ICC screenshots - our workflow

[UncleDoug]

Got another batch together.

Figured I'd put our little color managed reality under the spotlight right off the bat.

Since you are limited to two images per post, this will span 3 messages.

I have prepared 5 comparrisons. All are anotated as to what they are comparring.
(in order)
1. Our Tango profile vs. BestRGB (our default working space)
2. Monitor profile vs. BestRGB
3. Our InkJet profile for hi-gloss white film, hi-rez. vs. BestRGB.
4. All of our devices together vs. BestRGB
5. All of our devices together vs. ProPhotoRGB

Images 4 and 5 have the working spaces opacity pushed back abit to get a ghosting look inside to see how much volume of the working space is not covered by our devices.

[UncleDoug]

next....

[UncleDoug]

and the last two....

[UncleDoug]

Would love to hear any comments, questions, what ever.

I have my oppinions as to what all of this may mean or what is going on, but I'd love to hear it from you all first.

This has been EXTREMELY educational for me.
It has helped coalesce many concepts into a solid understanding, I think :D

[ssim]

ok, I'll bite. I'm not exactly sure how to read these. Care to go into a further /detailed explanation.:confused::confused:

[UncleDoug]

Sorry about that, a mild explanation is in order. :oops:

These are 3-dimensional plots of select color spaces intended to show their relation to each other in terms of the LAB color space. The color space that makes Color Management possible.

For each comparrison I made screen captures looking head on and from behind in hopes of giving an overview of these volumes and their relationships to eachother.
What gets clipped?
What is "wasted space"?
What are the trade offs of various workflow/color space combinations?
These are the basic questions I ask when looking at these comparrisons.

An example analysis from my perspective.

In the comparrison of our FJ-500 inkjet printer vs. BestRGB, BestRGB(our working space), is represented by the white/grey volume. The FJ-500 is represented in color.
Looking at the image almost all of the volume of the FJ-500 profile is encompased by the BestRGB profile, except in darker saturated blue-greens.
What does this mean?
Our chosen printer/media combination has the ability to render some colors, darker saturated blue-greens, that fall out side our working space.

Make sense?

This got me thinking.
I need to post a comparrison of scanner vs. printer.....

[UncleDoug]

Here is the comparrison of our Tango scanner vs. our InkJet.

See a similarity between this one and one of the previous ones?

Pretty interesting...

[Scottes]

OK, in the shots for Our Workflow vs BestRGB and ProPhoto... Why are the Our Workflow profiles different? If your profiling the same thing twice shouldn't they both be the same?


Next, how come you (High Mountain Imagery) only offer Large Format printing? Can't I get some 20x30" prints from you?
:-)

[jfrancho]

Are you saying that the scanner captures color that is out of gamut with the workspace? And if that is true, the comparison shows that the monitor won't show you all the available color in the workspace? However the scanner and the inkjet work in similar realms. I think I have this, but need more - only about 60 pages into Real World Color Management. Tell me more - what tools dod you use for this?

[UncleDoug]

OK, in the shots for Our Workflow vs BestRGB and ProPhoto... Why are the Our Workflow profiles different? If your profiling the same thing twice shouldn't they both be the same?

The device(color) volumes in the comparrisons you are talking about did not change.
All that was changed was the working space, BestRGB and ProPhotoRGB.
What I was intending to show there was how the two working spaces meshed with the devices in our loop, scanner, monitor, printer.
The ghosting of the working spaces may be affecting the appearance of the color volumes.

Next, how come you (High Mountain Imagery) only offer Large Format printing?

For production purposed we only run our FJ-500. Can print up to 50''x "how far do you want"?
Have not delved into the mini-lab yet. But who knows!

Can't I get some 20x30" prints from you?
:-)

No problem at all.

Contact me via our site, phone or e-mail is ok.

[UncleDoug]

Are you saying that the scanner captures color that is out of gamut with the workspace?

You got it.
Happens to a slightly lesser degree with ProPhoto, but info is still clipped right from the get-go.

And if that is true, the comparison shows that the monitor won't show you all the available color in the workspace?

Got it again.

However the scanner and the inkjet work in similar realms.

I think I might have sand-bagged you here, by not ghosting the last comparrison of scanner vs. printer. The Scanner, by definition, has significantly more volume than the printer space. If I had ghosted it you could see this easily.
What I thought was interesting is that there are some blue-greens that the printer can render that the scanner can not capture and the working space does not contain.

I think I have this, but need more - only about 60 pages into Real World Color Management. Tell me more - what tools dod you use for this?

