Luminous Landscape Forum
Raw & Post Processing, Printing => Colour Management => Topic started by: crwoo on February 01, 2015, 10:36:52 pm
-
Hi, I have a Colormunki and I don't think my colors on my printer are as good as they could be so I am wondering If it would be an upgrade if I git the older X-Rite i1 Pro Photo (UV Cut) to replace it? I know it is an old device, but it uses more colors, right? the new version 2 is ways over my budget.
I am running Yosemite and the printers are a Canon iPf8400, Epson 3880, and 3800.
-
I use the uv-cutting older model with my 8300 and a 3880, and I'm very happy with the results. Have never used the ColorMunki.
If you think something is wrong with your ColorMunki profiles, download the manufacturer's profiles for the same papers you have made profiles for. Compare the two with the gamut volume graphs on iccview.de. If the shapes look very different, then something may be wrong.
You would need to compare several different profiles to definitely make a conclusion. I get extremely good agreement with the manufacturer profiles from Museo and Canson, very slight differences with Hahnemuhle profiles, and rather large differences with Breathing Color profiles, comparing those to profiles I create with my iPF8300 and my I1Pro. That may have to do with the profiling systems those manufacturers use, and also with the condition of my high mileage printer.
Edit: have only used the old i1Pro on Win 7. Works with the current verion of the PC version of i1Profiler.
-
Since I'm generally cheap ;) I think you should probably try either using more patches via ArgyllCMS software and/or getting a custom profile or two made from someone who has/uses an i1Pro. Argyll is free and custom profiles are roughly $25+ each. Would you pay $25 to find out if you should invest in an i1Pro? What country/continent are you located in?
-
Bill, thanks for the info, thats kind of what i want to hear.
Howard, im un USA. I will surly try the free software, but before i start paying for software i will probably buy the older i1 and decide which to keep.
-
what papers (particularly for the 3880) are you interested in re-profiling? Also, there is some form of optimization step in the Munki software but I dont know how useful it actually is.
-
the project i'm currently working on is sublimation. the colormunki imprives the overall colors but I have made 3 new profiles and they are all different. attached is a picture of the nicest one, and if you look at the bottom of the bird's belly there are harsh blue blotches. some pics are a lot worse. perhaps if i would enhance the profile with a few sheets it would improve.
-
If you are using sublimation inks I assume that you are transferring the profiling targets to the final substrate to do the measurements? Different substrates may well need different profiles.
It's not easy to tell how "wrong" your profile is from the image you have posted, but it looks to me as if there are some serious problems. Have you printed one of the "standard" test images to get a better overall evaluation?
-
Hi, the way the colormunki works is it prints 50 patches and after you scan them it prints a page of another 50 patches. I sublimated those on alum and scanned them. there is an option to improve the profile with more and more patches, which i have not done yet. I am wondering if i would be better off getting the old i1 or continue this with the colormunki.
I have the EFI express that came with the 3880 Designer, but have not installed it yet, and according to EFI website it works with the i1, is that true?
-
The ColorMunki is a good device, I can't see how going back to an older Spectrophotometer would help and think it might be the software rather than the hardware. And that might be an issue due to how the Munki requires a two pass system to operate. If you can't get software support from the measured data, good as the instrument is, you might have to find another solution.
-
have you uploaded the profile (and maybe an image or two) to iccview.de and looked at the profile shape/volume etc etc?
-
no, i have not. truth is i dont know squat about profiles and up till now i have depended on the colormunki's automated process that worked ok. in my case now the colors are so far out that it doesnt work with just a few clicks.
I'm a photographer and own Photoshop cs6 and i have been hearing thats how i should create my profiles. can anyone link me to instructions or training sites to modify existing profiles?
thanks guys, hope I'm not sounding too random but i am kind of confused and am taking advice from a few sources. :)
-
no, i have not. truth is i dont know squat about profiles and up till now i have depended on the colormunki's automated process that worked ok. in my case now the colors are so far out that it doesnt work with just a few clicks.
I'm a photographer and own Photoshop cs6 and i have been hearing thats how i should create my profiles. can anyone link me to instructions or training sites to modify existing profiles?
thanks guys, hope I'm not sounding too random but i am kind of confused and am taking advice from a few sources. :)
I appreciate you dont know 'squat' but you're also not doing that much to help us help you. Not that I'm a wizard but there are people here who are.
As an engineer of many years, here's what's wrong w/ your 'plan'...
You propose to buy a new widget when you dont have a any hard evidence that it's a shortcoming of the device.
You propose to edit a profile (which at best should be left to others) that you dont know is any good or not in the first place.
