HP B9180 for Digital Negatives for Alt

[sanking]

All of the current Epson printers have lower UV blocking than some of their previous models, which limits somewhat their use in alternative printing. Some alternative processes, albumen and salted paper for example, have contrast needs in terms of negative DR that are greater than what can be delivered by printers like the 1800, 2400 and 3800. I recently tested the 1400, a dye based printer, and the UV blocking of the colored inks was also very low.

Well, there is another option. Ted Harris, who has been testing the HP B9180 and preparing a review of it, printed out some tonal palettes and mailed them to me for testing. I now have them in hand and have made some preliminary readings of UV density, and a couple of tests with carbon printing on smooth surface papers. I am very impressed with the potential of this printer for digital negatives. The UV blocking is well beyond that of any current Epson printer, and the color palettes printed very smoothly in carbon in the density range where I would be working.

This printer would calibrate easily either with Mark Nelson's PDN or with Michael Koch-Scultze's color arrays for any needed negative DR up to log 4.0 or even higher.

I don't yet own a B9180 so I am not able at this point to carry out any further testing of it. However, based on my readings of the tonal palettes and a couple of tests for smoothness with carbon I am very impressed with the potential of this printer for digital negatives.


Sandy King

 

[Loris Medici]

Hi Sandy,

Which negative substrate did you used for the tests? I'm very interested in this printer, especially if the inks work with Ultrafine Crystal Clear...

Regards,
Loris.

All of the current Epson printers have lower UV blocking than some of their previous models, which limits somewhat their use in alternative printing. Some alternative processes, albumen and salted paper for example, have contrast needs in terms of negative DR that are greater than what can be delivered by printers like the 1800, 2400 and 3800. I recently tested the 1400, a dye based printer, and the UV blocking of the colored inks was also very low.

Well, there is another option. Ted Harris, who has been testing the HP B9180 and preparing a review of it, printed out some tonal palettes and mailed them to me for testing. I now have them in hand and have made some preliminary readings of UV density, and a couple of tests with carbon printing on smooth surface papers. I am very impressed with the potential of this printer for digital negatives. The UV blocking is well beyond that of any current Epson printer, and the color palettes printed very smoothly in carbon in the density range where I would be working.

This printer would calibrate easily either with Mark Nelson's PDN or with Michael Koch-Scultze's color arrays for any needed negative DR up to log 4.0 or even higher.

I don't yet own a B9180 so I am not able at this point to carry out any further testing of it. However, based on my readings of the tonal palettes and a couple of tests for smoothness with carbon I am very impressed with the potential of this printer for digital negatives.


Sandy King

 

[donbga]

All of the current Epson printers have lower UV blocking than some of their previous models, which limits somewhat their use in alternative printing. Some alternative processes, albumen and salted paper for example, have contrast needs in terms of negative DR that are greater than what can be delivered by printers like the 1800, 2400 and 3800. I recently tested the 1400, a dye based printer, and the UV blocking of the colored inks was also very low.

Well, there is another option. Ted Harris, who has been testing the HP B9180 and preparing a review of it, printed out some tonal palettes and mailed them to me for testing. I now have them in hand and have made some preliminary readings of UV density, and a couple of tests with carbon printing on smooth surface papers. I am very impressed with the potential of this printer for digital negatives. The UV blocking is well beyond that of any current Epson printer, and the color palettes printed very smoothly in carbon in the density range where I would be working.

This printer would calibrate easily either with Mark Nelson's PDN or with Michael Koch-Scultze's color arrays for any needed negative DR up to log 4.0 or even higher.

I don't yet own a B9180 so I am not able at this point to carry out any further testing of it. However, based on my readings of the tonal palettes and a couple of tests for smoothness with carbon I am very impressed with the potential of this printer for digital negatives.


Sandy King

Sandy,

Can you address the issue of ink grain, particularly when printing on single grade or VC silver gelatin emulsions?

Also have you perhaps tested the ink calibration image that is used with QTR?

Thanks,

Don Bryant

P.S. What Loris asked, is of interest to me along with the Pictorico White Film in regard to ink grain and substrate ink loads and durability. One of the really nice features of using QTR and other RIPs is the more precise control of ink deposition on the substrate not to mention various other benefits.

 

[sanking]

Hi Sandy,

Which negative substrate did you used for the tests? I'm very interested in this printer, especially if the inks work with Ultrafine Crystal Clear...

Regards,
Loris.

