new digital neg system from cone

[andy]

have you all seen this?

http://piezodn.inkjetmall.com/

"PiezoDN is a set of tools and QuadtoneRIP curves for printing and calibrating high-quality negatives using Piezography ink. PiezoDN is also a set of quality standards and ideas for revolutionizing both darkroom printing and digital printing as a whole."

[Taija71A]

Andy...

Thank-you, for posting this information about 'PiezoDN'. Greatly appreciated!
Regards, -Tim.

[DennisD]

Sounds great ! And very interesting. However, no indication of the cost.

Clearly, a "work in progress" with kinks to be ironed out along the way.
Important to note the timeline schedule for certain milestones and functionality with different hardware.

Perhaps best suited for those already familiar with piezography.

I'm not a very good Guinea pig ! So, I'll probably wait till the system is further along and perfected.

I don't know Jon Cone, but I'm impressed by his dedication and many years of research that raised the bar w/r/t high quality inkjet printing.

[neil poulsen]

It may add to darkroom printing, but I don't think that it replaces it.

The test is whether it can convincingly print silver gelatin prints. That would be interesting to see.

[LesleyNowlin]

i'm really looking forward to trying this out!

[Michael Rosenberg]

If you are interested in printing on silver gelatin paper using digital negatives, you should check out my monograph on making digital negatives with QTR and Epson Inks. (http://www.mprosenberg.com/digital-negatives). It is free, and does not require switching out inks. I have had feed back from a lot of people who have tried this and they report success. The profiles and curves work with a variety of silver gelatin paper.

Mike

[bob carnie]

I would be interested in a ink system that has the same blocking power and sharpness as silver film, I am producing enlarged contact film for silver with very
good results, I use a Durst Lambda with 20 inch and 30 inch Rollie ortho 25 developed in HC110. We are able to make up to 28 inch x 40 inch live area on film.
I also make inkjet negs but have not seen great results to silver gelatin - basically a lot of light bleed through the ink.

We are using Ron Reeders QTR system on a Epson 7800- though great for alternative rag prints , not so great for silver contact.

Michael - could you describe how your system is different to allow better blocking power using ink sets and pictorico?

[sanking]

We are using Ron Reeders QTR system on a Epson 7800- though great for alternative rag prints , not so great for silver contact.



Bob,

Are you still printing digital negatives on the 7800 with the Matte Black ink installed? If so, switching to PK would improve greatly the smoothness of your negatives on any process that uses smooth papers. You would have to create a new profile, of course.

Sandy

[Michael Mutmansky]

Bob,

Are you still printing digital negatives on the 7800 with the Matte Black ink installed? If so, switching to PK would improve greatly the smoothness of your negatives on any process that uses smooth papers. You would have to create a new profile, of course.

Sandy

Yes, I'm curious about this too because I can get densities in the 3.5 or higher range using stock Epson PK ink on film. That's way too high for SG printing, so if I were looking to print SG, I'd be looking are developing a 'shades of gray' approach that meets the density requirements you want. In the case of SG, it may be possible to use a 7-ink printer and set the maximum somewhat higher for other processes (gum and pt/pd), and then only use the lower density inks for SG. That would help avoid gritty highlights associated with high density black ink that is too high at the individual ink droplet level.

I recently put together an all-black inkset for my old 4800 with help from Sandy, but I took a slightly different approach than he did because I was focusing on pt/pd and he focused on carbon.

Anyway, the point is that I used all 7 inks to put it together, but I'll bet that a solid negative could be produced using 5 inks for pt/pd and gum, and similarly for SG. What could be done is set the ink densities so that there are a few dedicated to pt/pd and a few for SG, and the rest are used for both. Two of each dedicated, and three used for both produces 5 inks in each system without changing the inks.

Something like this:




#1_________#2_________#3___________#4_______________#5___________#6___________#7______
pt/pd

PK_______.5PK/.5LK______LK_______.66LK/.33LLK________________________LLK

SG

_______________________LK_______.66LK/.33LLK______.66LK/.33LLK_______LLK_______.5LLK/.5base

This will allow you a wide latitude for laying down ink in both density ranges but still have reasonable steps in density so that the highlights aren't gritty, and the shadows have enough ink to not be real spotty as well.

You could use #5 in the pt/pd as well, but it is probably unnecessary, based on my tests and the general low response curve in the shadows for pt/pd. This, in particular, is one place where I made a substantial change from the normal approach that many people have used.

You probably wouldn't want to use #1 at all in the SG, and probably not #2. I don't recall the Dmax that this blend should produce, so I don't know how useful it will be for SG.

An alternate is to do 2 steps between the PK and the LK instead of one, and then just do a single step between LK and LLK. All of this will require some basic testing to determine what appears to be the best approach.


---Michael

[WalkerBlackwell]

I'm the one who R&D'd the system with Jon and did all the math and programming. (I've been a lurker here for a long while fyi.)