I stand by my original statement. Upload, view and take a screenshot and paste it here of the profile with an example image. Use Argyll to create a 1 page target and sublim. onto your substrate and try to make a profile.
-
Hi, i am working on making a profile with agryll and will post it
-
It might help if you posted one of your Colormunki profiles so that we can have a look at it.
-
Attached is the profile. Colors are MUCH improved vs the before, but it has some very harsh colors.
Here is what I am thinking. I use the enhanced matte on my 3880 and that uses WAYS too much ink, I have set the ink load all the ways down to -50 and that still uses too much. maybe that messes with the Colormunki patches?
How would I change the ink load on a profile with photoshop and save it?
-
the only way (w/o a RIP) is to pick a different base Media Type. I believe glossy/lustre uses less ink load.
So this profile you uploaded is made from Argyll software or the CM software?
-
That's one sick little profile. It's got a big, inverted sinkhole where the darkest of the dark colors should be. Looks like you have backed your black ink down to nothing, or you are missing one of the blacks.
What's your nozzle check printout like?
Enhanced Matte is a very ink-hungry paper. That's part of the reason it looks so nice for the price.
-
hmm interesting... what paper do you suppose uses least amount of ink besides plain paper? plain paper max dpi is 320. I have the same black in the matte and gloss black tanks so switching to gloss is possible.
-
Here is what I am thinking. I use the enhanced matte on my 3880 and that uses WAYS too much ink, I have set the ink load all the ways down to -50 and that still uses too much. maybe that messes with the Colormunki patches?
Clarification ... you using the enhanced matte setting, but is this for printing on dye Transfer paper? Or are you printing on the enhanced matte paper?
Calibrating for dye sub is more challenging than regular photo papers because of all the variables. To make a target you need to print the target image on the transfer paper. Do the transfer using the exact same temperature and exact same length of time. If the temperature or time vary at all it will affect the density of the transfer and thus the results/repeatability of using that profile. After letting the final image cool completely to room temperature you should then be able to read it ... you may even want to wait a day.
As to which media setting/how much ink you should use, the ink manufacturers should have good advice on that. Hard to say with dye sub because you can’t really judge the amount by looking at the printed media, and you usually need to print the image with extra density to get a correct density of the final transfer.
-
hi, obviously their customer service isnt as good as it should be otherwise i wouldnt try to figure it out myself.
i need to figure out what paper uses low ink load because when i lowered my ink load most of the colors got a lot better. some still arent nice and im assuming if i lower it more it will help. maybe luster would use less ink.
-
my temperature is stable on my press and my heat time is as close as i can...within a second i suppose.
I do realize that this is more complicated than regular papers. any tips are welcome.
-
I believe some of the bamboo and rice inkjet papers have rather sparing ink usage and modest profiles.
From the looks of that profile, the range of colors available on that media is restricted to colors brighter than medium bright. Your best approach may be to adjust your post processing to realize the image in only rather bright tones, and forget the dark ones.
The default printer dialogue for my 8300 offers the ability to adjust individual color levels in the "Main" tab, where you can find Color Settings->Color Adjustment. Playing with those sliders will defeat the significance of the profile, but might give you better prints. You could do testing on any cheap paper to get into the ballpark, before trying your problem media.
-
I made another profile with using a paper type that uses much less inks, do you see anything weird with it? on my test samples the colors are faded ...especially the greens. seems like the yellows are blotching a bit.
I still have the ink load all the ways to -50,
If there is nothing weird about the profile I will try to improve it with more patches and maybe bump up the ink load.
Thanks
-
Much better profile! At least it's not reversing on itself. It's still very odd to see the maximum density occuring so far from the center of the Luminance axis, but I can imagine this would make a better print, although with a blue bias in the dark areas.
-
Here's why the greens and yellows are not coming through. The wireframe is for Epson Cold Press Natural, which is itself of moderate gamut. I doubt if you will be able to get those greens and yellows much better on your media when using a profile.
You can try printing unprofiled and without color management which may give you stronger saturation in many colors, but at a loss of tonality. You'd need to try a lot of test prints.
-
Is there an English version of iccview.de / there's no translated version alternative on the page and the links I got from google return a Nt Found error.