Hi Loris,

The tests were done on Pictorico. As mentioned, I emailed Ted Harris some tonal palette files, which he printed on Pictorico and mailed back to me. I measured the UV blocking density of the inks for PDN and for the color arrays that MKS has posted on this site, and when it became clear there was very good UV blocking I ran some tests with carbon using several of the colors. I find the results, in terms of smoothness and graininess, to be about the same as the Epson 2400. But what I could actually test was fairly limited and I can not comment at all about how the inks might dry on the Crystal Clear material from Photo Warehouse. I also can not comment on whether the Venetian blind problem that was a problem with the Epson 2200 and 2400 is a problem with the 9180.

Given the choice between the Epson 2400 and the HP 9180 I would go for the latter because the stronger UV blocking of the inks gives you more options from which to choose.

Sandy

 

[clay]

There tends to be an implicit assumption that the PDN color arrays give a relatively complete picture of the UV capabilities of a given printer. In fact, the PDN palette represents an extremely small subset of a myriad of ink combinations. For instance, when I ran the PDN array on my 7800, it showed a maximum UV capability of 2.4. Since then, I have gotten familiar with the capabilities of the QTR RIP,and a basic plot of the ink evaluation test shows that the Yellow and PK ink are capable of blocking up to a UV density of over 3.5 - alone! When you combine inks, it is quite feasible to create UV densities bumping upwards of 4.0. As a system for making inkjet negatives, PDN is obviously capable, but there are inherent limitations in its approach, namely:

1) It is totally dependent on the Epson driver to determine the particular color, amount, and pattern of ink is laid down for a particular input value. The user cannot control this other than by changing the input value fed to the Epson driver.

In short, it is a black box.

This became completely evident when I attempted to use the PDN color array choice to print out the step tablet. I got more UV density at 90% than at 100%! This is strictly a function of the decisions being made by the Epson driver software.

2) the color arrays in PDN are binary - they show the possible ranges in colorization possibilities with only two primary colors. There is no allowance for adding a third or fourth color, or even black, for instance.

Because these are binary (and no black) choices, it is no surprise that some combinations yield grainy results. For instance, there are many reports, confirmed by my own experience, of graininess with some color combinations. When I looked at the UV behavior of each individual ink in the QTR ink evaluation palette, it became painfully obvious why the green colors tend to be grainy on the 7800 I am now using: It is composed of two colors that are very far apart in their UV blocking abilities : Yellow and Cyan. It is no wonder when you put these dots next to one another that graininess can be seen in the higher negative density areas.

In the last three weeks, I have discovered a lot of interesting facts about making digital negatives, and have seen many of my pre-conceptions demolished by experience. I guess what I am saying is that the inkjet negative world is a whole lot more complicated than some systems (including my own) would have you believe.

 

[sanking]

I agree with you about the limitations of PDN. That is why I also tested MKS' arrays.

It is also very clear to me that the issue is a lot more complicated than any present system or logic explains. The 2200 usees the same two colors, yellow and cyan, as the 7800 to form green, and that is the smoothest of any combination I have found on the 2200.

However, we may be making this more complicated than necessary. If the purpose is to make digital negatives for pt./pd. printing there appear to be quite a number of color choices that work well, and that the limit to smoothness is not the printer but the papers themselves.

For really smooth tones of the type needed to print on smooth silver papers, I suspect that the best results will come from QTR or some other RIP.

Sandy

There tends to be an implicit assumption that the PDN color arrays give a relatively complete picture of the UV capabilities of a given printer. In fact, the PDN palette represents an extremely small subset of a myriad of ink combinations. For instance, when I ran the PDN array on my 7800, it showed a maximum UV capability of 2.4. Since then, I have gotten familiar with the capabilities of the QTR RIP,and a basic plot of the ink evaluation test shows that the Yellow and PK ink are capable of blocking up to a UV density of over 3.5 - alone! When you combine inks, it is quite feasible to create UV densities bumping upwards of 4.0. As a system for making inkjet negatives, PDN is obviously capable, but there are inherent limitations in its approach, namely:

1) It is totally dependent on the Epson driver to determine the particular color, amount, and pattern of ink is laid down for a particular input value. The user cannot control this other than by changing the input value fed to the Epson driver.

In short, it is a black box.

This became completely evident when I attempted to use the PDN color array choice to print out the step tablet. I got more UV density at 90% than at 100%! This is strictly a function of the decisions being made by the Epson driver software.

2) the color arrays in PDN are binary - they show the possible ranges in colorization possibilities with only two primary colors. There is no allowance for adding a third or fourth color, or even black, for instance.

