QTR-Quadtone RIP

Yes, it would be great if there was a step by step screen shot(s) or better yet a tutorial video on the process. Hint hint!

Mark
stillrivereditions.com

> On Apr 7, 2016, at 9:29 AM, artodd138@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:
> 
> 
> 
> Hi Everyone,
> 
> I've just begun working with QTR and I'm having a bit of trouble creating my own custom curves.  I've been working through the documentation that comes with QTR, however I fear that some of the tutorials are slightly dated.  I was hoping someone on here could direct me to more recent tutorials on the process.  Below I've outlined my setup and where I hit roadblocks.
> 
> I'm working on a Mac 10.11.3, Photoshop CC 2015, QTR v2.7.7, X-rite i1 Pro and i1 Profiler, an Epson 7800, and Piezography Special Edition P2 inkset.  I'm able to install QTR and print with Jon Cone's P2 curves just fine.  I want to try creating custom QTR curves for papers without Piezo curves available.  Through the tutorials provided with QTR I am able to print the calibration sheet, determine default ink limits and crossover values for the 6 shades.  Where I run into trouble is placing these values into the text file and then creating the curve.  Supplied with QTR there are no ink descriptor files that are close to the set up I am using and I've tried to use something relatively close, but I am still unclear as to how to properly edit the file for this workflow.  I'm also confused as to how to then create the curve from this file.  I have not had luck dropping it onto the Drop-Quad-Profile droplet.  If I simply place the text file into the appropriate folder and run the install command it will create the curve and I am able to use it, but I'm not sure this is the preferred method.  
> 
> Any insight would be really helpful.
> 
> Thanks very much.
> Andy

Dana and I coming up with a tutorial and video on how to properly use Roy’s QTR linearizer for Piezography. In recent testing, to get the same linear character of the Piezo quads with Roy’s linearizer, it does require a slightly different target that has been encoded with a specific gamma so stay tuned.

I have beta (!!! very beta) targets and workflows (colorport/i1Profiler) here: https://drive.google.com/open?id=0B4gHbc7Tg9-SOWhsS01BVFJmUGs <https://drive.google.com/open?id=0B4gHbc7Tg9-SOWhsS01BVFJmUGs>

1. There is no documentation yet so please wait for that if you aren’t familiar with using ColorPort or i1Profiler targets and workflows.
2. These target files are for Piezography. They won’t exactly work for everyone else using QTR.
3. These target files are not the averaging targets (aka, 21x4) yet. 
4. We’ve had some trouble getting a cgats output to be properly read by the linearizer. I’ve included a sample cgats file that works in the folder above. The L*a*b* values can be swapped out with the measured lab values from your print as long as you properly trim the row count and make sure END_DATA is after the last measurement number.

///

5. Making master curves for Piezography is a whole other ballgame. Piezography requires very finely tuned under-printing that is variable and specific to each ink shade and printer type and media. It’s literally a dark art. I have been working on a set of master curves that are kind’of a new take on curve building, but that is all I can say at this point because they are so early in their development stage. Dana is also hard-at-work on a set of master curves that enable dual K4 and GO printing. Again, this is a multi-year long effort to make sets of piezography master curves. It requires thousands of prints and is very manual.

cheers all,
Walker

> On Apr 7, 2016, at 9:29 AM, artodd138@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:
> 
> Hi Everyone,
> 
> 
> I've just begun working with QTR and I'm having a bit of trouble creating my own custom curves.  I've been working through the documentation that comes with QTR, however I fear that some of the tutorials are slightly dated.  I was hoping someone on here could direct me to more recent tutorials on the process.  Below I've outlined my setup and where I hit roadblocks.
> 
> I'm working on a Mac 10.11.3, Photoshop CC 2015, QTR v2.7.7, X-rite i1 Pro and i1 Profiler, an Epson 7800, and Piezography Special Edition P2 inkset.  I'm able to install QTR and print with Jon Cone's P2 curves just fine.  I want to try creating custom QTR curves for papers without Piezo curves available.  Through the tutorials provided with QTR I am able to print the calibration sheet, determine default ink limits and crossover values for the 6 shades.  Where I run into trouble is placing these values into the text file and then creating the curve.  Supplied with QTR there are no ink descriptor files that are close to the set up I am using and I've tried to use something relatively close, but I am still unclear as to how to properly edit the file for this workflow.  I'm also confused as to how to then create the curve from this file.  I have not had luck dropping it onto the Drop-Quad-Profile droplet.  If I simply place the text file into the appropriate folder and run the install command it will create the curve and I am able to use it, but I'm not sure this is the preferred method.  
> 
> Any insight would be really helpful.
> 
> Thanks very much.
> Andy
> 
> 
> 
>

Hi Andy, I am making a whole series of videos to go along with my profiling tools that will be part of my up coming QTR book and course.