ColorThink (http://www.chromix.com/colorthink/) is the application.
Darn cool tool.

Here is a re-post of the scanner vs. printer comparrison, the scanner volume is ghosted to reveal the printer space inside. Same comparrison as las time, just from other angles to show the differences in the two.

[jfrancho]

Thanks for the link to the tool. The last comparo is really interesting. So, do we have a moral to the story? Do you get any unexpected results in that satyurated blue-green area?

[UncleDoug]

Thanks for the link to the tool. The last comparo is really interesting. So, do we have a moral to the story? Do you get any unexpected results in that satyurated blue-green area?

To me the over all moral to the ColorManagement story as illustrated here is;

There is no perfect answer to a workflow.
I feel, as well as others, that you need to look at all the device profiles in your workflow and find a working space that you feel bridges the gap the best between input and output spaces.

As far as getting unexpected results, no.
Because I am aware of the mismatch I mentally compensate, for both visual and numerical, when adjusting, and when necessary convert to the Output space and sent that directy to the RIP with color management at the RIP turned off. Just like you would do if you were using a DryCreek profile and asking the Tech at costco to not use any color management.

That is what is soooooo great about looing at profile spaces in 3-d, you can actually see what the heck is going on, know how your devices perform in realtion to each other and thus be able to leverage the performance of the devices to obtain optimal results.

[jfrancho]

Now, obviously you have scrutinized your workflow more closely than the average person with a scanner/digital camera, computer, editing program, printer. It doesn't seem that hard to perform a color mangement analysis ourselves, yet it is unlikely that many do. I am betting that many are way out of whack with what can be considered acceptable, even by hobbyist standards. It seems the only hurdle to maximising color management in the workflow is the time to learn how, and the cost of the tools. I'm gonna guess at a few months to read and experiment, and maybe $150 worth of tools. And some time to completely screw it up, and then fix it... What are we risking by blindly not taking control? Is this the source of a lot of the complaints and wierd solutions posted here?

As far as being able to see the profile in 3D, I totally agree. In fact, a few months back, I asked if there was value in being able to see RGB histograms represented in 3D. Apparently I was the only one that thought so, since no one responded. I still think it is a good idea. I'd like to ree the color of an individual image set against the profile at various stages/devices in the workflow.

It sounds like you already had suspicions about the mismatch in your workflow, and have enough experience to work with it. The whole management strategy looks pretty good to me, given the nature of such a workflow beast, and the compromises. What got you to this point, and is this just BAU for you?

And, BTW, thanks for taking all this time. Pretty cool to have this info as I am learning about it.

[UncleDoug]

Now, obviously you have scrutinized your workflow more closely than the average person with a scanner/digital camera, computer, editing program, printer. It doesn't seem that hard to perform a color mangement analysis ourselves, yet it is unlikely that many do. I am betting that many are way out of whack with what can be considered acceptable, even by hobbyist standards. It seems the only hurdle to maximising color management in the workflow is the time to learn how, and the cost of the tools. I'm gonna guess at a few months to read and experiment, and maybe $150 worth of tools. And some time to completely screw it up, and then fix it... What are we risking by blindly not taking control? Is this the source of a lot of the complaints and wierd solutions posted here?

I'd venture to say so. If you don't take control, you are at the whim of the companies your devices/software was built by.
Things right out of the box will get you into the ball park, but if you want to sit behind home plate, you need to do a little work. :D

As far as being able to see the profile in 3D, I totally agree. In fact, a few months back, I asked if there was value in being able to see RGB histograms represented in 3D. Apparently I was the only one that thought so, since no one responded. I still think it is a good idea. I'd like to ree the color of an individual image set against the profile at various stages/devices in the workflow.

It sounds like you already had suspicions about the mismatch in your workflow, and have enough experience to work with it. The whole management strategy looks pretty good to me, given the nature of such a workflow beast, and the compromises. What got you to this point, and is this just BAU for you?

YOU JUST ANSWERED THE $64,000 question!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Actually you just used the word that is integral to each and everyones individual color workflow.

COMPROMISE!

I can not state this enough.
You have to compromise in your choice of color workflow.
You are bound by the performane characteristics of your devices and the math behind working spaces.
You must compromise in color management, no exceptions.

Yes this is business as usual, at least these days.
Taken me 3 solid years to get to this point and I still have alot to learn. :D

And, BTW, thanks for taking all this time. Pretty cool to have this info as I am learning about it.

No problem at all.

Like I have said before, I'm probably learing as much as anybody in this forum regarding color because of these discussions.
It has forced me to coalesce the info in my head into words.