Because these are binary (and no black) choices, it is no surprise that some combinations yield grainy results. For instance, there are many reports, confirmed by my own experience, of graininess with some color combinations. When I looked at the UV behavior of each individual ink in the QTR ink evaluation palette, it became painfully obvious why the green colors tend to be grainy on the 7800 I am now using: It is composed of two colors that are very far apart in their UV blocking abilities : Yellow and Cyan. It is no wonder when you put these dots next to one another that graininess can be seen in the higher negative density areas.

In the last three weeks, I have discovered a lot of interesting facts about making digital negatives, and have seen many of my pre-conceptions demolished by experience. I guess what I am saying is that the inkjet negative world is a whole lot more complicated than some systems (including my own) would have you believe.

 

[clay]

I agree. My old 2200 printed just fine with a green color. No grain, etc. And I hope I don't come across like I'm telling someone to use one system over another. They all have their advantages, and if one is working, then changing things just for the hell of it is sort of beside the point.

I just ran into some surprising (for me, anyway) findings with this 7800, so I am trying to throw them out there for people who either have a 7800 or are just curious about what is or may be going on. Still, the biggest aha! thing for me is the black-box nature of the Epson driver. It really gave me fits when I was using it for making negs.

I agree with you about the limitations of PDN. That is why I also tested MKS' arrays.

It is also very clear to me that the issue is a lot more complicated than any present system or logic explains. The 2200 usees the same two colors, yellow and cyan, as the 7800 to form green, and that is the smoothest of any combination I have found on the 2200.

However, we may be making this more complicated than necessary. If the purpose is to make digital negatives for pt./pd. printing there appear to be quite a number of color choices that work well, and that the limit to smoothness is not the printer but the papers themselves.

For really smooth tones of the type needed to print on smooth silver papers, I suspect that the best results will come from QTR or some other RIP.

Sandy

 

[Michael Mutmansky]

It is also very clear to me that the issue is a lot more complicated than any present system or logic explains. The 2200 usees the same two colors, yellow and cyan, as the 7800 to form green, and that is the smoothest of any combination I have found on the 2200.

It's actually very easy to understand why this is the case if you take the time to think about it.

The 2200 and x800 printers probably use C and Y for most of the deeper range of green, but they are not the exact same inks within the printer cartridges.

The 100% UV density of the Y ink for the 2200 is 3.36D; for the x800 printers, it is 3.39. These are essentially the same value, but visibly, they are slightly different mixes of pigments.

The 100% UV density of the C ink for the 2200 is 2.15D; for the x800 printers, it is 1.22. Not only are they visibly different, they measure out to be radically different in UV density.

When you mix an ink (in a dot matrix) with a density difference of 1.21D (3.36D-2.15DD) the individual dots have a difference of about 4 stops (16x).

When you do the same for the x800 printers, the difference is 2.16D, which is in the realm of 7.25 stops (150x !!!). In particular, this is a problem because that difference essentially exceeds the DR of most of the processes that people are making the negatives for.

So, it should come as no surprise that the green color approach fails miserably with the x800 printers. It's also not outside current logic why this works the way it does. It may be outside the PDN system, but is certainly not outside the basic approach that Clay has mentioned here.


---Michael

 

[sanking]

It's actually very easy to understand why this is the case if you take the time to think about it.


---Michael

That analyis takes more than just taking the time to think about it. To begin with, you have to know the UV blocking density of the cyan and yellow inks for both printers, which I don't.

To carry this a bit farther. What Clay wrote was this. "It is composed of two colors that are very far apart in their UV blocking abilities : Yellow and Cyan. It is no wonder when you put these dots next to one another that graininess can be seen in the higher negative density areas."

What you have said appears to agree with Clay's conclusion that the reson for the grainess is the considerable difference in UV blocking density of the cyan and yellow dots next to each other.

Does the logic follow that if the dots are close together in UV blocking density the results will be smoother? If that is the case it would appear that one could fairly easily predict the smoothest combination of inks for any given process ES.

Sandy

 

[sanking]

OH, and would either you or Clay post the UV blocking of the magenta ink(s) of the 2200 and 3800?

Sandy

 

[Michael Mutmansky]

Sandy,

The reason I said that it simply requires thought is that even without the numerical data, it doesn't take much to hypothesize that the grain is being caused by differences in the UV density of the inks, and that the grainer result is probably caused by a greater difference in UV density. Clay said as much in an earlier post. I didn't know that to be the case because I'd not compared them before, but the data supports that as the cause of the grain.

I don't believe you can characterize closer dots as smoother, except if you also mean they are also smaller, in which case dot gain will become a factor. The closer together the more likely for overlap, or for dot gain to cause blending, which could result in a smoother looking result.