If you want to get started faster, I offer a one on one workshop and tools for QTR profiling up to 6 inks with i1 Profiler. This takes all the math and guess work out of the process and will get you a perfect set of curves with  just one sheet of paper. I've done a number of these one on one sessions over the last few months and have the whole thing down to a three hours. Here is a link on my page with some info. This goes far beyond just linearizing an existing set of curves with with the linearize-quad app, and will allow you to see what your printer is doing at each step of the process (and tonal scale).

http://www.bwmastery.com/store/remote-quadtonerip-workshop

I forgot to mention this. If you are building the profiles from the ground up you can skip the linearize= line and use one of these correction curve tools instead. Other people have found this to be better than the linearize function in the QTR app:


http://www.bwmastery.com/qtr-correction-curve-tool

Hope that helps,
Richard Boutwell

I’ve been working on getting more number data (rows) into the linearizer (129 at the moment) to get it to really sing. It requires some interesting number stuff to say the least . . . I’ll post on this if I can get it good.

Re Piezo curves vs other curves, really it’s all in the eye of the beholder. I’ve been a professional printmaker for well over a decade and have progressed from dual quad on a 9000 using (ImagePrint can you believe it?) all the way to really crazy K7 and K8 and K9 (alpha channel) setups that I built at Black Point Editions using StudioPrint (and not much underprinting compared to current Piezo curves.) I really liked how studioprint (working on a curve layout built by Jon) lacked much under-print. This enabled myself and Tyler and a few of the other professional labs out there to really get distinct hue-splits on certain inksets while maintaining either dual-quad or K7 setups. But with low underprinting, you get extremely damaged prints with a single nozzle drop half-way through the print. You see it! Also you have to relin every time you do a big print-job or when simple environmental or paper-batches change.

That said, I’ve done a ton descriptor file work for K3, and K4 inks with color toning. I was never happy with the matte printing but did a bunch of semi-gloss papers with custom .quads for the 11880 and was very happy with them. My descriptors were so different from what the norm in this community is that I never really published them . . . I guess I didn’t have much time to document and didn’t want the questions when I was busy just printing for people. I can get back in there if anyone here wants 11880 quads . . . The 9900s work pretty well with the 11880 too . . .

Every curve does something unique. Now that I’m in the belly of it I can describe a bit of how piezography curves are built. What we do with Piezography QTR curves is take all the variables that effect a print, paper variables, environmental variables, ink chemistry, ink hue and physical characteristics, head and nozzle characteristics and physics, and we build something that can work on a bunch of different copies of the same printer model for a long time and print smooth with minimal noise and maximum frequency/nozzles. We calibrate that curve in such a way that it’s mathematically smooth, meaning when another person goes and does a linearization on their own system it will work and they can physically print 256 separations of gray regardless of their linearity.* There’s a note there: someone linearizing a piezo quad has to first put good unexpired non-sedimented inks in there and have them in the right place and have whatever other ink they may have had in, flushed out properly. The other thing we do in Piezo-land is enforce a gray gamma 2.2 workflow from monitor though print . . . but there is no reason why someone can’t calibrate a piezo .quad for some other gamma regime . . .

cheers and happy Thursday,
Walker

> On Apr 7, 2016, at 6:34 PM, brian_downunda@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:
> 
> The short version is that you can't create your own Piezography curve for an paper that doesn't have one - only IJM can do that.  The best that you can do is take an existing Piezography curve for a similar paper and using your i1 and Roy's relinearisation droplet, relinearise that curve for the new paper.  Details on how to do this with current technology:  http://www.cyberhalides.com/piezography-printing/re-linearising-a-piezo-curve/ <http://www.cyberhalides.com/piezography-printing/re-linearising-a-piezo-curve/>
>  
> Walker is developing a new and improved way of using the relinearisation droplet, but the existing approach is working pretty darn well for those of use who have been using it since it was released by Roy last year.
> 
> Or you could create your own QTR curve from scratch, using the standard QTR documentation, or the service that Richard is offering.  However strictly speaking this won't be a Piezography curve.  How much better or worse or simply different would this be to a custom or relinearised Piezography curve?  I suspect that everyone one here will have a different answer to this question.  IJM will claim that Piezography is a system, and their particular style of curves are part of that system and part of the style of print that it generates.
> 
>

