Digi Neg Roundup

[CraigK]

Returning to digi-negs after a layoff of a couple of years, and in an effort to get up to speed with the latest advances, I have been delving deep into the archives of this forum and other on-line sources of information and inspiration.

One thing that has struck me is that there are now a lot more ways of skinning the proverbial digi-neg cat than there were when I first started way back when with the Burkholder book.

I thought it may be helpful to perhaps list the various methods being explored and tweaked at the moment.

I will get the ball rolling with the ones that I have come across

1. Burkholder image-setter method
2. Burkholder ink-jet neg method
3. Nelson's PDN method
4. Harman's Colour Ratio method
5. Keith Schreiber's method
6.?


There are certainly others out there. I understand that Bostick and Sullivan are offering a digi-neg service now but do not know if it is via one of the methods listed above or using something they have developed themselves.

Anyway, feel free to add to the list. I would especially like to hear about the differences between the various methods and the reasons you may have chosen one over the others. Personally, I am making my way through the PDN method, mainly because I bought the CD and it seems to be well presented.

 

[clay]

You left out the Ron Reeder and Brad Hinkel QTR RIP method. I have tried this method and I think this may be the best solution yet. They have a brand new book out describing this approach.

Returning to digi-negs after a layoff of a couple of years, and in an effort to get up to speed with the latest advances, I have been delving deep into the archives of this forum and other on-line sources of information and inspiration.

One thing that has struck me is that there are now a lot more ways of skinning the proverbial digi-neg cat than there were when I first started way back when with the Burkholder book.

I thought it may be helpful to perhaps list the various methods being explored and tweaked at the moment.

I will get the ball rolling with the ones that I have come across

1. Burkholder image-setter method
2. Burkholder ink-jet neg method
3. Nelson's PDN method
4. Harman's Colour Ratio method
5. Keith Schreiber's method
6.?


There are certainly others out there. I understand that Bostick and Sullivan are offering a digi-neg service now but do not know if it is via one of the methods listed above or using something they have developed themselves.

Anyway, feel free to add to the list. I would especially like to hear about the differences between the various methods and the reasons you may have chosen one over the others. Personally, I am making my way through the PDN method, mainly because I bought the CD and it seems to be well presented.

 

[CraigK]

Wow, must be the half bottle of cough syrup I took this morning. I completely forgot about the Hinkel/Reeder method...even though I ordered the book last week!!

Thanks Clay for jolting my pyro-addeled brain.

 

[Digidurst]

Another to include in your list is Bjorke's "ChartThrob"

 

[bradhi]

[to mention the competition]
I'd also list elevatordigital.ca - I've seen samples of their work and it looks excellent.
I think most people on this forum will want to do-it-themselves - but there are definitely some pro labs developing out there.

 

[Colin Graham]

I'm a little confused; are these methods specifically for Epson drivers or are they adaptable to other printers?

 

[Katharine Thayer]

Don't forget Michael Koch-Schulte's RNP-Array System.

 

[sanking]

I'm a little confused; are these methods specifically for Epson drivers or are they adaptable to other printers?

Why the confusion?

The issue now is about competing egos. That should clarify everything.

Sandy King

 

[Colin Graham]

I've probably misunderstood something basic...I thought that some of the digital negative methods dealt with specific printer drivers and pigments and might not be as adaptable to other manufacturers. I didn't want to, for instance, start with the PDN method and find out it didn't lend itself well to HP printers.

In rooting around some more I suspect I was mostly thinking of the Quadtone RIP method at Ron Reeder's site.

 

[Joe Lipka]

I started with Dan Burkholder's method when it first came out. When I started making digital negatives with my Epson, Michael Koch-Schulte answered my questions with his method, I switched. So much now depends on the inks you use in your desktop printer, it helps if you have a system that you can test to find your correct negative color.

 

[clay]

I'll answer this question without ascribing motives to anyone.

You are correct in that some methods, notably QTR/Ron Reeder's approach can only be used right now with Epson printers, because the QTR RIP only supports Epson printers.

Other methods, like PDN and Michael Koch-Schulte's, support any printer that has a printer driver that is installable on your system.

And there are yet other methods that people are using but are not talking about publicly for various reasons that use other RIPs such as Imageprint's or Bawhous IJC/OPM to do the ink distribution.

