Yahoo Groups archive

QTR-Quadtone RIP

Index last updated: 2026-04-28 23:12 UTC

Thread

Profiling a K8 monotone B&W inkset

Profiling a K8 monotone B&W inkset

2015-06-01 by roark.paul@...

I am setting up a 3880 with Eboni-6. I use 8 positions of Eboni ink -- one MK and 7 dilute positions. Some of the profiling approaches I use may have application to other similar inksets.


With many dilute carbon inks, I prefer to have at least 2 "channels" or "partitions" running at the same time. So, my "default" approach is the have a main "gray" partition using standard QTR semi-automatic partitioning. I then set up the LLK and LK, which in the Eb6 inkset I have duplicate the LC and C dilutions. (C is 30% MK; LC is 30% C [9% MK]). I can also increase the channels to 3 by splitting the main Gray partition into 2.


More overlaps can accomplish several goals. It hides the cross-overs and has more inks firing at any one point. This absorbs/hides paper and ink batch differences better, and also hides microbanding better.


Take a look at pages 7 & 8 of http://www.paulroark.com/BW-Info/3880-Eboni-6.pdf to see what these types of profiles look like. Note that in these samples the basic dilute ink limit was 50. When more than one partition or channel is going to carry the load, that ink limit needs to be manually split between (or among) the "Gray" and "Toner" partition(s).


With some profiling approaches I've seen with some other high-K position B&W inksets, it appears to me the ink loads get very high. I've observed bleed in print samples that is consistent with that. Mottling is often another symptom of an excessive ink load on the paper. I tend to have the ink limits for even the dilute inks at where the MK tops out. The denser inks clog the pores of the paper first. For simplicity I often just select that MK dmax point for the ink limit of the more dilute inks, even though they can be pushed up higher.


(On the samples shown, which were for Arches, an un-coated watercolor paper, I accepted more bleed in the K position because the paper does not have a coating to get overloaded, and I can achieve a higher dmax with higher ink loads *on that paper.* So for the ink limit approach, that sample shown is not representative of a coated paper.)


FWIW,


Paul

PaulRoark.com -- Paul Roark's Photographic Home

Re: Profiling a K8 monotone B&W inkset

2015-06-03 by richard@...

Hi Paul. I just took a quick look. Your k boost/limit seems crazy high. Every paper and printer I have tested with these inks shows no increase in the MK channel after about 40% and will eventually start to reverse as the limit gets higher. Have you noticed that with these profiles you are making now with the 3800? How are the Lab_L measurements from the unlinearized print?

Looking at your density settings for the gray/toner inks (gray_value= on the mac) your whole number approach is not accurate enough when slicing the grayscale up with so many inks (4 or more). I don't want to seem like I am just picking on your methods here, but you can easily solve the cross over point banding problem by setting them more accurately to begin with (by using an interpolation method mentioned in one of the QTR pdfs floating around—I'm detailing it better in my writing, and providing a easy workflow and worksheet for it). Also, trying to hide the banding with more inks is an option, but the same can be done with fewer channels and increasing the overlap/gamma settings. See my post here: http://www.bwmastery.com/blog/2015/in-search-of-the-perfect-qtr-profile

I am all for pushing what is possible with ink jet printing, but trying to solve something with more complexity where it can easily be solved by being more careful might be setting people back instead of pushing printing forward.

Richard Boutwell

Re: [QuadtoneRIP] Re: Profiling a K8 monotone B&W inkset

2015-06-03 by Paul Roark

Richard,

That was Arches. That's why the high K limit.

It all works just fine. No problems here.

Paul
Show quoted textHide quoted text
On Tue, Jun 2, 2015 at 9:27 PM, richard@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:

Hi Paul. I just took a quick look. Your k boost/limit seems crazy high. Every paper and printer I have tested with these inks shows no increase in the MK channel after about 40% and will eventually start to reverse as the limit gets higher. Have you noticed that with these profiles you are making now with the 3800? How are the Lab_L measurements from the unlinearized print?


Looking at your density settings for the gray/toner inks (gray_value= on the mac) your whole number approach is not accurate enough when slicing the grayscale up with so many inks (4 or more). I don't want to seem like I am just picking on your methods here, but you can easily solve the cross over point banding problem by setting them more accurately to begin with (by using an interpolation method mentioned in one of the QTR pdfs floating around—I'm detailing it better in my writing, and providing a easy workflow and worksheet for it). Also, trying to hide the banding with more inks is an option, but the same can be done with fewer channels and increasing the overlap/gamma settings. See my post here: http://www.bwmastery.com/blog/2015/in-search-of-the-perfect-qtr-profile

I am all for pushing what is possible with ink jet printing, but trying to solve something with more complexity where it can easily be solved by being more careful might be setting people back instead of pushing printing forward.