So, a bit of thanks from me to anyone that has posted to the threads regarding color is in order.

Thanks!

[DocFrankenstein]

I need a book about LAB color now, cause I have no idea as to what the axis mean.

Doug - isn't the whole point of using profiles is that you don't have to worry about the 3D clouds? WYSIWYG and everything?

[UncleDoug]

I need a book about LAB color now, cause I have no idea as to what the axis mean.

Doug - isn't the whole point of using profiles is that you don't have to worry about the 3D clouds? WYSIWYG and everything?

LAB

L=lightness on a scale of 0 to 100
A=Red-Green oppostition on a scale of 0 to 100
B=Blue-Yellow opposition on a scale of 0 to 100


In theory you are right, but a mild understanding of what is going on sure can help with things.
I think the WYSIWYG thing would be more of a reality IF there were more solid standards and all the soft/hardware companies agred to abide by the standards.
Kind of like the DVD format battle right now....

[jfrancho]

I've done a little more reading, and more research online. It seems that the small amount of loss you have in the saturated blue greens are pretty normal. In fact this IS the compromise we should seek, for if we try to encompass this are, the rest will really be out of gamet throughout the workflow.
The other thing I picked up was that scanners and cameras are always operating at wider range than can be handled by a colorspace, but that makes sense because the camera has to turn what it "sees" or is confronted with into data.
I think we may be a waiting a while for a true wysiwyg model. Color seems too subjective to emotion, opinion, mood and surroundings to package as a one size fits all product. Maybe not, it is concievable that with a calibrated monitor, and profiles that map to what that monitor is showing at all stages from work space to printer. I suppose there would be other technicalities that involve specific paper profiles and what not, but I am getting off topic here.
Anyway, this is a great topic, that will be largely ignored because it doesn't answer the ubiquitous "Should I use aRGB or sRGB in PS" questions. I'm always afraid to answer that question, now knowing what I know, because there many other prerequisite questions that need to be answered before you can answer that!
Thanks!

[UncleDoug]

Me thinks you got a handle on it now!!!!!!!!!!!!!!!!!!!!!!!!

I think we may be a waiting a while for a true wysiwyg model. Color seems too subjective to emotion, opinion, mood and surroundings to package as a one size fits all product. Maybe not, it is concievable that with a calibrated monitor, and profiles that map to what that monitor is showing at all stages from work space to printer. I suppose there would be other technicalities that involve specific paper profiles and what not, but I am getting off topic here.

Not off topic at all. Actually dead-on topic!
If you are going to pilot an airplane, why do you think they make you learn aerodynamics?


Anyway, this is a great topic, that will be largely ignored because it doesn't answer the ubiquitous "Should I use aRGB or sRGB in PS" questions. I'm always afraid to answer that question, now knowing what I know, because there many other prerequisite questions that need to be answered before you can answer that!
Thanks!

Yup. You got it.

Gongrats.!

[SuperFly]

I'm attempting to understand colour management doing some reading etc... just wanted to say this post has been helpful in showing me just how much more i have to learn :)

I think investing time in understanding this topic separates great prints from normal prints.

[DocFrankenstein]

Doug - you're the only one who can answer this question with any scientific basis:

What's better: CRT or LCD? I'd imagine if you are working with drum scanners and such, you'd be inclined to acquire a "good" monitor for post processing.

I heard NEC/mitsubishi is really good with CRTs. And I heard people with LCDs swearing by them.

How do the "clouds" compare? What monitors do you use in your printing facility? (If not a secret of course) :)

[UncleDoug]

Doug - you're the only one who can answer this question with any scientific basis:

What's better: CRT or LCD? I'd imagine if you are working with drum scanners and such, you'd be inclined to acquire a "good" monitor for post processing.

I heard NEC/mitsubishi is really good with CRTs. And I heard people with LCDs swearing by them.

How do the "clouds" compare? What monitors do you use in your printing facility? (If not a secret of course) :)

We use a mix. Eizo color edge LCD for adjusting and CRT's for the second monitor in a dual configuration. Those Eizo monitors are a chunk of change!
These days LCD is the way to go though. Consistent, even illumination and little drift over time. This is the main problem with CRT's - they drift way to easy in comparison to LCD's. The phosphors age and diminish in their "quality". We calibrate and profile our CRT's every monday morning. The LCD's every 2 weeks.

Mitsubishi monitors are supposed to be good but we had a Mitsubishi Diamond Pro that went to heck in a hand basket early in it's life, just after the warranty ran out, - bias -.