PiezoDN is drastically different than any other dig-neg process as it takes into account the full trip from camera to print including film acuity requirements, durability, physical ink bit-depth, 16bit calibration and print-path, nozzle frequency, dMax of final ink (4 to 5 inks overlapping in the dMax makes a black that is significantly more opaque than what you can make with epson OEM) and much more.

For those who print in the darkroom and use LF equipment, this is simply a logical extension of that idea and process.

We built a silver curve for it for the Epson 4900 that works well (Bessy 23c, Ilf MG IV, 16sec, f8, Sprint Dev is your baseline). Alan Vlach up in Maine can speak for the silver curve more eloquently than I. Another great photographer and artist (Ray Meeks) has used an early early variant of it at Light Work when I was managing that lab. He actually enlarged 4x5 versions of this negative but frankly I think it's best suited for contact-printing.

Our main goal was to build a truly professional system for Platinum and Palladium. We inadvertently also profiled full tonal separation into Cyanotype, Ziatype (one variant), Salted, and a few others.

We are Gold Master Release Candidate 1 as of tonight and will be releasing version 1 tomorrow morning. Over the intervening months and years we will slowly add support for all Piezography-capable printers and all hand-process types. Hopping on board (cost of $75) gives you access to all the tools and updates and community forever.

regards,
Walker

[WalkerBlackwell]

I guess the most important part of PiezoDN is that the tools included enable you to linearize a .quad directly from 129 printed patches using Roy Harrington's QTR-Linearize-Quad. This is a significant improvement over 51 steps + guessing in the dark (or 16 steps in a photoshop adjustment curve)

The error correction tool is built to correct for darkroom printed patches where light fall-off or brush-marks may be present.

[Michael Mutmansky]

How many of those are duplicate patches?

[Greg]

I have been making digital negatives for the past few years per Dan Burkholder's system as described in his book: The New Inkjet Negative Companion. Truly 90% successful with first "final prints". In the world of Pl/Pt printing... cost effective to almost the nth degree. For me: "If it's working and not broken, why change?"

[Michael Rosenberg]

Bob,

The negatives with my system look like a silver gelatin negative, with smooth transitions between tones, little to no grain (on the silver gelatin print), and very sharp. Sandy King visited me a couple of years ago and saw the negatives and the resulting prints. He has since adapted part of my method for carbon print negatives. My system builds on Ron Reeders method for alt pritns using QTR. The process is linear, so the higher density of a silver gelatin negative is the densest part of the negative; as you can see from the image of the ink curves they sort of look like a film density plot. My method differs from all other methods as it introduces a linearized curve in the ink descriptions,, rather than trying to impose a linearization on top of all the inks (see insert picture and my monograph). This is clearer when you look at the top part of the image showing the QTR Gui ink settings. Thus, all but one ink overlays each other.

LC and C are not used because silver gelatin paper does not respond to cyan/ink color. Because the inks overlap the ratio of magenta and yellow inks affects local contrast. Overlap of the inks also lays down more ink for blocking light and reducing grain. I use the Epson 4880 and 7880 printers with the Epson inks. One other big difference is the negative media; I found transparency and white film do not give very good results, as I explain in my monograph. Instead I have found that glossy print paper gives excellent results - less grain and very sharp. Not to mention it is much cheaper!151443

Several people who have tried my system have given me feedback that the two curves and profiles I supplied worked extremely well, and they did not need to change anything.

You can download the monograph from my web site, and if you send me your e-mail I will send you the files (My web host does not allow them to be downloaded from my site).




I would be interested in a ink system that has the same blocking power and sharpness as silver film, I am producing enlarged contact film for silver with very
good results, I use a Durst Lambda with 20 inch and 30 inch Rollie ortho 25 developed in HC110. We are able to make up to 28 inch x 40 inch live area on film.
I also make inkjet negs but have not seen great results to silver gelatin - basically a lot of light bleed through the ink.

We are using Ron Reeders QTR system on a Epson 7800- though great for alternative rag prints , not so great for silver contact.

Michael - could you describe how your system is different to allow better blocking power using ink sets and pictorico?

[Denny]

I have been making digital negatives for the past few years per Dan Burkholder's system as described in his book: The New Inkjet Negative Companion. Truly 90% successful with first "final prints". In the world of Pl/Pt printing... cost effective to almost the nth degree. For me: "If it's working and not broken, why change?"

I completely agree, I've been using Dan's approach for years, and I get great negs and prints in Pt/Pd. Relatively simple and effective.

[WalkerBlackwell]

@michaelmutmansky None of the patches are duplicate. We error correct in a different way. In fact we've error corrected in a different way since 1994 which is why we've been linearizing Piezography at 256 patches internally for decades.

PiezoDN is both linear and non-linear in that everything is baseline calibrated linear and then optionally perceptually pulled into contrast depending upon the dMax and slope angle of the actual physical print.

Included is a tool for burning contrast intents directly into the .quad as well enabling split toning of contrast intents directly in QuadtoneRIP on-the-fly pre darkroom print.