I actually do not have a starting point for K6/P2 because my (admittedly shortly-experimented) results and endeavors with qidf on more than K4 were not quality-equivalent to the piezo quads or studioprint (what I was running at the time). I think there’s a lot of people that do make K6 qidfs though . . .

However, for the rest of the people following this thread, our P2 archive is here: http://www.inkjetmall.com/tech/content.php?165-Piezography-2-Matte-and-Gloss-for-K3-printer-models <http://www.inkjetmall.com/tech/content.php?165-Piezography-2-Matte-and-Gloss-for-K3-printer-models> (I know, it’s impossible to find our quads. I’m working on it .)

I do realize that there are a few spectrums missing from the above list that disable people from linearizing for related paper types: high gloss and low-ink-load baryta papers..

Again, working on it. However, there is a possibility to linearize the P2-X800-X880-WN-HANptoRagBaryta quad with the droplet for glossier papers . . . . . you’ll need a spectro . . .

best,
Walker

> On Apr 7, 2016, at 11:39 AM, artodd138@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:
> 
> Thanks Walker - do you know of any sample QTR ink descriptor files that you could share as a starting point when using P2 inksets?
> 
>

For me as a printmaker for artists it has always come back to the inks regardless of what output system I’ve used (and I’ve used 5 different systems for printing Piezography ink so far).

Piezography curves are built for piezography inks and a specific standard of tonal reproduction that we all sort-of define as “Piezography". There is a real complex relationship between how the curves are made and the chemistry at play within the inks that I’m only starting to realize now that I work here but that I was a beneficiary of for over a decade. That really is what makes a piezography print: the inks and their underlying relationship per tonal value. Remember, Piezography has gone through many iterations: Iris Quad, Roland (whateverK# you want), image print dual quad, iQuad, bw ICC, Studioprint: k6, dual quad, k7, k8(with some people doing dual GO), QTR k6/k7,upcoming dualQuad. In each of these cases Jon and the team has made a system capable of creating an image that lives up to the standard set out for Piezography. It’s not just QTR . . . it’s an artistic ethos that starts at the ink and goes through the multiple generations of piezo print-head technology. It’s been around for a long time. This company started putting pigment ink into epson printers before epson was, and was doing inkjet before epson came out with the first home-use photo-quality printers. Even back then there was monochrome ink overlaps happening that related to ink chemistry and that started to define the parameters of a “Piezography” print. I think people forget the long history of how Piezography came to be defined.

QuadtoneRIP is certainly the most flexible system and has the potential when refined specifically for what ink is being used to attain a level of quality that is on-par with (or higher than) previous more closed systems. Certainly we’ve discovered this in our recent invention of a profiled negative system that finally attains that needed level of tonal reproduction. :) my two cents.

best,
Walker

> On Apr 8, 2016, at 10:05 AM, brian_downunda@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:
> 
> It's rather late here and I have an overly big day tomorrow.  I just wanted to say that I didn't think that we were arguing.  I was just contributing an additional perspective, which was what I've read Jon say about the reasons for the uniqueness of his system.  Ultimately the beauty of it is in the eye of the beholder of the prints, or not as the case may be.  As I note on my blog, opinion is divided.
> 
> 
> 
> ---In QuadtoneRIP@yahoogroups.com <mailto:QuadtoneRIP@yahoogroups.com>, <richard@...> wrote :
> 
> Brian, I am not arguing with you here. Those were just some mostly rhetorical questions to point out that that Piezography isn't just one thing. It is a whole system. 
> 
>

Dear list. This is related to the thread and pertains to linearizing for darkroom negative printing (something Richard just mentioned.)

