Luminous Landscape Forum
Raw & Post Processing, Printing => Colour Management => Topic started by: LucDelorme on July 15, 2013, 01:33:29 pm
-
I did a test recently, calibrating and profiling the same Dell U2711 monitor connected to my PC (win 8), and them on my Mac (10.8.4). I used the latest ArgyllCMS and DispcalGUI on both platforms, with the Same i1Display Pro sensor. I was quite surprised to see the difference between the two calibrations (see attached color accuracy measurements).
I used the same methodology and calculation spreadsheet for both, as detailed here (http://stephenstuff.wordpress.com/2011/06/26/complete-display-calibration-with-argyll-cms/) - Great tool BTW!
The Win8 calibration and profile is shockingly accurate, while the Mac OS one is quite horrible. I have to wonder if the problem lies within the ArgyllCMS packages on either platform, or in the color management system within the OS. Visually, I could tell that some of the patches were off on the Mac also.
On another note; I was very pleasantly impressed with the performance under Win8. I didn't think the U2711 would have been that accurate. There is some variance in uniformity of about -15% in the lower left corner, so it's not perfect, but for the price, it's quite impressive.
-
Print a known color target with a wide range of memory colors letting Photoshop manage colors between the two platforms.
Which calibrated display matches the print the closest viewed under a Solux lamp or equivalent daylight balanced viewing light?
Maybe ArgyllCMS tunes its software toward the way Window's systems deal with color managed previews. It's been proven in the past software engineers tend to favor one platform's API over another.
But thats just my guess until you do the WYSIWYG test to determine the reference point for "accuracy" over another. We don't view color patches to judge accuracy, we judge real photos of scenes which involves a completely different way of seeing things.
-
I did a test recently, calibrating and profiling the same Dell U2711 monitor connected to my PC (win 8), and them on my Mac (10.8.4). I used the latest ArgyllCMS and DispcalGUI on both platforms, with the Same i1Display Pro sensor. I was quite surprised to see the difference between the two calibrations (see attached color accuracy measurements).
I used the same methodology and calculation spreadsheet for both, as detailed here (http://stephenstuff.wordpress.com/2011/06/26/complete-display-calibration-with-argyll-cms/) - Great tool BTW!
The Win8 calibration and profile is shockingly accurate, while the Mac OS one is quite horrible. I have to wonder if the problem lies within the ArgyllCMS packages on either platform, or in the color management system within the OS. Visually, I could tell that some of the patches were off on the Mac also.
On another note; I was very pleasantly impressed with the performance under Win8. I didn't think the U2711 would have been that accurate. There is some variance in uniformity of about -15% in the lower left corner, so it's not perfect, but for the price, it's quite impressive.
Same issue here but on Win 7. That`s why I gave up using Mac Os 10.8.4. I just keep my MacBook Pro late 2008 to run Win7 on Boot Camp. In my own experience, everything changed after Apple released Snow Leopard. I don´t know why but I couldn´t calibrate my monitor correctly on Mac anymore. On Win7, the problem was solved (and besides that, I can still use my iOne Display 2 and APS to profile my monitor and HP13200ps printer).
-
I too had horrible results on Mac, and tried to find a solution via the mailing list support with no success. How did I solve? Since I use the i1Display Pro I just kept using the maker's software, which is much easier and gives high quality, repeatable results.
-
I forgot, I have a Dell U2711. Anyway, even my MacBook Pro's display had quality issue.
Definitely there are troubles with ArgyllCMS, more than OS X alone, since the other commercial products work just fine on a Mac.
-
> Maybe ArgyllCMS tunes its software toward the way Window's systems deal with color managed previews.
It is not.
-
> That`s why I gave up using Mac Os 10.8.4
Have you tried replacing Argyll binaries with those Graeme published here http://www.argyllcms.com/disptools.osx64.tgz
More here: http://comments.gmane.org/gmane.comp.graphics.argyllcms/11230
-
Yes, I have replaced those binaries for Mac. Without them, ArgyllCMS cannot deal with wide gamut on Mac at all.
-
Sorry for your trouble. Maybe to try contacting Graeme through argyllcms list at http://www.freelists.org/list/argyllcms ?
-
That would be a logical next step. It's not a huge concern for me right now since I do most of my work on PC now that I've switched from Aperture to Lightroom 5, but it's still intriguing.
-
> Maybe ArgyllCMS tunes its software toward the way Window's systems deal with color managed previews.
It is not.
It is.
If you can't explain it simply, Iliah, you don't understand it enough according to Albert Einstein. I think I'll go with what ol' Al baby says.
-
It is.
If you can't explain it simply, Iliah, you don't understand it enough according to Albert Einstein. I think I'll go with what ol' Al baby says.
So you understand and can explain Mr. Einstein's theory. Good for you. Where is that raw btw?
-
It is.
Argyll is open source, so Tim you can point that particular piece of code there, right ?
-
So you understand and can explain Mr. Einstein's theory. Good for you. Where is that raw btw?
If you're going to dance the "obtoosy" with my quoting Einstein's views regarding explaining complex concepts (i.e. API driven color managed algorithms, not the theory of relativity), then you're just going to have to continue dancing in the dark by yourself. I don't see any explanations from you that indicates you actually know what you're talking about.