The basic idea is to get the printer to deliver ink to a substrate in densities that allow one to make a print from the negative that matches what is seen on the computer screen. The rest amounts to details in how to achieve that end. Either through the application of curves to the file before sending it to the printer, or the application of curves in the driver itself, it really boils down to getting the ink density right. There are many roads to enlightenment here.

I've probably misunderstood something basic...I thought that some of the digital negative methods dealt with specific printer drivers and pigments and might not be as adaptable to other manufacturers. I didn't want to, for instance, start with the PDN method and find out it didn't lend itself well to HP printers.

In rooting around some more I suspect I was mostly thinking of the Quadtone RIP method at Ron Reeder's site.

 

[Colin Graham]

Thanks for the clarifications Joe and Clay, very helpful. I had started to read Reeder's manual and it became stuck in my head that all methods were RIP based, or had code written for specific drivers. No fault of Reeder's obviously, just me own pyro-addled brain, to quote CraigK. Thanks for clearing that up.

 

[sanking]

I mentioned the confusion of competing egos because of a tendency observed here and elsewhere to tout certain methods of digital negative making as superior to others, and with some caveats I believe such claims are not, and can not be, supported by actual comparison of prints. Hardware, appropriate use of software, and choice of the proper ink system are all important, but less so than individual skill. More about those choices later, but first a brief over-view of the history of digital negatives.

The earliest digital negatives were usually produced on image setters in service bureaus. This was a good way to make high quality negatives, but with the obvious drawback that most people had to farm out the work. Although others were involved Dan Burkholder is the person who came to be best known for this work. This was in the late 80s and early 90s.

As for negatives made with inkjet printers one could categorize the systems.

1. Early systems like those of Dan Burkholder and Keith Schreiber that were highly printer specific, either to the Epson 2000p, 2200 or the 1280. None of these systems provided a method for matching negative density range to process exposure scale so the canned curves usually did not match the process. In most cases printing black composite negatives with these printers gave too much maximum density, on the order of log 3.4 and higher, so the adjustment curves had to be very steep to adjust for the excess of density, which of course resulted in a reduction in the number of tones in Photoshop.

2. Color ratio systems such as Mark Nelson’s PDN, Michael Koch-Schulte's RNP-Array System and Clay Harmon’s color ratio system. The advantage of these systems over what Burkholder and Schreiber were doing is that the color system provides a method to visually match negative DR to process ES, which means that the adjustment curve can take advantage of most of the tones in Photoshop. The color choices of PDN do not produce enough UV blocking density with most current Epson printers without adding ink density, but it does work well with the current HP and Canon pigment printers. The systems of Clay and Michael allow adding black to increase density. These color systems are used with process adjustment curves, and there are various ways to create the curves. Chart Throb is not a system, but a method of creating an adjustment curve.

3. The newest system is Ron Reeder and Brad Hinkel QTR RIP method which is outlined in their book, and a subsequent chapter is now available on their web site that better explains QTR. This system has a lot of potential, but has without question a steeper learning curve. This system does not use a process adjustment curve but depends on producing linear output by adjusting output from each ink.

Another method which words very well is to simply print in composite black, and either adjust your process with exposure controls, or adjust printer output with plus or minus ink density. Persons with access to a UV densitometer can determine the right printing density in a very short period of time with this method, though subsequent adjustments may need to be made to optimize results. I have seen some extremely smooth platinum and palladium prints made with the Epson R1800 and the 3800 using nothing but the composite black inks controlled by the Epson printer driver. So smooth in fact that I seriously doubt that the QTR RIP system can do any better with pt./pd. given the inherent limitations of the process itself.

I am not a proponent of any system, but for those looking for a new printer the following three options are all good ones IMO, and two of them are fairly economical.

1. Epson R1800 and replace the ink set with the Piezography K7 set and run through the Harrington QTR. Simplify by using an adjustment curve once you determine correct ink density setting for your process, or use QTR to linearize output. You could also just use the composite black ink with the Epson driver, but negative DR is a bit low for palladium with no restrainer.
2. Epson 3800 and print with the Epson driver and composite black. Add a little ink density for palladium.
3. HP 9180 and print in either composite black or in color with PDN.

