QTR-Quadtone RIP

You need to actually embed Adobe RGB or Gray Gamma 2.2 into the target to get the proper printed values.

Then linearize, and then your curve will print linear for any image embedded with a gray gamma 2.2 space (specifically Adobe 1998 and Grey Gamma 2.2)

best,

Walker

On Dec 6, 2016, at 8:20 PM, nigglefish@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:

I am using PrintTool and QTR on Mac OSX. I have been creating and linearising curves but I am a bit unsure of how to proceed to actual printing. I'm sure I can work it out through trial and error, but that won't help me understand the underlying color management concepts any better. So I thought I would describe what I have done and ask how I should proceed. (A few details have been simplified.)

1) I created a .quad QTR curve using my own software and installed it.

2) I created an 8-bit 18 patch grayscale target image in Gimp with each patch having a value 15 units higher than the last (0, 15, 30, 45, ..., 255). When I open this in PrintTool it reports the embedded profile as 'untagged'.

3) I printed the target using PrintTool selecting 'No color management'. I used the QTR driver and selected my curve in the QTR print settings.

4) I measured, adjusted, linearised and printed many times.

5) I finally printed the target and measured linear lightness readings across the patches. That is, after measuring with a ColorMunki the difference between each adjacent patch in terms of L* readings was (just about) equal.

I'm happy to have gotten to this point, but I'm not sure if I have been headed to the right place or not. Maybe the desired readings for each patch should have been different? I don't know. I now have a photo I want to print. It reports an embedded profile of 'Adobe RGB (1998)' in PrintTool. I'm not sure what settings I should choose in terms of color management now to print that. I am assuming I should select 'No color management' again. But again I don't know.

So I have two questions:

1) Given my target and printing options what lab readings should I be trying to get for each patch?

2) Once I have achieved those readings what setting should I use to print a photo with an embedded profile.

Happy for any other information, background or guidance. Thanks in advance.

What I would add to Roy's useful summary is that (i) approaches 3 & 4 require a measurement device, which you have, whereas 1 & 2 don't; and (ii) the difference between approach 3 on the one hand and 2 & 4 on the other hand seems to reflect a difference of view about what sort of print you want - 2 & 4 tend to deliver more 'pop' and contrast, whereas 3 tends to deliver more open shadows and shadow detail, particularly on matte papers. I've written a partly tongue-in-cheek commentary on this difference of view here: http://www.cyberhalides.com/piezography-printing/the-piezography-heretic-to-convert-or-not-to-convert/ . This is written from the perspective of a Piezography user, but it applies to QTR with any inkset.

This article on using the Munki with QTR may be helpful, but note that it is a bit dated now:
http://www.northlight-images.co.uk/better-black-and-white-profiling-with-the-colormunki/

---In QuadtoneRIP@yahoogroups.com, wrote :

?? Targets for QTR linearization or for ICC profile creation are ALWAYS printed without

Color Management. "No Color Management" specifically ignores the embedded profile

even if one exists. So tagging targets has no effect if printed with NoCM.

QTR linearization creates a straight-line graph of K-values vs L--values, where K is the whole

range from K=0 (white) to K=100 (black) and L ranges from dMin to dMax. This is quite

different from Gamma 2.2 which is not a straight-line.

There are numerous ways to deal with this difference and you'll see that everyone's got

a favorite that they "swear by". The basic ones are:

1) get used to it, edit your files till they produce the print you like.

(might be surprising but I bet this is the most used -- after all that's what we did in darkroom)

2) add correction curve in Photoshop. It's a very convenient idea for Windows.

See www.paulroark.com for info on this.

http://www.paulroark.com/BW-Info/Eboni-4-Plus.pdf page 16 has a very good description

of the issue and of the solution

3) make ICC profile for soft-proofing. basically what you do here is edit your image under

soft-proofing so that output print matches what you want. nice for Windows but your files

are now targeted for a specific printing output

4) finally there's the standard color management approach. Make an ICC profile and print

with it. On a Mac with Print-Tool this is by far the easiest and in a sense the "standard" way.

With nice linearized QTR profiles the generic "QTR driver Gray Matte (or Photo) Paper"

ICC profiles do quite well but a customized one with all the QTR driver blending selections

etc is the best. Personally I like this the best but it's less convenient for PCs. Also if you are

used to #1 above you'll find it disconcerting at first because you have to re-get-used-to-it.

Roy

Roy,

Regarding Option 4 above ("standard color management approach)" you say "Personally I like this the best but it's less convenient for PCs".

Why is it less convenient for PC users? I'm still learning QTR but I do understand that QTRGui does not support using an ICC profile, and to me this means "it is not possible" rather than "it is less convenient".

So I'm wondering if I'm missing something based on your statement. Is it somehow possible to print with QTR on Windows (not using QTRGui??) and use an ICC profile when printing? If not, is there any hope of getting an updated QTRGui that would support this (and get on par with Print-Tool in this regard)?

I'm really interested in this since I have invested in the new Piezography Pro system from IJM and it seems that I'm going to be partly handicapped on the Windows platform without the ability to print and use an ICC profile. I fully understand I can make the ICC profile and soft proof, but as you say, this means baking the file.

Thanks,
Dave

Thanks so much for the discussion. I think I can put my finger on what I was missing now. Can I run over my final understanding for confirmation?

So, I have a .quad curve that gives lab-linear readings when given an untagged image of linear RGB patches and I print using No Color Management. To print my photos I could (but won't) convert them to use a lab-linear ICC profile (adjusting the pixel values) and then print with No Color Management again.

Or, I could (and will if this makes sense) follow suggestion 4 and take readings of my final target print, construct an appropriate file from the readings, drop it onto QTR-Create-ICC and make an ICC Profile that characterises the non-linearities left in my curve. I would then select Application Managed color management and select the created ICC profile. My photo files could be tagged with any profile after that.

Is my new understanding correct? If not I am probably more confused than ever :)

Just to clarify another point; is the reason that the generic "QTR driver Gray Matte (or Photo) Paper" ICC profiles are sub-optimal because they won't characterise the non-linearities in my curve and they don't know what my dmax is when working out a Y -> L* conversion?

Also, at one stage I installed a program called ColorFaker which replaces the default OSX sRGB ICC profile to allow some programs to work properly (can't remember why I installed it - google told me to). I was accidentally printing an untagged image with QTR and Print-Tool with No Color Management with ColorFaker turned on (the profile replaced) and the results were different than with it turned off. I was wondering where in the flow of things the system sRGB profile came into it.

Thanks again!