We have built a system we are calling PiezoDN for Piezography using QTR that lets you linearize a negative .quad yourself directly from darkroom-printed patches using Roy’s linearizer. This is also then accompanied by a Create-ICC grayscale profile that does a final Perceptual linearization of the negative to match contrast between screen and print (optional).

We are calling this system PiezoDN and will be selling install packages for the curves/iccs starting with curves for Palladium and Potassium Oxalate printing (on BerggerCOT320,HahnemuhlePatinum,ArchesPatine,LegionRevere) but eventually hope to calibrate the history of analogue from silver/salt/cyanotype/gravure/etc. Launching for 1430/3880/4900 printers in the coming weeks.

What we have done is build a brand-new internal profiler and master curve builder that makes master curves for each alt-process to the desired contrast ratio (dmin to dmax) required by the light-sensative material. Our resulting master curves are so linear when printed with K6 and GO that they are essentially ready to go. The linearity is archived by a relationship between the ink chemistry and dMIN curve shape and dMAX curve shape + underprinting characteristics that went through about 700 iterative changes to achieve dMAX and paper-white dMIN for any given process/printer using a single sheet of OHP UltraPremium film.

 If different darkroom chemistry is being used, these curves can be simply tuned with Roy’s linearizer and then iteratively tuned/profiled with Roy’s create-icc.** In most cases one only has to build a create-icc or simply use ours if their chemistry and workflow match what we had during the R&D of the curve/profile. For brand-new processes, PDN works perfectly still to do photoshop limiting of the density but we can also build custom limited and linearized 256 patch masters directly off customer’s alt-process work if the negatives they are printing are too dense for their process. Or, a simple straight-line Photoshop curve can be applied to a target patch set and then applied to all images printed from the resulting calibrated .quad or ICC (Both PDN and straight-line limiting diminishes bit-depth in the image. This is partly why we built PiezoDN.)

The smooth piezography master curves for negative printing were calculated at 2x to 4x the bit depth of normal .quad linearizing (and then iteratively refined with 256 step prints.) We have spent nearly a year building and fine-tuning the internal profiler and master curve maker that lies at the core of this new negative-printing process and adds to the long history of Piezography and fine-art printing.

This system uses shades 1-6 (no longer needed are shades 2.5 and 4.5). For pro printers, it also does dual Gloss Optimizer printing (at 10% to 20% or so) for added negative durability (very important).

We will be launching this system soon, but it’s working in-lab currently for Palladium and we are teaching it starting this spring and throughout the summer.

This is a small screen-shot that I did today of a linear Palladium step print result graph that we are using to build a 129 patch Create-ICC profile: facebook dot com /greendirt/posts/10154141435327220

cheers and happy darkroom v2 printing,
Walker

**as long as one has an i1 or other spectro.

> On Apr 14, 2016, at 10:20 AM, richard@richardboutwell.com [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:
> 
> The at is partially correct. The curve creation program that is run when "installing" an ink descriptor file is a black box. All the partitioning formulas and gray curve adjustments can't be seen because they are all just executable binaries that are called at different times when installing the profile. I guess if you were a hacker type you could get them out of there some how...
> 
> 
> Edibility of quad curves: there have been a number of threads over the last few years that have gone into this, and a few people have worked it out with different kinds of spreadsheet methods. All of that was not for the feign o heart and was before Roy released the linearize-quad app. That works well for a lot of users, and is the first thing that should be used. 
> 
> 
> All that being said, for the past few weeks I have been working on a way of taking the individual quad curves for a linearized profile and parsing them into 21 steps so I can manually adjust the shape of the ink curves in Excel or Photoshop to get smoother shoulders and long trailing edges (while still maintaining the same ink load). I then put them back into an new ink descriptor file as Curve_ink="0;0 5;25 ... 100,0 " or ACV curves made with PS (and then change the limits to 100 for each of the inks used so the ink load in each of the curve stays the same), and then using my own linearization tools to fine tune the final linearization. I have all that figured out for inkjet positives on paper, but doing the same for digital negatives is what has me hung up now (well, 21 steps is easy, but refining with 51 to 218 steps is a little messy still with digital negatives). 
> 
> Hope that doesn't muddy the waters too much 
> 
> Richard Boutwell
> 
> 
> 
>

Does anyone know if the Epson Stylus Pro 7000 prints decent photographs without converting to all black and white?  I notice that Quadtone RIP is listed to work with it.
Thanks,
Curt

Perceptual Linearity is a bit of a misnomer. What the approach of darkening the shadows actually does is help to match the print contrast to the monitor assuming the monitor has a set DARK value that is pretty dark and light value that is pretty light and the paper has some specific dMax value that is lighter than the monitor’s dark value. So, in effect, the printed values are changed to simulate the monitor. Both monitor and print are now no-longer linear but they kinda look the same.