Because everything is done in the .quad directly (except for optional perceptual print to screen matching which is done in an ICC that you print with in Print-Tool), this maintains the 16bit data-flow.

best,
Walker

ps: There is no reason why people wouldn't still be able to use their own systems however. The real magic is the baseline curve overlaps and ink frankly.

[bob carnie]

Hi Walker

Do you guys ever offer one or two day training on your systems?
I for one would be quite interested to learn

Bob

@michaelmutmansky None of the patches are duplicate. We error correct in a different way. In fact we've error corrected in a different way since 1994 which is why we've been linearizing Piezography at 256 patches internally for decades.

PiezoDN is both linear and non-linear in that everything is baseline calibrated linear and then optionally perceptually pulled into contrast depending upon the dMax and slope angle of the actual physical print.

Included is a tool for burning contrast intents directly into the .quad as well enabling split toning of contrast intents directly in QuadtoneRIP on-the-fly pre darkroom print.

Because everything is done in the .quad directly (except for optional perceptual print to screen matching which is done in an ICC that you print with in Print-Tool), this maintains the 16bit data-flow.

best,
Walker

ps: There is no reason why people wouldn't still be able to use their own systems however. The real magic is the baseline curve overlaps and ink frankly.

[Michael Mutmansky]

Do you work in pt/pd or other alternate processes? I think the variability of the coating process, the nature of matte surface papers, and the inherently low contrast of many alt processes would make a 129 patch linearization effectively impossible to achieve success using the Linearize function. It is inherently unable to properly work if the readings deviate too much from what it considers a properly increasing dataset. Unless I am not understanding you correctly.

[bob carnie]

Good Point Michael - I like using the numbers in PS and trying to see how they relate to the printed process.
I have found the best % of accurate lay down of tones was on my digital fibre prints where in the top end and lower end I see separation on print as read by the info pallete.
I have seen 96L values read as highlight density on silver , and 4 L values read as shadow density.
There is less tone separation on inkjet, But as you increase the gloss media we get better results.

I like what Sandy King said to me years ago , make a chart of 100 steps from 0 -100 and actually see on print how many you can achieve.. I think this is a critical step for any way one wants to make digital neg's.

As Michael points out getting all steps to read on the final print as they are seen in PS is almost impossible and there are some compromises to be expected. IMO this is where more ways of creating digital negatives is going to be on everyone's radar in the future.


Do you work in pt/pd or other alternate processes? I think the variability of the coating process, the nature of matte surface papers, and the inherently low contrast of many alt processes would make a 129 patch linearization effectively impossible to achieve success using the Linearize function. It is inherently unable to properly work if the readings deviate too much from what it considers a properly increasing dataset. Unless I am not understanding you correctly.

[sanking]

I suspect there is no hand coated alternative process that would allow use of the LINEARIZE command with a 129 step wedge. Even a 51 step wedge is incredibly difficult to reproduce with sufficient accuracy to permit use of LINEARIZE.

Sandy

[WalkerBlackwell]

Huh. Than a guess the last 300+ prints I've seen run on 129 linearized PiezoDN .quads aren't real. ;)

But for sure, there are "reversals" that show. The trick is to algorithmically correct for said reversals (laterally along the line instead of by measurement of multiple patches) and glean more accurate data from the total line than you would from a 21step.

That said, every target file of PiezoDN has a 21 at the bottom to enable self-checking.

best,
Walker

[WalkerBlackwell]

There is a screen-shot I included in the about section of the PiezoDN website that shows the smoother tool in action. I intentionally included SUPER rough reversals in there so you can see what it's doing.

QTR-Linearize-Quad droplets performs much better when you give it more data-points as I don't think it's doing truly stable interpolation for low amounts of patches (21). The 129 patches + control of the ink actually effectively makes Roy's linearizer into a very high-quality tool indeed.

Upcoming in PiezoDN is a 31x4 tool that will output 129 quadratic data to be used by the Linearizer. This may be just as good as 129 in that it will keep the linearizer from getting "wiggly" when given low amounts of data.

[Michael Rosenberg]

Bob and Michael,

Yes, the image in PS remains unaltered, and is how it appears in the negative. My test target has tone values with ink and k values making it easy.

I agree, it is very hard to see subtleties in tonalities in the silver gelatin print. I attribute this to the paper curve, where the slope blurs slight differences in tonalities. Try printing a Stouffer 21 step tablet. I found a 21 step target to be more than adequate, and the test target I provide has a 21 step strip and a Kodak Zone strip to allow either reflection densitometry or visual inspection. The ink curve is a PS curve, and it extends to the left to provide better shadow detail; for highlights I define LLK as grey ink and then use black boost (which only works on grey ink) to get increase in density at the right end of the curve.