-
Here's a tutorial I just made of how to use Colormunki with Argyll. That's a cheap way of "upgrading" the instrument, as the instrument is much more capable than the bundled software makes it.
http://www.ludd.ltu.se/~torger/photography/argyll-print.html
The main weaknesses of the Colormunki instrument seems to be that 1) it nees larger patches than i1Pro so you need to waste more ink/paper for the same patch count 2) it's more tedious to scan as you have no ruler and it's better to scan a bit slower as the instrument is a bit more noisy and gain from averaging and 3) it's UV cut so you cannot do OBA correction
The value of OBA correction can be debated though (as you lock down your print for a specific UV content in the viewing light and viewing light may not be that constant in real conditions), and if you use OBA-free paper like Hahnemühle Baryta for example it makes no difference at all of course.
If you make lots of profiling with many different papers I think it can be worthwhile to upgrade to get a more efficient workflow. If you're not interested in OBA correction and only profile occasionally I think "upgrading" by using Argyll (which is free) is a good choice. Although I haven't been able to make an A/B comparison between the instruments I have no reason to believe that a Colormunki+Argyll profile would be noticeably less accurate than a profile made with i1Pro and bundled software, unless we talk about OBA correction which the Colormunki can't do.
From tests I've made using Argyll instead of the bundled software will make measurably more accurate profiles, even if you just scan a single A4 sheet, but I think four sheets (840 patches) is about the ideal to get "as good as it gets" without overkill when using a modern inkjet. The bundled software may have say DE2000 3.7 as max and 1.2 as average, while a 840 patch argyll profile will have say 1.8 max and 0.5 average.
I'm keeping my Colormunki.
-
Here's a tutorial I just made of how to use Colormunki with Argyll. That's a cheap way of "upgrading" the instrument, as the instrument is much more capable than the bundled software makes it.
http://www.ludd.ltu.se/~torger/photography/argyll-print.html
The main weaknesses of the Colormunki instrument seems to be that 1) it nees larger patches than i1Pro so you need to waste more ink/paper for the same patch count 2) it's more tedious to scan as you have no ruler and it's better to scan a bit slower as the instrument is a bit more noisy and gain from averaging and 3) it's UV cut so you cannot do OBA correction
The value of OBA correction can be debated though (as you lock down your print for a specific UV content in the viewing light and viewing light may not be that constant in real conditions), and if you use OBA-free paper like Hahnemühle Baryta for example it makes no difference at all of course.
If you make lots of profiling with many different papers I think it can be worthwhile to upgrade to get a more efficient workflow. If you're not interested in OBA correction and only profile occasionally I think "upgrading" by using Argyll (which is free) is a good choice. Although I haven't been able to make an A/B comparison between the instruments I have no reason to believe that a Colormunki+Argyll profile would be noticeably less accurate than a profile made with i1Pro and bundled software, unless we talk about OBA correction which the Colormunki can't do.
From tests I've made using Argyll instead of the bundled software will make measurably more accurate profiles, even if you just scan a single A4 sheet, but I think four sheets (840 patches) is about the ideal to get "as good as it gets" without overkill when using a modern inkjet. The bundled software may have say DE2000 3.7 as max and 1.2 as average, while a 840 patch argyll profile will have say 1.8 max and 0.5 average.
I'm keeping my Colormunki.
Does the Colormunki photo stays good, also after a few years or does it need recalibration from time to time? (I can use an old colormunki.)
-
I suppose you could get it re-cal'd but considering that they charge $350 or so to re-cal and re-certify an i1Pro, it seems hardly worth it.
WRT i1Pro, it's not obvious to me at least exactly how bad the lamp aging can be and how much of a function it is of # of seconds of usage vs. something like drift even if it's not in use.
Do any of the resident color gurus have experience w/ that (esp. in a non-professional environment)?
-
Does the Colormunki photo stays good, also after a few years or does it need recalibration from time to time? (I can use an old colormunki.)
Save your money. If it calibrates to it's white reference, if the product produces profiles that appear OK, I'd not worry.
-
If it calibrates to it's white reference, if the product produces profiles that appear OK, I'd not worry.
I'm using my ColorMunki to learn more about color science. Part of the process is using it with Argyll to profile a cheap HP Envy 7640 four color printer I got to print the odd web page that would benefit from being in color (I have a laser printer for any other utility printing.) But my profiles aren't quite right. I haven't posted about this yet because I am still checking my workflow to make sure that I'm not making a mistake somewhere. (I also made profiles with the bundled XRite software and they aren't right either.)
One thing that makes me suspicious is that when I measure a ColorChecker chart (full size, about six months old that has been stored and handled very carefully, and also a several year old ColorChecker Passport) and the ColorMunki L*a*b* readings don't match other references:
Dry Creek Photo ColorChecker images (https://www.drycreekphoto.com/Learn/monitor_calibration.htm#charts)
BabelColor ColorChecker page (http://www.babelcolor.com/main_level/ColorChecker.htm)
Robin Myers Imaging Spectral Library (http://www.rmimaging.com/spectral_library/library_index.html)
I mostly used ArgyllPro ColorMeter and Argyll spotread. But also did a few readings with the bundled ColorMunki Design software (that is installed on a different computer. My Win 7 desktop computer is set up for Argyll and my Win 8.1 laptop is set up for ColorMunki Design.)