This perceptual non-linear icc curve outlined in the blog post linked below is dependent on dMax values. Perceived contrast is much higher when print dMax is lower (L 2.5 etc) and there is less of a need to simulate a higher contrast by pulling quarter-tone shadows down. This curve becomes stronger when dealing with dMax in the range of L 24 (Platinum prints). 

What we’ve done with Piezography (for the past decade) is two things. We’ve encoded our targets in gamma 2.2 and we’ve calibrated our monitors (good monitors) in gamma 2.2. We’ve also made sure to set the contrast ratio of the monitors to a specific target of between 180:1 and 250:1 by adjusting both the white and black levels. We also build an RGB create-icc to soft-proof with. The result of the above steps, gives an actual linear monitor image that matches to a print that is actually linear. There is no need for the ICC. The reason for using a linear workflow is that fine-adjustments of the shadows become much easier and predictable (even with monitors that aren’t super awesome).

However, there-in lies the rub. When dealing with images created by non-linear monitors not using the Piezography monitor setup described above, the shadows have been modified to print a bit too open when just printed. Often people like this, often people don’t. In our print-testing with create-icc, we’ve found that the shadows tend to block just a bit too much. They become a little un-stable to work-with. This is the case for paper-printed shadows; however, we’ve found that when using a Create-ICC profile to print negatives that are then printed in the darkroom, Create-ICC works really really well as a final perceptual match-calibration to the monitor.

As a “practitioner” of piezography in the past, I’ve used create-icc to great effect. It’s enabled me to anticipate the contrast of my client’s images and quickly match previous print processes. The tonal values of Piezography are linear, the tonal values of Create-ICC are variable based on dMax that sometimes pulls everything a bit too dark, the tonal values of ABW are a combination of the two and are also just weird sometimes (less shadow gamma is pulled down and the midtones to highlights are usually pulled just a bit up).

If someone is using the create-icc profile as-is to print with, I would suggest calibrating their monitor and workflow to match a print created using it. If someone is using the Piezography linear .quads with Gray Gamma 2.2 images, I would suggest calibrating their monitor and workflow to match a print using that. What we are working on internally at InkjetMall is a way to modify the "perceptual" linearity of either a .quad directly or a Create-ICC profile (indirectly) to enable more flexible tonal matches in real-world environments. That, however, is a hack only used when one can’t properly calibrate a monitor to a defined environment. In the flow of work, one really should decide on a print standard and keep it that way. Otherwise everything gets uber complex.

best,
Walker

> On Apr 15, 2016, at 11:18 PM, brian_downunda@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:
> 
> I've hesitated to respond to this, because this is an issue that I've struggled with over the years, and I'm not sure whether my view is shared around here.  At the risk of being seen as a shameless self-promoter, I posted my thoughts on this issue in another blog post <http://www.cyberhalides.com/piezography-printing/the-piezography-heretic-to-convert-or-not-to-convert/>, written to answer a slight different question, but covering essentially the same material as your question.
> 
> If you've already seen this post, there is some new material that I inserted today in light of your question, particularly what happens if you convert the standard 21x4 to an ICC generated by QTR-Create-ICC, print and measure.  Since Roy's objective in designing the ICC generator was to create prints that were perceptually linear, this new graph is in effect his answer.  The article also includes a link to Roy's most recent statement on perceptual linearity.
> 
> Another practitioner on this forum has hinted to me that he has developed an approach that combines the open shadows from printing in GG22 with the perceptual linearity of converting to an ICC for printing.  If I've understood this correctly and if this approach gets released, then there will be another option.
> 
> 
> 
> ---In QuadtoneRIP@yahoogroups.com <mailto:QuadtoneRIP@yahoogroups.com>, <artodd138@...> wrote :
> 
> On a side note - does anyone have a graph of L* values that would appear linear to a standard observer?  Looking quickly at the cyberhalides link and other references it seems strange that a perfectly linear response in ink density is desirable.  Even though it is measured as linear wouldn't it not necessarily be perceived as linear because of the nature of standard human vision?  What seems to make piezo curves so nice is that they appear linear, and in fact when measured and plotted are not perfectly straight; the slope increasing in the deep shadows to increase separation.  Am I incorrect in thinking that human vision is not as good at separating dark values as compared to light and therefore an increase in slope at the shadow end of a curve would be required to produce a perceptually linear grayscale?
>  
> 
>

