P2 and the R1900/R2000

piezography

#1

On the Piezography 2 Matte and Gloss for K3 printer models page there is the following statement: “[I]Piezography 2 is NOT available for non-K3 models, including the 1400/1430, 2200, R1800/R800, R1900, R2000, or 7600/9600[/I]”. This has always disappointed me. I’ve never really been convinced about the value of shade 7, and as stated in another post today I find the MK/PK swap unnecessarily wasteful, given the associated head cleans.

Now I can’t speak for all the other printers, but I have discovered that this is [B]NOT[/B] true for the R1900 and therefore I assume not for the R2000 either. These hi-gloss printers are not K3, but they in fact heavily disguised K3 printers. They use the same DX5 printhead as the R2880, from what I can gather, and some of the larger and older printers K3 also use some sort of DX5. The R1900 is virtually identical to the R2880 physically, so it was surprising to me that P2 doesn’t suit the R1900/R2000.

It is true that the piezo QTR curves are different for the R1900 - the shades are in a different order. So as an experiment I remapped the P2 SE x880 curves so that the shades appeared in the correct places in the quad files for the R1900, and checked how linear the result was. Here is the result for HPR:


I have to say that I was staggered. This is better than the standard curve for HPR performs on my 4½ year old 1900, which has a certain amount of drift. Roy’s new droplet pulls it into a straight line. I’ve converted all the eleven available special edition P2 curves, and for those papers that I have, the results are similar, although not quite that good. I’ve also used the HPR curve for EEM, and HPR-Baryta for IGFS, and again got similar results. I’ve also converted all the neutral P2 profiles for Jeff’s new P2000.

It seems to me that P2 [B]IS[/B] available for the R1900, once the curves are remapped. I’ve done 40% of then - could IJM do the other 60%?

Why bother? Well, as I said elsewhere, I am far from convinced that shade 7 is necessary for most people, and I have a strong aversion to the waste arising from the MK/PK swap.

There is another side benefit to this discovery. One problem with the R1900/R2000 is that the range of available curves is more limited than for the x880 printers. IJM has been reluctant to create more, as the work wouldn’t be justified for such a niche printer. But this isn’t a problem it seems - the x880 curves can be remapped to the R1900. Problem solved.

I haven’t attached the 22 remapped curves, because it isn’t clear whether IJM would welcome that. But I am happy to if they would, and I’m happy to provide them to anyone on request.


Piezo Digital Negatives on R1900
#2

Brian,

Piezography 2 is offered by InkjetMall for those who have printers that require a black ink change involving the moving up and down of ink levers and the wasting of ink in order to do so. The black ink change on those models takes several minutes and consumes a lot of ink - so Piezography 2 makes sense from an economic view point.

We do not market Piezography 2 for the 3880 for example because it has both blacks installed at the same time and the black ink change uses only black inks rather than wasting a portion from all carts.

We do not market Piezography 2 for desktops because it takes about 30 seconds to change out the black cartridge from matte to photo black.

We did not produce a Piezography 2 for the 1400/1430 because it would require reducing it to a 4 black system in order to make room for both the GO and the PK. The 7600 would require a 5 black ink system in order to have enough room for the PK and GO. We try to market Piezography as a highest standard product, but you can certainly lower the amount of grays that are printed if that better suits you. But, we are not in that same position because we are trying to promote the concept of highest standards. In the end, we think its best to run Piezography at the highest available amount of black inks. So you have more flexibility than we do, although I do hope that you appreciate our efforts. And I actually run some funky ink setups on my own systems - but it would be very difficult to market tons of funky systems and then provide support and instructions for them all. So we stick to the highest possible standards in what we offer.

I am sure that if there are other R1900 or R2000 or R1800 users who would like to switch to Piezography2 they will greatly appreciate your efforts and you can certainly post a zip of your curve library here. I would only ask you to use a different nomenclature for your curves than we do. Thanks for your efforts!

Ultimately - we use this section of the forum as a technical support board for our products - and perhaps its better to move this thread into the discussions areas and get that more active! We welcome your insights.

Best,

jon


#3

I have used names like [I]P2-1900-SpecEd-HANptoRag.quad[/I] - is that sufficiently different while conforming to the overall naming schema?


#4

No, please use your initials or something.
Perhaps [I]BS-[/I][I]P2-1900-SpecEd-HANptoRag.quad
[/I]
That way we can quickly differentiate in a technical support issue.

thanks!


#5

I’m happy to name as you suggested, although they’re not my curves, or my relinearisations for my printer, they’re just your curves remapped. The only creative content of my part is the remapping, which you could just as easily do yourself and label them as your R1900 curves.

[I have a speadsheet that I created to automate the remapping process using formulas. I am happy to share that if anyone wants to remap the other three inksets.]