I don't know a lick of code and don't claim to and I don't have to. I also don't need to use ArgyllCMS, but common sense tells me when a piece of software acts differently between two platforms when similarly functioning commercial software doesn't, well all I have to say is when all things are the same...blah, blah, blah...I believe your vernacular or at least it might as well be going by past exchanges with you.
I resort to my old screengrab samples I made of Mac OS 8-9 API driven color managed preview color errors as an indicator of how the math can screw up accurate color just for demonstration purposes here. At least I can show something to back up what I said. How about you, Iliah?
-
I can show something to back up what I said.
You said "Maybe ArgyllCMS tunes its software toward the way Window's systems deal with color managed previews. It's been proven in the past software engineers tend to favor one platform's API over another."
I considered it as an honest "maybe".
Next you replaced your "maybe" with "it does".
Now you are to show how. And please show that raw with "colormetrically accurate by the numbers (http://www.luminous-landscape.com/forum/index.php?topic=80130.msg647217#msg647217)" CC24. Or follow your own note (http://www.luminous-landscape.com/forum/index.php?topic=80130.msg647132#msg647132) to yourself - "Ignore Iliah's postings from now on for they're a big waste of time to read"
-
Try a different calibration & profiling package, if for nothing else, just to see if there's a difference like with argyll. I recommend basICColor's display. They have a free trial (14 days perhaps) which would let you try it on both.
The video cards are obviously different. I don't know Mac's at all. Perhaps your Mac video card doesn't support loading LUTs (lookup tables), or has some other limitation causing the issue. (Only applicable if other calibrating & profiling packages have same difference.)
As Iliah mentioned, try the argyll list. The author is impressively active, especially for free software with a solo author/maintainer. (This might be something not diagnoseable for him without being at your computers, but who knows.)
-
Where is that raw btw?
Must be with the proof you haven’t provided that the ACR engine has an invisible issue. In the box you labelled "all in good time".
-
"Ignore Iliah's postings from now on for they're a big waste of time to read"
I'm only going to respond to you when you first counter with useless information and will stop responding once I see that's all you've got left. When you stop making sense, I lose interest to respond.
So far you don't come across as very knowledgeable on the subjects discussed here and on the other thread you linked to due to your useless, indecipherable comebacks.
-
Must be with the proof you haven’t provided that the ACR engine has an invisible issue. In the box you labelled "all in good time".
Are you saying that raw does not exist?
-
So far you don't come across as very knowledgeable on the subjects
You to judge? LOL
-
Are you saying that raw does not exist?
I'm saying and have said your idea about this invisible problem with the ACR engine has yet to be shown let alone proven. Despite your promise to do so!
-
I'm saying and have said your idea about this invisible problem with the ACR engine has yet to be shown let alone proven. Despite your promise to do so!
LOL. I asked a simple question, where is the raw. Not some study, that takes time and caution, but a raw from back. what, 2007?
So, where is it? Is it yours? Or Tim called Mom?
-
LOL. I asked a simple question...
So did I. You just refuse answer it.
I'd be happy to tell you where all my raws are. I can't comment for others.
-
So did I. You just refuse answer it.
I'd be happy to tell you where all my raws are. I can't comment for others.
I asked Tim :) He can answer for himself.
-
I asked Tim :) He can answer for himself.
But can you? So far, you haven't.
-
I seem to remember a few issues coming up with OS X Lion and loading calibrations where they didn't stay loaded. Have you tested to make sure the calibration actually got loaded before you built the profile and stayed loaded once you tested it? You can use dispwin to download the calibration information to a file, which you can then compare it to the original .cal file. Since I've been using Spectraview/Multiprofiler I haven't had any problems, but I remember pulling my hair out for a while because occasionally the system default profile would "magically" become set instead of the custom profile. That was a while ago, so I don't remember which OS it was.
Do you see the custom profile in the color tab of the display preferences?
-
But can you? So far, you haven't.
I see Tim needs Mom.
-
I asked Tim :) He can answer for himself.
Can you refresh my memory on why you need me to supply the Raw file of the 24CCchart in the other thread? I'll be happy to supply it but I'm not going to do it if I don't learn something new you have to offer. I did this before with another contributor on why blue shifts hue when converting from ProPhotoRGB to sRGB and it was a waste of time because the contributor didn't even bother to post process the image in order to normalize it so the blue could shift. He left it in its dark Raw unprocessed state which was totally a waste of time.
I don't plan on doing the same especially with you.
And an added note about that colorimetrically accurate 24CCchart I posted, I had to edit primarily the contrast of the image to get the Lab numbers to match, but I was using old CS3 PV2003 processes. After playing around in LR4.4's PV 2012 processes that same Raw image won't deliver the same Lab numbers.
Just FYI before you respond for your reasons for my supplying the file.
-
Can you refresh my memory on why you need me to supply the Raw file of the 24CCchart in the other thread?
Because there is a problem with that shot, and I want to run an analysis on it.
-
Because there is a problem with that shot, and I want to run an analysis on it.
What's the problem you see and how does this problem relate to the reasons I posted the image in the other thread which I emphatically made clear?
And just FYI if you're going to continue to give cagey responses like the one above, you can forget about my cooperation on this. You're not sparking my interest on this, Iliah, with those kind of answers.
-
What's the problem you see
That is exactly why I'm asking for the raw - to demonstrate the problem. And it is really an interesting one.
Now just to stay on topic - have you found the proof for your "It is" in this very thread?