Let's dispel the myths about color management and ICC profiles.

[mikling]

For many years I was sold into the myth of ICC profiles will give you accurate colors as well as screen to print matching.

(This is NOT a topic for the inexperienced.) but it is interesting.

Is it real? Well it depends a lot on the workflow you follow. Very few follow it.

Let's get the myths out of the way. Colormunki, Xrite, Datacolor will sell you packages to give you printing Nirvana. Well it depends...most believe, you get the profile , use it and print photos bam,....excellent prints. Hardly any softproofing is in the workflow. Improved prints but it does not end there. Have we ever been told otherwise?

Read this article thoroughly, I've posted it before and I was surprised not many responses because it is largely misunderstood or is counter to what many many even pro photograhers believe. Using ICC profiles in printing..helps a LOT but what is behind what it was meant for is not appreciated nor understood when cases arise when it needs to be used properly.

I will take a pause and let some read the article and then digest it. Step back and think about what is happening and what the author is really doing. Then an understanding of what is going on may open up.

The third paragraph is important....very very important. It hints at the depth of color management. In the end I will show that what Hat uses....for color management is the "direction" that one also needs to take as well but in a more sophisticated way.

http://www.pixelgenius.com/tips/fraser-softproofing.pdf

 

[Roy Sletcher]

Interesting!

As a practitioner of colour management I will be interested in seeing how this discussion plays out.

The CMS concept is largely overplayed by those marketing the concept and its products for financial gain. I am not judging the marketing financial gain, but rather implying proceed with caution. "Caveat emptor"

Aside comments:
Bruce Fraser, the author of Miklings attached paper, has been dead for over 10 years. The best he ever saw was Photoshop CS, and commercial output devices that struggled to reach even a portion of the sRGB colour space. 66% of sRGB was close to cutting edge then. Colour reproduction software, and output devices have come a long way since then.

Not to denigrate what he wrote. He was a genius ahead of his time, sadly cut short by an untimely death. His comments are largely still valid because knowledge about the physics of colour predate the 20th century.

If you want to verify google Dr Alexander Munsell. A German Physicist and a pioneer in explaining colour through the colour wheel. A concept I learned about as a Polytechnic student over 50 years ago. (One of the few concepts I learned then that has not changed over time) His principals and writings are equally valid today even though MS, Apple, HP like to imply they invented colour reproduction and undoubtedly own patents in the field.

Bottom line
If: colour reproduction is critical on the images you produce or work on,
AND: you produce these images for output on multiple output device,
AND: you share these images with several people/customers/colleagues who have the same exacting colour standards

THEN: without doubt you need a full colour managed reproduction environment.

ELSE: you will go insane or bankrupt.


Now dismounting from my "hobby horse"


rs

 

[The Hat]

@milking, That’s the second time you have posted this PDF, and I still can’t get it to open, I’d copy the darn thing if I could just see it..

 

[Ink stained Fingers]

we all in this forum have probably some or some deeper understanding of color management and will probably agree on the important things, and able to discuss variations why somebody is doing something this way, and not the other way. The core question for me is - who currently does not know about color management, is somehow not happy with the outcome of his photography, on display or in print, and how can the essentials of and awareness for color management be spread. It's not always that automatic functions - auto white balance, automatic presets , image enhancement settings in the drivers etc can handle it all, and can even be counterproductive in a range of situations. Photo magazines don't talk much about it and sometimes even seem to create the impression that it's all too complicated

 

[mikling]

Hat, the link is working. It's something with your browser. An addon is probably disabled. Try going to pixel genius directly.

Over the weekend I ran an exercise with toolman Joe. Here I went over several printers and more importantly compared the output of Canon's desktop as well as the P800, 3880, P600, R3000, R2880. All the Epsons used the same ink and profiling tools software etc. All the Epsons looked largely similar but there were differences...even in color. The question to many would be why? and what was happening? Did they not use the same ink? why the differences?

How do we explain that?

I'll invite you to step back and think...what if we put in orange instead of yellow ink. What kind of profile would be produced and how would it print?

 

[palombian]

I was lucky to buy a Colormunki Photo at half price so I am not bankrupt.
Guess I recovered the money already since I can print now on any paper.

Not every paper/printer/ink combination works but - as example - I am able to proof on A4 with a MG8250 dye on an unknown cheapo satin paper and print afterwards on A3/A3+ with a PRO9500 II pigment on - what I suppose is - old stock Ilford Smooth Pearl with the same result.

In fact what I do is using the quite good gamut power of a 6 ink dye printer on beneath average paper to achieve the same gamut on a (now also grandfather) 10 ink pigment printer on again not the latest paper.

But I have consistent and pleasant results at very very acceptable costs.

It took me some time and a lot of trial profiles and prints but IMO profiling can work.
It is the logical complement of refilling.
Why cheat on ink if you have to spend it on paper afterwards ?
(although results on expensive papers are better )

 

[Roy Sletcher]

Hat, the link is working. It's something with your browser. An addon is probably disabled. Try going to pixel genius directly.