Can I just clarify, where should I upload? It’s your support forum and I can put them wherever you prefer. Personally I’d have thought that “QTR Curves and Workflow” would be the obvious place to find and discuss curves, but you moved this thread. I could just attach them here, or I could put them in “QTR Curves and Workflow”. You may prefer to keep that section of the forum as a technical support board for your products, but my point remains that in effect these are your curves.

I didn’t mean to suggest that P2 as an option for the 14x0. Not on a six cart printer. Also it clearly doesn’t have a DX5 printhead with two channels cut off. I tried remapping the 1430 K6 curves to see if they could be used as the basis for the R1900 P2 matte curves and they can’t. The numbers in the quad file are far too low. I’m not familiar with 7600, but there are similar issues and its K6 curves don’t migrate either. The R1900 & R2000 are special cases, and should be grouped with the x880 printers.

I replied about ink waste in the other thread. Since the advent of chip resetters for these carts which allow me to reset all carts at refill, I get very little printer-generated ink waste. Almost none. Regular MP/PK swaps would increase it substantially.


#6

What you are offering is not a technical support topic - and we need to keep technical support for technical support issues so that we can quickly help our users work through problems.

What you are suggesting is an experimental practice of sorts - or a custom procedure - and you are offering to produce something that we do not directly offer support for. I believe that the best thing to do is to provide the suggestion and method and your spreadsheet - rather than the actual curves themselves. This way - users learn something practical - and we do not add confusion to what are official Piezography curves. The spreadsheet will be really good for users that wish to do this for themselves - and you will help others to produce a custom setup of their own.

We could rename this thread if you like to Creating Piezography2 curves for Non-supported Printers. And you should be able to attach the zipped spreadsheet directly to a post in it. But, we could also open a new type of technical articles section written by users in the Articles section of the board. Right now we have Product Manuals and Instructions and we have Frequently Asked Questions. I’m open to suggestions.

FWIW: You still get about 30-35% waste if you operate an Epson printer - regardless of when you reset chips. The printer primes the heads constantly and eject ink to keep them moist. If you install a waste ink bottle, you will cry. :frowning: It’s amazing to see how much ink builds up over the course of a printing day without your ever doing a single head cleaning.


#7

[QUOTE=jon;8787]FWIW: You still get about 30-35% waste if you operate an Epson printer - regardless of when you reset chips. The printer primes the heads constantly and eject ink to keep them moist. If you install a waste ink bottle, you will cry. :frowning: It’s amazing to see how much ink builds up over the course of a printing day without your ever doing a single head cleaning.[/QUOTE]

I’ll reply to the rest of this post once I’ve had a chance to do some user testing of the files that I intend to post.

In regard to the 30-35% figure, as explained in my post in the parallel thread, that’s not the case here with the R1900. I most certainly have waste ink bottles on all my non-Pro printers. For the R1900, I fill and reset the carts, empty the waste bottle, and when it’s time to refill there’s not much in that bottle other than the ink from initial head clean, and usually that’s not much. I might try and measure this for the next cycle. I think the percentage would be in single figures. I guess it’s possible that this is one aspect of the R1900 where it differs from the K3 printers. The real killer in my experience is the ARC chips resetting sequentially if you let them. And lots of MK/PK swaps.


#8

User testing is complete. The spreadsheet is attached.

In line with Jon’s request, I haven’t attached the remapped curves. Personally I don’t see the point of not attaching them, since anyone looking to do the migration will just want the curves and will not want to have to go through a technical learning and remapping exercise in order to do so, as even with the tool, it’s tedious. That’s no way to encourage piezo uptake. [U]So I have put a zip of the remapped and renamed curves on my blog and am happy to PM the link to anyone who asks.[/U]

Ostensibly the attached spreadsheet tool is to take an x880 P2 curve and remap it to an R1900/R2000 P2 curve, but as I said towards the end of my initial post, you can also use this tool to remap an x880 K7 curve to an R1900/R2000 curve. So should you need a K7 curve for the R1900/R2000 for a particular paper and there isn’t one but there is an x880 curve, you can use it for that as well. So for that reason, having the spreadsheet here is useful.

I wanted to disagree with a couple of things Jon said. I don’t regard what I’ve done as [I]experimental[/I]. A discovery perhaps, but not experimental. Ink mixing is experimental. Remapping is not experimental. IJM offers it as a commercial service for printers with a few but not too many irredeemably clogged nozzles. Remapping for the R1900/R2000 is new. But IMHO all I’ve done is demonstrate that [I]these two printers really are part of the x880 family[/I], as far as peizo is concerned, and should be treated as such.