I suggest you download and read The New Piezography Manual. It’s the first place to look for Piezography workflow related stuff, and the documentation on this is there. Some things certainly need updating, but that will come in the 2016 edition.

Regards and cheers,
Walker

> On Apr 18, 2016, at 7:36 PM, brian_downunda@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:
> 
> Having read this forum on and off for much of the time that it's been around, it's my recollection that there's more of the science of visual perception in the way that the ICC was designed and impacts on the image than simply improving the monitor to print match, hence the term used.  But finding some of those posts now is like looking for the proverbial needle in a haystack and the really old ones can't be found.  Roy can correct me if my recollection is incorrect.
> 
> Re "The result of the above steps, gives an actual linear monitor image that matches to a print that is actually linear. There is no need for the ICC".  In my experience it's true that there's less need for the ICC on high dMax gloss papers.  The impact of converting to it, or doing a preserve numbers soft proof, is subtle.  But it's still there, and if the subtleties matter to you then there's some value in it.
> 
> However I've never encountered a situation with matte papers where the ICC has no impact, no matter how well I've calibrated by Eizo monitor for print in the manner described.  For an image with shadow detail, the effect of the ICC is quite marked.  For matte papers there is the binary choice of converting to the ICC for printing and soft-proofing in the conventional manner, or not converting and using the ICC for a preserve numbers soft-proof.
> 
> This requirement to do a preserve numbers soft-proof when printing in GG22 is rarely made clear anywhere, including in IJM material.  Jon first made the point in a post a long time ago <https://groups.yahoo.com/neo/groups/piezography3000/conversations/messages/29322>, even when working with a monitor calibrated for print, but it rarely gets mentioned, and is partly the reason for my blog post.   
> 
>

Back to the thread at hand!!! (This is the important part, it’s the QTR forum after all.)

Like I said a few posts back, perceptual linearity is a misnomer so it’s very hard to graph what would appear linear because that makes assumptions about one’s monitor calibration and the dMax of the printed piece. It’s about matching to a monitor. So yes, if the dark curve makes your print match your monitor you are looking at a "perceptually linear" (aka monitor matching) print. If not (if it makes your print darker in the shadows because your monitor/workflow is calibrated differently) than it is no longer a “perceptually linear” print. Because we use a different method of calibrating the on-screen image at IJM, an ICC-printed print with pulled down shadows is now a bit dark and we’ve found reasons why the opposite (non-icc) method works better, mainly for shadow stability. Catch my drift? In this line of work, image data and monitor comes before print and the word “perceptually linear” is really a replacement for “how it looks on my screen". My observations of the underlying math at play between ICC-printed dark dMax and light dMax prints are just observations (all-be-it from 10 years of printing with Create-ICCs) so I can’t really comment on that part. It’s partly obfuscated by how Photoshop utilizes one’s monitor ICC as it is an “ICC aware” application. In the past, the way I (and probably many other people) used Piezography was printing with Create-ICC. This pull-down of the shadows worked well for how I calibrated my screen although sometimes it would make things annoyingly too dark. Pre-Create-ICC I used a combination of Gamma and Dot-Gain in studioprint to match to a gamma 2.2 calibrated 19” LacieCRT. This was (and still is) the closest monitor-to-print match I’ve ever seen but this was a LacieCRT calibrated at 2.2, an image encoded in 1.92, and the printed through a dot-gain of 19. That was a weird-ass route, but it worked.