Richard Boutwell


Re: [QuadtoneRIP] Re: Profiling a K8 monotone B&W inkset

2015-06-03 by Paul Roark

Not a problem, Richard. I only had that one set graphed. For the 3880 I'm going to be stressing and profiling mostly inkjet papers, not watercolor papers. I'm really liking that printer. With the Eb6 setup I have in it, it's hard to make a bad print even with the Epson driver. What is rather amusing is that with the 8 carbon inkset in there, even the ABW mode produces outstanding prints -- maybe better than when all the "colors" are distributed by the driver in "color" mode. I'm not sure why.

At any rate, between QTR, the Epson driver, and the Epson driver with QTR-made ICCs, the 3880 is capable of pumping out some top notch 16x20 prints or panoramas very easily.

The post was really aimed and those who might want to try doing their own profiles with the Piezo inks. The banding I was talking about was the banding near the start and end of a print where both rollers are not involved. Some may want to test some of these alternative profiling approaches to see if they help there. I can't do it all.

The mottling that has been noted might be a result of overloading the inks. I didn't want to inject myself randomly in another thread. So I thought, again, some might want to take a shot at a different profiling approach that might help with this issue. High K B&W inksets can be profiled in a number of different ways.

Basically, I am simply encouraging people to experiment with alternative approaches to QTR that might help with their inksets. I do try to keep them as simple as possible, including some shortcuts that might not be technically perfect but work very well in practice. I used to make the assumption that those who use QTR know how to do their own profiling. I was very wrong in that assumption. So, simplifying the process as much as possible is, I think, necessary to encouraging the use of this valuable tool.

Paul
Show quoted textHide quoted text
On Wed, Jun 3, 2015 at 5:20 AM, richard@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:

Paul,


I am sorry, I was looking for the paper that was being profiled in your PDF, and don't know how i missed it in your posting here.

RB


Re: Profiling a K8 monotone B&W inkset

2015-06-03 by brian_downunda@...

I don't have a 3880, but my understanding from those that do is that you don't get the microbanding in the first and last inch on the 3880. I've seen 3880 prints done in QTR and there was none. Are you saying that you had some?

Do you see any advantage in making our own profiles for Piezo inks? I'm not sure I see the point if my IJM-supplied curve is close enough to linear. Of are you suggest that we make K8 curves using our own dilutions of piezo. I'm not sure I see the point of that. I think K7 is one more than necessary, and even Jon implies this with his P2 inks.

---In QuadtoneRIP@yahoogroups.com, <roark.paul@...> wrote :

The post was really aimed and those who might want to try doing their own profiles with the Piezo inks. The banding I was talking about was the banding near the start and end of a print where both rollers are not involved.

Re: [QuadtoneRIP] Re: Profiling a K8 monotone B&W inkset

2015-06-03 by Paul Roark

I have not seen microbanding on the 3880. The 1400 family is where I see the most.

I don't use Piezo, but the comments seem to indicate that users have less ability to modify or re-linearize them. Also, I have seen significant bleed in the lettering of test strips. Some also mentioned mottling. These might all be reasons for individual users to want to do their own profiling. I, frankly, would want to be in control of my own system if I used the inks.

Paul
Show quoted textHide quoted text
On Wed, Jun 3, 2015 at 4:29 PM, brian_downunda@... [QuadtoneRIP] <QuadtoneRIP@yahoogroups.com> wrote:

I don't have a 3880, but my understanding from those that do is that you don't get the microbanding in the first and last inch on the 3880. I've seen 3880 prints done in QTR and there was none. Are you saying that you had some?


Do you see any advantage in making our own profiles for Piezo inks? I'm not sure I see the point if my IJM-supplied curve is close enough to linear. Of are you suggest that we make K8 curves using our own dilutions of piezo. I'm not sure I see the point of that. I think K7 is one more than necessary, and even Jon implies this with his P2 inks.

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

The post was really aimed and those who might want to try doing their own profiles with the Piezo inks. The banding I was talking about was the banding near the start and end of a print where both rollers are not involved.


Re: Profiling a K8 monotone B&W inkset

2015-06-04 by sanking@...

I have never seen micro-banding at the top and bottom of the page with QTR and either the Epson 3800 or 3880, or with any of the wide format professional models. I definitely get it with QTR and the Epson R3000, and see it in a friend's work with the Epson Artisan 1400. I do not get the micro-banding with the Epson driver and the R3000.

Sandy

Re: [QuadtoneRIP] Re: Profiling a K8 monotone B&W inkset

2015-06-04 by asharpe@...

So, out of all the folks who say they don't see microbanding on these
printers, are any of them using black-only workflow (single ink)? That is
where the 1400/1430 has difficulty, and especially the 1430, where I have
to lose at least a full inch on the leading and trailing edge using a
single ink to eliminate the microbanding. On the 1400, I could get away
with 1/2".