By linearizing the inks it makes it possible to predict the effects on the negative density and changing the print. When Ron Reeder adapted QTR for making digital negatives for alt processes he started with paper curves, and then linearized using grey curve. If you changed one ink you changed the deposition of all inks. In my system if you change on ink you only change that one in, either increasing or lowering density in the negative. It is intuitive that the ink curve resembles the film curve for TMax films (you can even make it look like the film curve for other films too), and then relates to the paper curve. Remember, you can only correct the ink curve so far before running up against the paper curve. I know - I ran into this myself!

The system I developed and the files I provide are all plug and play. They work for Ilford and Adox papers. I have been using it for the past 5 years with no problem.

If you want to split contrast it is very easy to do. Simply add a separate PS curve for yellow or magenta inks using the toner function of QTR. I have done this a couple of times, and with the test target I provide it is easy to work out..

Again, send me a PM and I'll send you the files of curves, profiles, target image, and etc.

Mike

[WalkerBlackwell]

I am not here to defend or compete with anyone's system. Just wanted to add to the conversation. I'm super excited about PiezoDN because I think it hits a sweet spot of quality and simple adaptability needed by experimental analogue printmakers of all stripes (including myself and Jon of course). However, I'm not here to say it's better than anything else.
The users and viewers of PiezoDN will decide yay-or-nay after seeing its fruits in the flesh. There is a nice print at Latitude (possibly still there) that I did that anyone in Chicago is welcome to look at. And of course there will be many many more down the line.

Currently this system is bed-rocking the entire historic-process lab at Maine Media Workshops. Anyone in that region should check it out (or take a class!). There are some amazing genius darkroom people there. I was truly impressed by their abilities!

cheers all,
Walker

edit> www dot mainemedia dot edu/workshops/photography/historic-process-mentoring

Brenton is amazing

[Michael Rosenberg]

There is a screen-shot I included in the about section of the PiezoDN website that shows the smoother tool in action. I intentionally included SUPER rough reversals in there so you can see what it's doing.

QTR-Linearize-Quad droplets performs much better when you give it more data-points as I don't think it's doing truly stable interpolation for low amounts of patches (21). The 129 patches + control of the ink actually effectively makes Roy's linearizer into a very high-quality tool indeed.

Upcoming in PiezoDN is a 31x4 tool that will output 129 quadratic data to be used by the Linearizer. This may be just as good as 129 in that it will keep the linearizer from getting "wiggly" when given low amounts of data.


Walker,

Can you help me understand what you mean by QTR linearize Quad droplets? I am unfamiliar with this term. Are you referring to the ink droplets? And what do you mean stable interpolation?

Mike

[LesleyNowlin]

Currently this system is bed-rocking the entire historic-process lab at Maine Media Workshops. Anyone in that region should check it out (or take a class!). There are some amazing genius darkroom people there. I was truly impressed by their abilities!

cheers all,
Walker

edit> www dot mainemedia dot edu/workshops/photography/historic-process-mentoring

Brenton is amazing

Brenton is amazing! Such a great teacher. I love the Maine workshops too. Took the alt-process class and I learned tons of processes and found my love for Platinum a few years back. I'm looking forward to trying the PiezoDN. What if I need help setting it up? Do you have suggestions on where to go?
Thanks,
Lesley


Sent from my iPhone using Tapatalk

[WalkerBlackwell]

@lesleynowlin It's certainly a bit of time to set it up.

It requires a kit: Piezography ink, the cartridges, the software. All of the instructions can be downloaded right from each product and we also offer direct support at the private forum at PiezoDN. There is an extensive Piezography manual that comes with the system as well that covers far more than one can teach in a normal printing class.

best,
Walker

[WalkerBlackwell]

@bobcarnie yeah, we teach this. It's really the best way sometimes for people to come actually see what is possible. I think all our group workshops are sold out (they sold real quick so we increased from 5 to 6 students) but we do individual day-long stuff as well.

best,
Walker

[sanking]

"But for sure, there are "reversals" that show. The trick is to algorithmically correct for said reversals (laterally along the line instead of by measurement of multiple patches) and glean more accurate data from the total line than you would from a 21step."


Is the software that algorithmically corrects the "reversals" in the PiezoDN package?

Sandy

[Michael Mutmansky]

"But for sure, there are "reversals" that show. The trick is to algorithmically correct for said reversals (laterally along the line instead of by measurement of multiple patches) and glean more accurate data from the total line than you would from a 21step."


Is the software that algorithmically corrects the "reversals" in the PiezoDN package?

Sandy

Clearly, he's not using linearize without scrubbing the data first... which makes all the steps pointless with alt process. Might as well go with the 21-step curve at that point and not mess around with 129 steps.

At that point, a nice fine tuning of the curves can be done in finer steps successfully, but the linearize doesn't need to be done to so many steps.

[Michael Mutmansky]

Walker,

Can you help me understand what you mean by QTR linearize Quad droplets? I am unfamiliar with this term. Are you referring to the ink droplets? And what do you mean stable interpolation?

Mike

He's referring to an app that is used on the Mac (and also on the PC I think, not sure) that linearizes the file with an adjustment curve automatically.