---Michael

 

[sanking]

Sandy,

The reason I said that it simply requires thought is that even without the numerical data, it doesn't take much to hypothesize that the grain is being caused by differences in the UV density of the inks, and that the grainer result is probably caused by a greater difference in UV density. Clay said as much in an earlier post. I didn't know that to be the case because I'd not compared them before, but the data supports that as the cause of the grain.

I don't believe you can characterize closer dots as smoother, except if you also mean they are also smaller, in which case dot gain will become a factor. The closer together the more likely for overlap, or for dot gain to cause blending, which could result in a smoother looking result.


---Michael

Michael,

The reason I suggested that there may be more to this than the current logic indicates is because the UV blocking is not based on color as we might anticipate in looking at the spectrum. That is obvious from the differnt blocking of the cyan and yellow pigment dyes of the 2200 and 3800. And we don't know the bandwidth of the UV blocking, or how the bandwidith of the two pigment inks correlate to the exposing light.

It is a perfectly logical hypothesis that two colors far apart in UV blocking will cause more grain than two colors close together. And, as you say, the data in the case of the 3800 appears to support that view. But if it is true, does the same logic not also suggest that the smoothest tones would be achieved by printing with a single ink high in UV blocking, say the yellow?

Sandy

 

[Michael Mutmansky]

Sandy,

I think that using the Yellow on it's own (with a UV density of 3.54D on my film with the 4800) will cause dot gaps in the image as the density decreases. That's worse than having two inks with strongly different UV densities, as an ink gap will effectively have B+F density.

This is why blending several inks seems to work, as long as they have somewhat similar UV densities.

This is also in particular why yellow is such a problem, because the Y ink has no LY counterpart that it can be blended into as density decreases, but the C has a LC that will be blended in, which increases the difference between the UV density of the ink dots being used.


---Michael

 

[sanking]

Sandy,

I think that using the Yellow on it's own (with a UV density of 3.54D on my film with the 4800) will cause dot gaps in the image as the density decreases. That's worse than having two inks with strongly different UV densities, as an ink gap will effectively have B+F density.

This is why blending several inks seems to work, as long as they have somewhat similar UV densities.

This is also in particular why yellow is such a problem, because the Y ink has no LY counterpart that it can be blended into as density decreases, but the C has a LC that will be blended in, which increases the difference between the UV density of the ink dots being used.


---Michael

Michael,

That makes sense, it has definitley been my observation that the yellow ink does not produce results as smooth as blends. Though I speculate that the type of dithering pattern could play a strong role here.

Do you have the UV blocking density for magenta?

Sandy

 

[Michael Mutmansky]

Sandy,

I don't have a 3800 here. The 2200 produces 1.03D. Those numbers all include .11D for OHP B+F.


---Michael

 

[clay]

100% magenta ink on the 7800 : 0.98 logD UV

100% yellow ink on the 7800 : 3.14 logD UV

 

[sanking]

100% magenta ink on the 7800 : 0.98 logD UV

100% yellow ink on the 7800 : 3.14 logD UV

Clay,

And yet, on the 3800 at least, Red gives the smoothest tones I have seen from this printer with UV processes.

Sandy

 

[Kerik]

Clay,

And yet, on the 3800 at least, Red gives the smoothest tones I have seen from this printer with UV processes.

Sandy

Sandy,

The RGB BW negs I'm making are very bit as smooth as the red negs I tried. The red negs didn't develop enough density and behaved strangely relative to the linearity of the UV density values.

 

[sanking]

Sandy,

The RGB BW negs I'm making are very bit as smooth as the red negs I tried. The red negs didn't develop enough density and behaved strangely relative to the linearity of the UV density values.

Kerik,

OK, even if the RGB BW negatives are as smooth as Red ones, the fact remains that the Red negatives are also very smooth, and we are combining two colors, Yellow and Magenta, with UV blocking density difference as great as the difference between cyan and yellow. That is my point here.

As for the Red, the max UV blocking density is about 1.6, and I am not seeing any linearity problems at all. However, you can increase the UV blocking density a lot bylaying down more ink, which is feasible because the yellow and magent inks appear to be very fast drying. A DR of 1.64 is too low for my needs, and for several reasons I will probably go your route with the RGB BW negative, or do a QTR, rather than layng down more Red ink. But that does seem to be a viable option based on reports I have heard from others.

But who hijacked this thread? It was about the HP 9180!!

Sandy

 

[donbga]

But who hijacked this thread? It was about the HP 9180!!