Andrew
http://andrewsharpe.com
Show quoted textHide quoted text
On Wed, June 3, 2015 5:03 pm, sanking@... [QuadtoneRIP] wrote:
> I have never seen micro-banding at the top and bottom of the page with
> QTR and either the Epson 3800 or 3880, or with any of the wide format
> professional models. I definitely get it with QTR and the Epson R3000,
> and see it in a friend's work with the Epson Artisan 1400. I do not get
> the micro-banding with the Epson driver and the R3000.
>
> Sandy

Re: [QuadtoneRIP] Re: Profiling a K8 monotone B&W inkset

2015-06-04 by richard@...

There are two different kinds of banding that are being discussed here.

1: With the small format 1.5 pl printers the first and last 1" or the page are unsuable because, as Roy and others have said, the dithering and micro weaving needed where the paper stays "in place" under the print head is different from when it is passing through the printer. This ink firing pattern is built into the Epson driver on the computer on the smaller desktop printers and not into QTR. There are some complicated controls going on behind the scenes to control what nozzle is firing in what order... The larger format printers like the 3800 and up have that weaving function built into the firmware of the printer and you can print right up to the edge of the paper with no micro banding (i think the 2880 might be free of the micro banding as well, but don't quote me on that).

2: The kind of banding that I was referring to shows up in smooth gradients as horizontal bands where the inks from one shade overlap the inks of the other shades. This is the banding that can be hidden with the additional inks in the toner sections of the ink descriptor file. I have done this using a different color than the main color of the inkset, like a 4 or 5 carbon profile with two shades of selenium as a toner.

This kind of banding comes from how QTR uses a partitioning function based on the cross over points of each ink (the relative dmax for each shade of gray where it intersects on the K curve). If these cross over points are not set accurately or if there is not enough overlap built in from one shade to the next there will be noticeable places there are flat spots or small reversals in smooth gradients. With 4-7 ink profiles I do not consider whole numbers or estimates between the 5% steps to be accurate enough.

I am not sure if using all 8 inks with carbon provides any advantage over a well made 6 gray carbon profile. I can see where it might be better to hide the dots with more inks and lower ink limits, but with the inks as dilute as they are, you will be hard pressed to see any dot when using more than 4 black/gray inks. I think the best option for utilizing all 8 channels would be to made a dual quad printer and blend the two profiles at print time using QTRs multiple curve printing option.

As for the benefit of making your own QTR profiles with the Cone inks: I personally think the Cone carbon has a better color dilution than the Eboni Carbon inks—I know other people feel differently, but that comes down to personal preference. I prefer Jons way of partitioning the inks starting with bezier curves, but it does require using his gray dilution percentages to work properly. I also like that I can more or less swap out one tint to the next without the prints changing drastically. If the Peizography inks and profiles are working properly then I don't see an advantage of making your own QTR profiles with those inks. The shade 7-4 ink curves are much better with Piezography than with the increased overlap settings when made with the QTR tools. The QTR workflow is definitely more demanding, but can be just as good. It is only when Piezography don't work correctly that it is nice to be able to take over and continue to make equally beautiful prints.

Richard Boutwell
Blog



Re: [QuadtoneRIP] Re: Profiling a K8 monotone B&W inkset

2015-06-04 by asharpe@...

It also begs the question that, if the mechanism for printing at the
leading and trailing edges are handled by QTR for the lower end printers,
then why, when printing with the same profiles and the same QTR, would the
banding be worse on a 1430 than a 1400, both at the highest resolution
(2880/unidirectional)?

Andrew
Show quoted textHide quoted text
On Thu, June 4, 2015 1:16 am, brian_downunda@... [QuadtoneRIP] wrote:
> I thought that this brief post by Roy dealt with the issue of the
> microbanding rather succinctly:
> https://groups.yahoo.com/neo/groups/QuadtoneRIP/conversations/messages/12
> 758
> https://groups.yahoo.com/neo/groups/QuadtoneRIP/conversations/messages/12
> 758
> Give this, I'd very surprised if the R2880 didn't suffer from it.

Re: Profiling a K8 monotone B&W inkset

2015-06-05 by brian_downunda@...

You are referring to microbanding in the first and last inch? If so then I don't know. I find that it varies from image to image, or at least it appears to vary. It's not an exact fixed pattern.


---In QuadtoneRIP@yahoogroups.com, <asharpe@...> wrote :

It also begs the question that, if the mechanism for printing at the
leading and trailing edges are handled by QTR for the lower end printers,
then why, when printing with the same profiles and the same QTR, would the
banding be worse on a 1430 than a 1400, both at the highest resolution
(2880/unidirectional)?

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.