However, it struggles to properly linearize a curve if there is data the isn't constantly increasing, and even if it is increasing, sometimes it fails to succeed (presumably because it is a little too discontinuous).

In the PC, the process to develop the curves is a little different because of the GUI program that was developed, so I don't think the app is used in the same way, but I believe the same calculations are made when using the linearize function in the Curve GUI.

[sanking]

He's referring to an app that is used on the Mac (and also on the PC I think, not sure) that linearizes the file with an adjustment curve automatically.

H

And a bit of clarification, if necessary, it is a Linearize Quad droplet that was developed by Roy Harrington. The creation of the aplet was preceded on the QuadTone Rip forum on Yahoo by extensive discussions about how to farther linearize an existing Cone profile, since these can not be linearized with traditional QTR creation tools. If we create a profile with the regular QTR tools, and linearize it with GRAY-CURVE or LINEARZE, we can adjust it later by going back and re-doing the curve, but the rest of the profile remains intact. With the regular tools really good profiles can be created with 21 point curves, 51 is overkill for most applicaitons and definitely problematic with alt processes. But it appears that the use of the droplet to correct a Cone profile requries a lot more points than would be needed for the first linearization with regular QTR tools.

One of the drawbacks, IMO, of the previous Cone digital negative methodology, Methodology 2, was that there was no way for the user to adjust the profile with an internal curve. Rather, the user had to apply the curve to the image file. If the droplet works as hoped, it will make the Cone digital negative system much more flexible.

Sandy

[WalkerBlackwell]

"it appears that the use of the droplet to correct a Cone profile requries a lot more points than would be needed for the first linearization with regular QTR tools."

It actually doesn't "need" anything more than what other methods take. By default the base-negs are algorithmically and UV/Pan density smooth. To properly explain the potential of this system, what we realized is that with Piezography inks and proper ink values for all 6 shades and the proper calibration tools set, we could measure and calibrate 256 patches. I built a linearizer in-house to do this for negatives. Roy's linearizer only does up to 129 so we limited it to 129 for PiezoDN customers. That said, we didn't see much different between the 129 and 256 other than when we were doing the initial master curves (that was good at 256).

This system is so way beyond 21 it's hard to really compare it. But 21 would certainly work for it just like it does for every other QTR curve.

What we did find was that if there are sharp changes in curve data on a 21 step target, the linearizer can over-correct. (This is the case with all QTR utilities. It's the same base math.)

best,
Walker

[George Pappas]

Walter,

I appreciate the explanations as I have been a Cone B&W printer with the Selenium inkset for some time and looking for a system like this. Can you comment on the smoothness of the results for Silver Gelatin printing? This has been a common issue with past systems using inkjet negs compared to straight silver gelatin printing with "original" large format negs. Printing with more textured papers and other alt processes (Silver Gelatin is now an alt process!) do not bring out lack of smoothness the way that fiber based silver gelatin printing does.

Much of the PiezoDN material speaks to platinum/palladium so not sure how good you think this really is for Silver Gelatin (still my favorite) printing.

Thanks in advance for your candor....

Regards,
George

[bob carnie]

I too am interested in how the performance is against silver film, Walter can I send a file to you to print a neg and I will match size and make a silver neg here and print both negatives it on Ilford Warmtone, my paper of choice? I would pay for your negative output of course.
I would love to see a practical comparison on 16 x20 paper.
Bob

Walter,

I appreciate the explanations as I have been a Cone B&W printer with the Selenium inkset for some time and looking for a system like this. Can you comment on the smoothness of the results for Silver Gelatin printing? This has been a common issue with past systems using inkjet negs compared to straight silver gelatin printing with "original" large format negs. Printing with more textured papers and other alt processes (Silver Gelatin is now an alt process!) do not bring out lack of smoothness the way that fiber based silver gelatin printing does.

Much of the PiezoDN material speaks to platinum/palladium so not sure how good you think this really is for Silver Gelatin (still my favorite) printing.

Thanks in advance for your candor....

Regards,
George

[George Pappas]

Excellent Suggestion, Bob! I'll help with the costs if Walter is game so that I can see the real thing in the flesh - would split the negative charge and ask that you make a second print for me. For me, it is all about highlight differentiation...the result will tell the tale. If it meets the test, then it will be worth the investment.

Regards,
George

[bob carnie]

I would ship both negs to you as well George so you can actually see with your own workflow and paper. Should be a good test for those wanting to see the difference.
I have a vested interest in the silver neg output so it would be good for other workers to compare apples to apples.


Excellent Suggestion, Bob! I'll help with the costs if Walter is game so that I can see the real thing in the flesh - would split the negative charge and ask that you make a second print for me. For me, it is all about highlight differentiation...the result will tell the tale. If it meets the test, then it will be worth the investment.