So in the manual we lay out a method for making one’s screen match a print (that doesn’t utilize icc, just gamma 2.2 images). Included in there is how to do the soft-proofing.

best,
Walker

a note: We spend most of our time at IJM making ink, making ICCs/.quads, shipping ink, selling ink, QC’ing said ink, and taking calls from people who don’t read the manuals that (literally) come with their products. Another 10 percent is spent making new products. The final 10 percent is spent handling the over-load of documentation that has been produced over a 16 year period and also documenting new methods and procedures along with engaging a community of very passionate followers (one of them used to be me).  I was hired to do this but only recently. The person I was replacing was also running the world’s first digital fine-art print-lab (Cone Editions) at the same time as much of the stuff above. So, we are working on it. :) It’s fun. It’s still a lot of work and takes (will take) some time to do it right. Partly why I was so eager to come here and work was to tackle the very problems outlined below in your comment Brian.

upcoming:
Both the Inkjetmall and Piezography website and forum are going to be collected into one place.  (email with details about that coming soon to the IJM mailing list)
Manual is going to be re-written, re-designed, edited, with included best practice information and diy linearization workflow (thanks Roy for the droplet), and updated regularly. It is in need of this, although the information is there. I’ve gotten a ton of useful information from that manual in the past as a printmaker before joining IJM fyi. It’s very useful even as a reference! There are things in there that long-time practitioners can learn!
Piezography OS X install packages will be created with accompanying documentation. These will require QTR to be installed but will enable us to publish updated .quads very frequently in one organized place without bogging down Roy’s distribution schedule.
All old documentation (over 1000 individual urls at this point) will 301-redirect to current documentation (this is the bulk of the work fyi).
And of course more to come (like PiezoDN and Pro), cheers.

> On Apr 19, 2016, at 7:15 AM, brian_downunda@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:
> 
> You're kidding me, right?  Clearly you've settled into IJM pretty well, because RTM is the standard IJM response.  I've read and reread the ("new") manual and this stuff is not in there.  That is, not sufficiently clearly and some issues not at all.  People use Piezography despite its documentation, rather than becauseof it.  Although you dismiss the use of an ICC, the preserve numbers workflow is supported in the NPM (p66).  However the overall workflow is particularly poorly described, and it would take a forensic archaeologist to construct a concise and complete one from the NPM.  Nor is it complete - useful information bordering on essential occasionally accidentally dribbles out on the IJM forum, and you have to monitor it like a hawk to catch these gems.  The manual can only really be understood fully by people who don't need it.  I'd like to think that the 2016 NPM update is going to be better, but that's probably just blind optimism, because it's going to be written by the same folks who think that the current version is fine.  Prove me wrong.  
> 
> Coincidentally, I received an unsolicited email via the blog today, and I'll quote some of the contents without naming the author, because it summarises my views and those of others that I speak to: "Unfortunately I find InkjetMall and the Piezography web sites to be a little difficult to digest. Information can be contradictory, incomplete, scattered or just not explained as fully as I would like.  It does not help that it is an evolving process and often posts are written for a previous time, and may not apply today.  I do wish the New manual was clearer, and I suspect I am not alone."  No, that person is most definitely not alone.  Which is precisely why I created the blog.  
> 
> Getting back to the OP Andy's request for "a graph of L* values that would appear linear to a standard observer", my suggestion was to convert the standard 21x4 to an ICC generated by QTR-Create-ICC, print and measure.  I suggested this because it remains my understanding that this was the science behind QTR-Create-ICC.  Is there a better suggestion?
> 
> 
> 
> ---In QuadtoneRIP@yahoogroups.com <mailto:QuadtoneRIP@yahoogroups.com>, <forums@...> wrote :
> 
> I suggest you download and read The New Piezography Manual. It’s the first place to look for Piezography workflow related stuff, and the documentation on this is there. Some things certainly need updating, but that will come in the 2016 edition.
> 
> 
>

I have much less experience in building packages for Windows but at the vary least I will be able to build an installer that can install a folder that has all the documentation and needed curves that can be placed inside of the QuadtoneRIP profiles folder manually. This will be a little while longer than the OS X install packages. When we publish the OS X install packages, there will be a .zip download of that folder for Windows. This should simply things tremendously. We can also continually update the installer/zip and keep a change-log online.