And that’s my recommendation to IJM - stop treating them as outliers, bring them in from the cold and into the warmth of the x880 family and promote them. You didn’t sell many R2000 carts, but that’s not surprising the way that it was marketed, or not, for piezo. You can still change that. The R2000 is the last printer that you can still buy new that has carts-on-head, so you can swap inksets and hibernate cheaply, and pigment agitation is much less of a problem. You’re always talking up the R2880, but this printer for all practical piezo purposes is the same, can use the same curves (once remapped), and you can still buy it new! Why bother buying up reconditioned R2880s, when you can have new R2000s??? It’s the same printer, but with inks in different positions and different chips on the carts. (Actually, the R1900 is the same and the R2000 is only a minor upgrade.)

And as for [I]technical support[/I], I’m telling people how to use your inks and your curves in cartridges that you still sell, in ways that are currently possible on a x880 printer, but until now haven’t been for the R1900/R2000. If that’s not technical support … But I think we’ve crowded enough angels onto the head of this particular pin.

Zip of the spreadsheet tool: Curve Migration 3880-P2 to 1900 and 2000-P2 - NU-HANptoRag.zip (59.1 KB)


#9

[QUOTE=Brian_S;8804]And that’s my recommendation to IJM - stop treating them as outliers, bring them in from the cold and into the warmth of the x880 family and promote them. You didn’t sell many R2000 carts, but that’s not surprising the way that it was marketed, or not, for piezo. You can still change that. The R2000 is the last printer that you can still buy new that has carts-on-head, so you can swap inksets and hibernate cheaply, and pigment agitation is much less of a problem. You’re always talking up the R2880, but this printer for all practical piezo purposes is the same, can use the same curves (once remapped), and you can still buy it new! Why bother buying up reconditioned R2880s, when you can have new R2000s??? It’s the same printer, but with inks in different positions and different chips on the carts. (Actually, the R1900 is the same and the R2000 is only a minor upgrade.)
Curve Migration 3880-P2 to 1900 and 2000-P2 - NU-HANptoRag.zip (59.1 KB)[/QUOTE]

Loud and Clear! :slight_smile:


#10

It’s 2016 now! :slight_smile:

With fewer R2880’s out there, isn’t it time?

Thanks to Brian’s work, I think I’m about to pull the trigger on a refurbished R2000.
Larry


#11

In the 19 Dec 2016 IJM newsletter, underneath a photo of a smiling Dana, there is this text: “[I]Dana at Cone Editions Press printing final targets on an Epson R3000. R2000 were completed earlier signalling support for that underrated model![/I]”. Out of interest, were the Pro curves for the R2000 created from scratch or by remapping the x880 / K3 curves, and if from scratch, why not by remapping? As a number of us have demonstrated quite clearly, remapped curves are fine, and in a number of cases better (more linear) on our printers that the stock curves. I’d have thought that remapping and checking linearity would be simpler and faster than building all the curves from scratch, especially when you’re all under so much pressure.

I’m pleased that the underrated R2000 is getting some love. A little over a year ago, Jon said that “P400 is already on our radar!”. Any update on this?


#12

With Pro curves we are seeing linearity differences between R2000 and 3880 and 3880 and R3000 . . . This is on [I]​new R2000 and new 3880.[/I]

best,
Walker


#13

[QUOTE=walkerblackwell;12458]With Pro curves we are seeing linearity differences between R2000 and 3880 and 3880 and R3000 . . . This is on [I]​new R2000 and new 3880.[/I][/QUOTE]

This comment intrigues me. I assume what this means is that the ink placement for Pro inks in the R2000 has been specified such that the x880 K3 curves can be used [I]without[/I] remapping. If so, smart move!

Given that the R2000 is not a Pro printer you’d expect more sample variation than for the Pro printers, and I suspect that any such minor linearity differences could easily be relinearised. That’s certainly the experience of a number of R1900/R2000 P2 users, and I can’t imagine that Pro inks would be any different. To be honest, I’d be (pleasantly) surprised if a linear 3880 curve was exactly linear on an R2000, given sample variation, but it should be pretty close.


#14

When Walker says he sees linearity differences, I have to ask how much. I was able to relinearize existing papers such as EEM and HPR as they were very close to being linear. I have failed to relinearize Aurora Art Natural. I have now forgotten what curve I used by now, as I tried two. However, the non-linear curve shown here was too far out to relinearize using Roy’s Linearize-Quad.exe program. I assume I’d have to have this one linearized by IJM. I’m not sure if the Pro tools I could buy from IJM would work on this or not. Aurora Art Natural is not an expensive paper, and it’s not worth spending a lot of time, effort and money on.

The point I’m trying to make is that depending on how non-linear the differences are that Walker’s talking about, they may or may not relinearize.



#15

I think the issue of using a curve for one paper to create a curve for another paper is a separate issue, Larry. This thread is about taking a curve for a given paper on a K3 printer and reusing it for that same paper on an R2000.