Regards,
George

[WalkerBlackwell]

I would love to but do not have the printer in-house that we R&D'd silver densities for. We actually donated that to Maine Media Workshops (our one epson 4900). The 9900 that we are currently doing R&D on actually has a different head and nozzles (so different densities). We are getting a device (and building a workflow) for us to translate densities from one film-type/printer to another without going back into the darkroom every time. The second that is done I can do sample silver negs. Anyone with a 4900 and selenium can do it right now with the base .quad. We actually don't have a silver darkroom in our little rural facility in Vermont but if this device takes too long to arrive I'll do the 256 silver neg density translation by hand to get it rolling on the 9900. But the truth be told, I the sample silver prints we ran (almost as an off-shoot) at MMW were smooth and fully matched the monitor.

[bob carnie]

So could you recommend someone who could make a good neg using your system that I can send a file too for my testing purpose?


I would love to but do not have the printer in-house that we R&D'd silver densities for. We actually donated that to Maine Media Workshops (our one epson 4900). The 9900 that we are currently doing R&D on actually has a different head and nozzles (so different densities). We are getting a device (and building a workflow) for us to translate densities from one film-type/printer to another without going back into the darkroom every time. The second that is done I can do sample silver negs. Anyone with a 4900 and selenium can do it right now with the base .quad. We actually don't have a silver darkroom in our little rural facility in Vermont but if this device takes too long to arrive I'll do the 256 silver neg density translation by hand to get it rolling on the 9900. But the truth be told, I the sample silver prints we ran (almost as an off-shoot) at MMW were smooth and fully matched the monitor.

[WalkerBlackwell]

Of the group of people who bought the system (at launch yesterday) I'm not sure who has a 4900. I'll send the call out.

In the interim I'll work on getting these densities translated.

best,
Walker

[Michael Rosenberg]

Bob and George,

I would be glad to print negatives for you both on 16x20 media. Only pay for shipping. I have done this for others.

Mike

[bob carnie]

Ok - I will send you back the neg I make same size so you can see what I am up too.
George if you want to supply the image I am ok with that as well

Bob
Bob and George,

I would be glad to print negatives for you both on 16x20 media. Only pay for shipping. I have done this for others.

Mike

[bob carnie]

I would love to test your system as well Walker.

let me know

thanks
Bob

Of the group of people who bought the system (at launch yesterday) I'm not sure who has a 4900. I'll send the call out.

In the interim I'll work on getting these densities translated.

best,
Walker

[bob carnie]

Hey Michael I see your from the Durham NC area, two of my friends have just moved there and are setting up a bit of an art center.
I hope to visit them soon.

Bob

Bob and George,

I would be glad to print negatives for you both on 16x20 media. Only pay for shipping. I have done this for others.

Mike

[Michael Rosenberg]

Bob,

Please stop by when you visit. Where are they setting up the art center? I am in a gallery in downtown Durham. Have them contact me.

Bob and George,

I think the best way to go is for each of you to put an image in DropBox for me to download. Then I will print them out and send you the negatives as well as a negative of the target for you to establish the exposure and development times.

BTW, the curves work for all x880 printers (I have the 4880 and the 7880), and presumably the newer printers if you don't use the other ink channels.

Mike

Mike

[George Pappas]

Thank you Walter and Mike. I have been off the grid all day. I'll send a fuller reply tomorrow. Regards, George

[Richard Boutwell]

I don't check the LF forums all that often and have some things to add. I have been working on a system very similar to what Walker has developed for Piezography. The way he uses linearize-quad and ICC profiles is kinda genius. I built my own quad linearizer before Roy released the linearize-quad app and just use my own for everything now—prints and a modified one for negatives. The quad files from my smoothed/linearized QTR curves look pretty different than the PiezoDN curves and I am anxious to see how differently they end up printing.

129 steps: I can say that using 128-256 steps is definitely better than 21 and 51 steps—not so much for linearizing the midtones (which is the easy part) but for getting perfect gradations from 0%-2% and 98-100%. That is where the extra steps that break the gray scale into less that 1% increments is really helpful. I also have to add that the data smoothing with that many steps is totally necessary—I had the same issue with my system at first—but it doesn't do anything to drastically alter the shape of the linearization curve that is created.

I did some tests a few weeks ago on Lodima contact printing paper using 5 shades of selenium inks and my method of making smoothed curves and my own linearizer (without the simultaneous GO). The subtle highlights and shadows are perfectly smooth, but some of the tones between 40%-60% still have a grainy look similar to what I found with the older PZDN inkset and a Photoshop correction curve. I am setting up a 1430 tomorrow and a 3800 with the new PiezoDN on Monday for a few quick initial tests, but am then setting up a second 9900 as soon as possible so I can to get closer to the ideal printing conditions. My hope is that with more overlap along the whole scale the new PiezoDN system will get rid of that graininess. I realize the 4900-9900s might be ideal, but in some ways using the 3800 is a good real world test of the profiling system as a whole.

M Rosenberg: I saw your prints several years ago at the Michael and Paula workshops... How do your gelatin silver prints from your digital negatives compare to those analog enlargements?

Walker and Bob Carnie: If I can get a negative linearized for Lodima with the new PiezoDN system when I am in the darkroom monday or tuesday I will send up a few sample prints.

