inspected"? OK zero relevant experience. And disparaging the whole subject by dismissing it as a "pin head dance" doesn't change that, or the view of others here with significant experience that the factors being discussed mean something and have practical consequences.
-
"Inspected"? OK zero relevant experience. And disparaging the whole subject by dismissing it as a "pin head dance" doesn't change that, or the view of others here with significant experience that the factors being discussed mean something and have practical consequences.
If you’ve only imagined it, you haven’t experienced it; yeah, virtually no experience from someone with a habit of posting incorrect concepts. Evidence to ignore!
-
............ Evidence to ignore!
What "evidence" is there to ignore? None of those observations at his local art gallery or from his Epson All-In-One, which I do not doubt Tim observed as he reported, have a bearing on the technical aspects of the discussion. Doug Gray started this thread making some technical observations that deserve to be explored and discussed at that level.
-
What "evidence" is there to ignore?
Evidence that someone here, with no experience in testing this subject, on an imaginary all-in-one undefined printer, should be ignored. ;)
None of those observations at his local art gallery or from his Epson All-In-One, which I do not doubt Tim observed as he reported, have a bearing on the technical aspects of the discussion.
We are in violent agreement, hence his so called observations (to be so kind) should be ignored.
-
I'm not sure the sort of changes some people might apply to an image to make it attractive ("make it pop") fall into the typical expectation of "perceptual" or even "saturation". "Artistic" might be closer to the mark.
But I tend to see it as a clash of traditions - the ICC tradition is founded on measurements and color accuracy, with perceptual/saturation mappings based on cLUTs and the idea of gamut mapping, while the other tradition is based on hand tuned curves and other algorithmic or manual tweaks made on the basis of visual assessment or per channel type measurements such as density, with the explicit goal of making images "look good".
Graeme,
ICC tradition is made for printing eg. magazines where a whole batch of images should be gamut-mappped consistently. In fine art printing it is reasonable to use adaptive gamut mapping. Klaus Karcher wrote a droplet which implements this with Argyll.
Edmund