Should I be suspicious of my ColorMunki?
Wayne
-
Could be your target.
-
....
Should I be suspicious of my ColorMunki?
Wayne
I would also be suspicious of their spectrometer and targets.
I don't think an exact standard exists since Kodak left the scene with their control strips. They produced a set/control exposure strip, and you hoped your processing matched their standard strip plus whatever corrections they applied to it. With inkjet processing, it's pretty much pot luck, imho. Even x-rite in all their webinars mentioned no two monitors are exactly alike and no doubt same applies to their spectrometers. I have a couple of their spectrometers, and one is warm and the other is cooler in their reading values just messing with them this AM in ink mixing. Go figger! No doubt their is a tolerance in their Colorchecker charts too. Okay, I am also suspicious of the tolerance of the ink tints in various batches too. From my mixing of inks this morning, one drop can really screw up a cart of ink and how much of that works into an ICC profile I don't know.
Best we can do is come up with our own standard and work from that, given our equipment.
SG
-
Could be your target.
If you are referring to my profiles, then yes; there could be a problem with the target. As I said, I am still going over my profiling workflow. But ColorMunki Design is full on auto--I can't change the target when I make a profile with ColorMunki Design.
But forget I said anything about printer profiling. I'm not ready to report on this. I'm talking about the ColorMunki itself. Shouldn't I be able to measure official XRite ColorChecker charts?? And get values that are reasonably similar to the measurements from the sources I listed in my previous post?
And a 2nd question about my 7640. I measured the large pure black patch on the Digital Outback Photo Printer Evaluation Image (http://www.outbackphoto.com/printinginsights/pi048/essay.html) and L is never below 20. It is usually about 22. I had this image printed at WalGreens and the black patch measures about 6. I've also tried measuring the black patches from targets printed on my 7640 and they are always >= 20. This is with any of my profiles (Argyll and ColorMunki Design.) And also when I run the printer driver in "auto" mode (with Photoshop set to "Printer manages colors.")
Wayne
-
I don't think an exact standard exists since Kodak left the scene with their control strips.
I'm beginning to suspect that "color science" is an oxymoron. I'm not seeing much science anywhere. Only cargo cultish circular logic.
"The name derives from the belief that various ritualistic acts will lead to a bestowing of material wealth."
https://en.wikipedia.org/wiki/Cargo_cult (https://en.wikipedia.org/wiki/Cargo_cult)
Wayne
-
I'm beginning to suspect that "color science" is an oxymoron. I'm not seeing much science anywhere. Only cargo cultish circular logic.
Wayne
You may be right. Maybe "color art" since it seems more theoretical, or a WAG, at times.
If x-rite provided a "Standardized Colorchecker" target along with the numbers that they read off their reference spectrometer, one might get closer or at least interpolate a bit. As it is, it really is a crap shoot with who says what it should be. Might cost a small fortune to buy one though.
I've found the same nonsense trying to figure out whose white balance Kelvin is right among various software packages as well. Some may report 1,000 Kelvin difference from another software brand. So who is right? Hopefully, the engineers who build these cameras know best (with their own software too) than the third parties software.
A recent visit to the body shop I had a talk with the paint mixer guru. It really is a mess too in that paint dries down into a different color than their computerized mixer says it should be. He thinks yellow is the hardest to mix up. So he has to mix a batch up, spray it, and wait a day and re-anayize and then spectrometer says it's wrong on day two, but it is right on the car. Go figger. He doesn't think it's a science either. Helpful to get close, but it still takes a bit of eyeball to get it there over what his $20K gizmo says it should be.
SG
-
interestingly, I tried to essentially print out a color chart today. I had an actual paint sample, ran spotread on my i1Pro2 and created a PS document and filled it w/ that color (from the Lab space w/ integer rounding). after printing it on some luster paper (fully color managed), I was surprised that the L was right, the 'a' was pretty close (within 1.5 units) but the b was off by 5-6 units. it was 50/-1 / 40 approx. greenish yellowish grayish.
trying to figure out where my error is vs. color science/art. I strongly doubt its out of gamut of my 3880 but I will go back and check that.
-
Well my experience is different, I find colour management rather predictable. Usually it's me that forgot something ;-)