Over the weekend I ran an exercise with toolman Joe. Here I went over several printers and more importantly compared the output of Canon's desktop as well as the P800, 3880, P600, R3000, R2880. All the Epsons used the same ink and profiling tools software etc. All the Epsons looked largely similar but there were differences...even in color. The question to many would be why? and what was happening? Did they not use the same ink? why the differences?

How do we explain that?

I'll invite you to step back and think...what if we put in orange instead of yellow ink. What kind of profile would be produced and how would it print?

Simple explanation.

There has to be a spectral balance between ALL the colour that comprise an ink-set. When, or more accurately IF, this could be be achieved, then the combination of the colours would combine to print a perfect black using our CMY colours, and without the need for a separate black. The fact that this does not happen is because of defects or limitations in the spectral purity of our pigments and dyes, which are man made, and subject to all the variables of measurement and mass production. Good ink-sets get close, but as yet, NO cigar!

Unless we are privy to their print driver formulae for conversions from RGB to CMYK we cannot make any conclusions about the different printers. If you are using a commercial RIP instead of a proprietary printer driver then you have more information to work with and more control of the final output. The downside being cost and learning curve.

Substituting an orange for yellow. as suggested above, intensifies not mitigates the problem of colour or spectral imbalance.

There is a wealth research and the associated literature on this subject going back decades by the makers of inks and coatings for the commercial printing industry.

Colour, falling in the realm of physics, depends heavily on mathematics for verification, and not simple maths either. To debate or discuss this meaningfully gets very intense.

As amateurs we are probably best advised to work within the simple tools at our disposal.

rs

 

[Ink stained Fingers]

Over the weekend I ran an exercise with toolman Joe. Here I went over several printers and more importantly compared the output of Canon's desktop as well as the P800, 3880, P600, R3000, R2880. All the Epsons used the same ink and profiling tools software etc. All the Epsons looked largely similar but there were differences...even in color. The question to many would be why? and what was happening? Did they not use the same ink? why the differences

Did you use the same paper in all printers ? did you compare the gamuts of the various printers - if the gamuts are not very similar different areas may be out of gamut and create a different color perception - if you use different papers they may have a different amount of OBA's - different printers probably have different drivers - and you may have to play with the driver settings to get matching gamuts and a better matching output- And I'm afraid there are a few more of those variables to consider like paper selection - print quality

 

[mikling]

In the last sentence of the third paragraph .
"If you want great, rather than just good, you need to
optimize images for different output processes, because something always
has to give, and each image demands its own compromises."

Essentially what it is saying is that when you use an ICC profile, as best as you can get it, it will compromise each and every image. It will give you a "good" print but it will compromise. Essentially the same was said in the same paragraph as well. "
So no output profile, however good it is, does equal justice to all images."

Let's figure out why this is so. What is the compromise. Review the article and see how the author is recommending shifting colors, adding layers. This is Hat's way of color management. See a parallel? This is kind of counter to ICC profiles right? Generally when we have someone complain of poor color reproduction we say get a custom ICC profile instead of playing with the sliders. But darn, after getting an ICC profile, to use it properly we supposedly are left playing with sliders and curves again! wasn't that great ICC profile supposed to get us away from that playing around. Hit the print button in combination with the profile and you get perfect color. Heck, even the pro reviewers of printers are advocating this. So what gives?

Oh to the prior question. Yes, obviously same papers and it also raises questions about gamut......we hear ink sellers saying their ink has larger gamut.......even the printer mfrs say this. Is it inks or is it in combination with something else? We're looking at this thing at a deeper level to gain an understanding of what goes on.

I've been thinking about this for over a year.

 

[Ink stained Fingers]

Over the weekend I ran an exercise with toolman Joe. Here I went over several printers and more importantly compared the output of Canon's desktop as well as the P800, 3880, P600, R3000, R2880. All the Epsons used the same ink and profiling tools software etc. All the Epsons looked largely similar but there were differences...even in color. The question to many would be why? and what was happening? Did they not use the same ink? why the differences?

I did something like that on a much more limited scale - I'm using the same pigment inks on a Pro 7600 and a R265, I'm getting similar gamuts when printing onto the same paper what I normally don't do. The drivers of these two printers are very different - the 7600 gives me the option to change the overall ink density, not like with a RIP per color but just as an overall adjustment, desktop printers don't offer that adjustment. This parameter allows me to prevent ink puddles, or on the other side increase the ink amount if the paper takes it - and that makes some difference in the profiles, in the gamut, specifically in th darker range below L=40. And there are other differences in the drivers like resolution setting, the droplet size of the printers are different, and some more. And it is interesting to see that printouts look very much alike from these printers with a resp. profile. And if I tweak the driver settings such that a diferent gamut can be expected this becomes visible in the printouts in the saturated color regions with some more detail definition. Overall I'm not surprised that you get different results from different printer models. And it may take some effort to identify that variable which actually creates that difference.