[Michael Rosenberg]

Richard,

Good to hear from you! I have kept up with you on your web site over the past few years.

The images with the digital negatives are the same quality in terms of grain, smoothness of transition of tones, and identical in that you can't tell it is from a DN. In another sense they are better because the image is not projected through an enlarger lens and does not suffer from diffraction, and thus resolution is better; but I can also control sharpness through PS. It is possible to make a digital negative that is too sharp. Of course when I printed before, particularly the Tobacco Factory series, I was dealihg with large contrast ranges and had to control for shadow and highlight details by using USM, sharp masks, dodge masks, and burn masks,

The issue of grain was a problem I encountered when starting out. I realized that there was two causes of grain: First was depositing enough ink in the lighter tones to result in the ink droplets to overlap on a microscopic scale (certain the drops from the print heads were very fine). The second was I needed a negative material with a surface structure that allowed more ink to be deposited and dry quickly. For the latter, I found that OHP and white film had a coarse structure (I read up on this), I realized that glossy thin (cheap) print paper had the desired surface structure and works very well. The only caveat is that it takes a little longer for exposure - a non-issue. The paper is UV opaque, and so won't work for alt-processes.

The problem of having enough ink overlap and reduce grain was solved when I introduced a linearized curve in the ink descriptor. Look at the ink curve I posted earlier - you can see all the inks overlap over the full range of the negative, and differ only in slope of the individual inks. Sandy adapted part of my system for making carbon prints introducing a curve in the ink descriptor of a single ink and saw a decrease in grain in his prints. Ron Reeder had done the same for a UV blocking ink to get better shadow detail in his prints.

The need in many systems to perform extensive linearization is be because the ink profiles were adapted from those used for digital prints, where inks rise and fall on the x axis. Linearization in gray scale is an attempt to change the extent of overlap of the curves over the range of negative tonalities, complicated by the fact that not all colors block light equally. I discuss this more fully with examples in my monograph (available for download free from my web site). By the way, it took me two rounds to linearize, and any further attempts just made me realize that I was measuring the paper curve.

Mike

[bob carnie]

Hi Richard

Michael sent me a box of Lodima to test last week, what I would like to see is a test negative on the different systems as well as my image setter to compare on different papers.
I also promised him to output a negative for you to test.
Durst is coming tomorrow to install my blue laser and knock on wood I will be back up and running within weeks for film output.

I think we are all going at the same thing here, very interesting , as I get lost in the inkjet techno talk, but I have been creating image via neg or pos since 1976 printing for others, so
my strengths are in working the materials, thank goodness for the geeks of the world to figure out the math. Ron came to my shop and has been a strength for us for years, as well Sandy
always has the patience to walk me through all the tech talk.

For those making prints from digital negatives to silver I think the true test is a glossy NOT matt paper.

I have made 16 x20 prints on Ilford Warmtone Glossy, using a 4 x5 negative, then scanned the negative and made an exact size silver neg from my image setter and then contacted the negative onto Ilford Warmtone Glossy.
The resulting prints were an eyeopener for me as they were indistinguishable to my eyes, the photographer and others who have seen this.
I would love to sent a file around to others here to make their negatives to a common size then I will print them on Ilford Warmtone Gloss. Very open then to send all the negs to another worker to
do the same test on their paper of choice.

For me the proof will be on the print. I have noticed with pt pd and rag textured silver some leeway to great results, In my darkroom I have not seen the same results on glossy silver paper.
Must be noted I made the digital negs on a epson 7800 using Ron/Roy's QTR program, and I will openly admit that I believe there are others posting here that have a better handle on ink jet negatives than
I did three years ago and probably today, It would be great to be able to offer both types of negatives to my clients as output.


Anyone notice a problem with images transferring to one another when stacking a grouping of separations within the same folder?. I am using matt inks on Pictorico Premium Overhead material
and if I do not keep the negs separated for a good while I find image transfer from one another. Solved this by complete separation and sleeves for each neg, but a real can of whoopass about three months ago
when I was preparing a 60 print show. Lots of Redo's .





I don't check the LF forums all that often and have some things to add. I have been working on a system very similar to what Walker has developed for Piezography. The way he uses linearize-quad and ICC profiles is kinda genius. I built my own quad linearizer before Roy released the linearize-quad app and just use my own for everything now—prints and a modified one for negatives. The quad files from my smoothed/linearized QTR curves look pretty different than the PiezoDN curves and I am anxious to see how differently they end up printing.

129 steps: I can say that using 128-256 steps is definitely better than 21 and 51 steps—not so much for linearizing the midtones (which is the easy part) but for getting perfect gradations from 0%-2% and 98-100%. That is where the extra steps that break the gray scale into less that 1% increments is really helpful. I also have to add that the data smoothing with that many steps is totally necessary—I had the same issue with my system at first—but it doesn't do anything to drastically alter the shape of the linearization curve that is created.