Sandy

I think I may havve when I mentoned QTR. Unfortunately I wasn't thinking clearly. QTR doesn't support HP printers.

Don

 

[clay]

I's true we may have gotten off on a tangent here. But the point remains. There is a lot more going on with any printer and the native driver than we realize, which can affect the results. FWIW, you can get a lot of the same control over the print head with something like the Bowhaus OPM/IJC with non-epson printers. I really like these discussions, which in my mind is why we come to these forums. (Other than watching the occasional flame war from the side)

I think I may havve when I mentoned QTR. Unfortunately I wasn't thinking clearly. QTR doesn't support HP printers.

Don

 

[Michael Mutmansky]

Ultimately,

I think ANY real RIP can do what is being done in QTR.

It just so happens that QTR is shareware, and if you don't have a conscience, free. Otherwise, it's $50. I paid my $50 a long time ago, because it's clearly a great program for printing B&W prints, and it is now proving to me to be an excellent method for building precise, high quality digital negatives. Other RIP's are in the hundreds to thousands of dollars depending on the printer.

The reason this came up was your generalization that the recent Epson printers cannot comfortably build the density for good alternative process negatives. That may be true in the narrow confines of PDN and other RGB color blending methods, but is certainly not true with other approaches.


---Michael

 

[sanking]

The reason this came up was your generalization that the recent Epson printers cannot comfortably build the density for good alternative process negatives. That may be true in the narrow confines of PDN and other RGB color blending methods, but is certainly not true with other approaches.


---Michael

My comment about the inability of the current generation of Epson printers to build sufficient UV blocking density was based on the use of the printer out of the box, with Epson printer dirvers. If you have to resort to QTR or more sophisticated RIPS to lay down enough density, the "comfortable" factor disappears for many persons.

My interest in the HP 9180 was based on the desire to find a printer that would work well with PDN and other RGB systems like Clay's tertiary system or Michael Koch-Schultze's array system. All three methods, once understood, are relatively simple and offer a direct approach to matching process exposure scale to a color that produces the correct UV blocking density, without the need for special printer drivers and specialized equipment such as UV transmission densitometers.

Clay mentioned in an earlier message that with his RIP it is possible to get UV blocking of log 3.5 to 4.0 with the 7800. Does the Harrington QTR allow this as well? If so, it would certainly appear to be a good buy for those working with a X800 machine with processes like albumen and salted paper that need negatives of very high DR.

Sandy

 

[Ben Altman]

Clay mentioned in an earlier message that with his RIP it is possible to get UV blocking of log 3.5 to 4.0 with the 7800. Does the Harrington QTR allow this as well? If so, it would certainly appear to be a good buy for those working with a X800 machine with processes like albumen and salted paper that need negatives of very high DR.

Sandy

Hi, I've been using QTR to make inkjet negatives with the 2200 and now the 1800. So far I've made a lot of test wedges and not a lot of prints... but I have a couple of thoughts.

1. With a little ingenuity you can get QTR to do pretty much anything. It allows you to control the ink curve for each individual ink. So you can combine the inks in any way you want. The standard profile files that come with the program are not set up that way, but everything is there to do it. I use Excel spreadsheets to generate the curves. Unfortunately I don't have a UV densitometer yet, so I can't say what the actual UV densities are. I'm testing with Pd Ziatype, but there's room to block lots more UV than I need. Is this a user-friendly plug-and-play approach? Absolutely not, but if you want control, it's a good method.

2. One of QTR's neat tricks is to give you a 0% to 100% print-out of each separate ink. This instantly reveals the blocking of each ink, of course, but also shows the dot-gain characteristic of each ink. So you can take that data and figure out the ink combos that work at each density level. This approach allows creation of a combined ink curve that gives a linear response on your alt. process of choice, thus minimising the compensation needed in the image file in Photoshop.

3. I started with the standard 2200 inkset, but now I'm using refillable cartridges with LLK, LK, Y and a mixture of LK/Y. What I notice is that the 2200 and the 1800 give different ink curves for the same ink (and same dither). Something about the head design or dot size, I guess; the 1800 blocking is generally less than the 2200. Nice small dots, though!

Ben

 

[Asher Kelman]

Update for 2011?



Guys,

So is there a new thread that resolves these issues? Where are we today? What are the alternatives for reliable printing of digital negatives for the next process in the workflow, be it silver gelatin or Carbon transfer and so forth?

Where does HP stand in this? Are the series 800 printers worse than the newest versions? Is UV densitometry necessary today?

Thanks,

Asher