The one problem with piezography/qtr on windows is that it does not yet allow grayscale ICC while printing, something that our new PiezoDN system will be using. 

regards,
Walker

> On Apr 19, 2016, at 11:50 AM, dtrout@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:
> 
> Walker,
> I'm reading this thread intently, as I may decide to wander into the world of Piezography and QTR for my B&W work.  (What appears to now be a permanently lost block of Green nozzles in my 7900 is making this move even more likely.)
> 
> In your previous post you mention that upcoming you plan to release "OS X" install packages.  Do you plan the same for Windows users?  And perhaps maybe you can comment on whether Piezography as a printing system really "requires" a Mac in order to make it sing.  I have QTR and have used it some on Windows, and I know there are a few differences in how things work vs. the Mac.  But my assumption up to now is that Piezography doesn't have any real dependencies on a particular OS.  If that is not the case, then I need to understand that now.
> 
> Sorry if this is a thread hi-jack.
> 
> Dave
> 
>

Grayscale ICC printing means you select a Create-ICC built ICC profile in PrintTool (OS X only), select perceptual mode, hit print, select your curve, and print. The Create-ICC profile pulls the shadows down slightly before the image hits the .quad and prints. Many people really like this way of printing.

PiezoDN will utilize CreateICC profiles for this final dark pull-down. It really seems to work great when dealing with Alt-Process printing that utilize wildly different dMax values (anywhere between light reflection of 3% to 30% in the blackest patch). However, if you are working on a PC and you have the PiezoDN curves, you will still be able to print VERY linear negatives and you will also still be able to linearize those .quad negative curves with the Quad-Linearizer droplet. By no means will the PiezoDN system not work on the PC, however, you may find that the screen-to-print match is not always as exact as you would wish with some alt-process methods of printing utilizing only the linear .quad and a GrayGamma 2.2 image. 

That said, OS X is free now, and it can be installed in VMware. It’s basically 70 bucks to get OS X running virtually inside of a Windows machine if one wants to run Print-Tool but they only have a PC.

regards,
Walker

> On Apr 19, 2016, at 12:53 PM, awidener@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:
> 
> "The one problem with piezography/qtr on windows is that it does not yet allow grayscale ICC while printing, something that our new PiezoDN system will be using."
> 
> What is grayscale ICC while printing?  
> 
> Also, does this mean that PiezoDN will not work on a PC since it's something that the new PiezoDN system will be using?
> 
>

This is a fascinating discussion. Good question Brian. I feel like it would touch off an ICC standard if the grayscale world were larger.

My (experiential only) understanding is that we see colors and tonal values in relation to each other. Darks next to other darks appear lighter than darks next to lights. In some images pulling the darks down below linear works to build a perceptually “true” (for lack of a better word) print and in some images it doesn’t. Maybe this is why I think it un-stable . . . 

//

Piezography is linear but with a big fat *. There are a lot of variables that Piezography depends on for linearity but the main one is gg2.2 data. I will share data when I get to it but anyone with a custom profile from us and our 256 patch target can validate that.

//

Related to VMware and OS X, I’ll send details off-list tomorrow when I’m awake Roy. Thanks for chiming in. 

cheers,
Walker

> On Apr 19, 2016, at 8:17 PM, brian_downunda@... [QuadtoneRIP] <QuadtoneRIP@...m> wrote:
> 
> @Roy - Back in post #23, artodd138 asked "does anyone have a graph of L* values that would appear linear to a standard observer".  This question does not mention monitor-to-print matching, but all my attempts to guess an answer to this question have been dragged off into the weeds of matching.  I'm going to make one last attempt at this. 
> 
> Let's suppose we are in a room with no monitor, so no concept of matching.  We place a piece of paper in front of a viewer with a series of unlabelled, evenly-spaced grey patches, i.e. evenly spaced luminosity values, ranging from white to black, and ask them if they are evenly spaced.  It's my understanding that they would say no.  So the question is, what series of patches would they consider evenly spaced?  Again my understanding is that if the evenly spaced luminosity patches are converted to an ICC from QTR-Create-ICC and reprinted, then the observer would probably say yes.
> 
> I seem to recall a post somewhere in the past where you (Roy) said this, but can't locate it.  Is this an accurate memory, or a manufactured, false one?  If it's accurate, then there's the answer to artodd138's abstract, conceptual question.  If my recollection is false, then I'll pull my head in and go away, in which case artodd138 will have to find the answer elsewhere.
> 
> 
> ---In QuadtoneRIP@yahoogroups.com <mailto:QuadtoneRIP@yahoogroups.com>, <roy@...> wrote :
> 
> There's a lot of pretty loose talk about linearity.
> 
>