I did some tests a few weeks ago on Lodima contact printing paper using 5 shades of selenium inks and my method of making smoothed curves and my own linearizer (without the simultaneous GO). The subtle highlights and shadows are perfectly smooth, but some of the tones between 40%-60% still have a grainy look similar to what I found with the older PZDN inkset and a Photoshop correction curve. I am setting up a 1430 tomorrow and a 3800 with the new PiezoDN on Monday for a few quick initial tests, but am then setting up a second 9900 as soon as possible so I can to get closer to the ideal printing conditions. My hope is that with more overlap along the whole scale the new PiezoDN system will get rid of that graininess. I realize the 4900-9900s might be ideal, but in some ways using the 3800 is a good real world test of the profiling system as a whole.

M Rosenberg: I saw your prints several years ago at the Michael and Paula workshops... How do your gelatin silver prints from your digital negatives compare to those analog enlargements?

Walker and Bob Carnie: If I can get a negative linearized for Lodima with the new PiezoDN system when I am in the darkroom monday or tuesday I will send up a few sample prints.

[bob carnie]

Re: new digital neg system from cone

((( Richard,

Good to hear from you! I have kept up with you on your web site over the past few years.

The images with the digital negatives are the same quality in terms of grain, smoothness of transition of tones, and identical in that you can't tell it is from a DN. In another sense they are better because the image is not projected through an enlarger lens and does not suffer from diffraction, and thus resolution is better; but I can also control sharpness through PS. It is possible to make a digital negative that is too sharp. Of course when I printed before, particularly the Tobacco Factory series, I was dealihg with large contrast ranges and had to control for shadow and highlight details by using USM, sharp masks, dodge masks, and burn masks,}}}}}



Hi Richard was your project on Gloss or Matt paper?, I think Sandy mentioned he attended a show and was very complimentary of the prints. Somehow I think your work was it.

[Michael Rosenberg]

Bob,

Just to cover a few points: I only use glossy papers, and my favorite two are now Adox MCC and Ilford WT. My tobacco factory work was all done with Forte, and I bought a couple of extra boxes to store in a chest freezer when they closed to be able to make more prints.

I always dry my negs over night or with a hair dryer for immediate use. Never had an issue with image transfer. I store them all in a box and never found the need to use interleave paper with no degradation.

Mike

[sanking]

Hi Richard was your project on Gloss or Matt paper?, I think Sandy mentioned he attended a show and was very complimentary of the prints. Somehow I think your work was it.

Bob,

I assume you meant to ask this question of Mike, not Richard?

I visited Mike about three years ago at his home near Chapel Hill, NC and had an opportunity to see his well-equipped work space and many of his beautiful silver prints, some made with traditional analog methodology, others printed with digital negatives.

Mike's method of creating a profile with QTR using the Epson K3 ink set got my attention and I decided to experiment with it in my carbon printing. Essentially Mike is using a single gray value system where all curves follow the K, and on top of that he also applies a special paper curve. Previously I had worked with a number of methods to make digital negatives with the Epson K3 inkset, including color spectral negatives made with the Epson driver, and with QTR with Ron Reeder methodology. What I quickly realized was that Mike's single gray value system provided a simple mechanism to balance UV blocking density by setting the specific ink limit of each useful ink in the K3 set (only K, Y and LK are useful blockers) so that all three inks block the same amount. This is highly useful because it assures optimum overall smoothness, and it provides a simple mechanism for increasing or decreasing overall density range, making it easy to adjust for any process. The profile can also be adjusted internally for linearity with either the K curve, or with the LINEARIZE command, or in fact one method over the other. Since then I have taught this method exclusively at my carbon workshops and many people are using it, and I have also shared it with my friend Sam Wang, who uses it exclusively in crafting negatives for his cyanotype over palladium and cyanotype over gum work. I am attaching the .txt file and a .quad view for anyone who might be interested in this methodology, which is essentially Mike Rosenberg's system adapted by me for alternative printing.

In my personal work with carbon I am using a custom all-gray inkset with a QTR profile I built using the regular QTR tools, linearized with an iOne spectrophotometer. I worked briefly with the Cone Methodology 2 digital negative system. I was pleased with the smoothness of these negatives, but did not like the work flow of having to apply a correction curve to the image, so I decided to start from scratch and create my own profile.

Sandy

[bob carnie]

Yes Sandy I did mean Mike . thanks for your explanation.

[Michael Rosenberg]

Sandy,

Thanks.

Mike

[George Pappas]

Mike,

Back online..appreciate the offer. Will select a file and upload - send you a link. Very much appreciate the gesture. Will take me a week or so as I have a heavy load over the next 7 days.

Bob, look forward to hearing your experiences with Mike's negative/print as well.

We will wait to hear from Walter on availability to test the new Cone system.

Best Regards,
George Pappas

[WalkerBlackwell]

@sandy the new PiezoDN has the same smoothness (possibly more) as classic Piezography Digital Negatives but enables true linearization + custom tuning directly in the .quad

cheers,
Walker