Options 1 and 3 are the least expensive but limit to 13” width. Option 2 is more expensive, but prints up to 17” wide.

Sandy King

 

[Colin Graham]

Wonderful info, many thanks Sandy. This would make a great sticky thread.

 

[Ron-san]

confusion of competing egos

Sandy King

Sandy--

As one of the potentially "competing egos" out there I would like to say that was a masterful summary of the history and current state of the inkjet neg field which you wrote.

I continue to pursue the Quadtone RIP approach because it offers COMPLETE control over ink deposition. Any time I hear or think of a new wrinkle in neg printing I can make the inks dance and try it out if I wish. BUT, can I prove to you that my approach results in a superior print? I am coming to think, probably not. We seem to be coming to a point where we are limited by print head quality more than by the specific method we use to set neg contrast and linearize midtones. It is great to have options, though, is it not??

Cheers, Ron-san

 

[sanking]

Sandy--

As one of the potentially "competing egos" out there I would like to say that was a masterful summary of the history and current state of the inkjet neg field which you wrote.

I continue to pursue the Quadtone RIP approach because it offers COMPLETE control over ink deposition. Any time I hear or think of a new wrinkle in neg printing I can make the inks dance and try it out if I wish. BUT, can I prove to you that my approach results in a superior print? I am coming to think, probably not. We seem to be coming to a point where we are limited by print head quality more than by the specific method we use to set neg contrast and linearize midtones. It is great to have options, though, is it not??

Cheers, Ron-san

Ron,

Thanks for your comments. I thought some over-view might be useful for some of the readers new to making digital negatives.

It is nice to have options, and I personally am in a transition state trying to figure out the best approach for my own work. I need a "better" negative for my carbon transfer printing on smooth surfaces than what is required for pt./pd. My belief at this point is that the very best option, if 13" wide is enough, is to dedicate an Epson R1800 to digital negatives and replace the Epson inks with with a Piezography K7 ink set and run with the Harrington QTR RIP. Of all the prints I have seen from digital negatives the smoothest ones have been from this printer with the Epson driver printing in composite black, and I think the small 2-picoliter drop size may have something to do with it. Piezography would offer a small improvement IMO, if not in smoothness or grain, but in the possibility of more UV blocking.

BTW, I have been very busy recently scanning some of the work I did earlier in Spain and more recently in BC and have not had a chance to resume the comparison testing we discussed. However, I just out your materials and I am going to give this a try today or tomorrow. BTW, I measured the UV density on one of the negatives that you sent (with the horses in foreground), using the image of the sun which produced paper white, and it measured log 3.47 on my densitometer. That seems pretty consistent with what you are measuring with your X-Rite.

Sandy

 

[Katharine Thayer]

Sandy, I agree with others that your summary of digital negative categories is well-done and useful; however one very slight quibble: I wouldn't say that Dan Burkholder's method was "highly printer specific." I never used the same printers, or sometimes even the same category of printers, that his sample curves were made for, and yet I found his instructions quite general enough to be useful for my situation. His method was really very much like what Michael posted the other day when he walked us through a way to create curves by hand, except somewhat less sophisticated, as you point out. But it was a way to create a curve that matched whatever printer you had to whatever you were printing and your own way of printing it; in other words, you printed the test image/step tablet using your own equipment and your own process and your own paper and everything, compared the printed test image to the original, adjusted the curve and so forth until you got as good a match as one could. I never saw it as a method intended to be used with any specific printer, and I don't think he ever presented it that way. When he expanded his method to include inkjet printers as well as imagesetters, after inkjet printers starting getting more possible for digital negatives, he provided curves for one or two popular printers just as a starting point for those owning those particular printers, but not owning one of those printers wasn't any particular drawback to using the method. This is not to say that I would recommend Dan's method now over currently available methods, but just as a matter of historical accuracy.

 

[seans]

...
2. Epson 3800 and print with the Epson driver and composite black. Add a little ink density for palladium.
...

Sandy King

Sandy - as others have stated - thank you for the succinct summary.
One follow up question -

As someone who has moved away from PDN and just completed a successful calibration of QTR on a 2200 - I am intrigued by your statement on the 3800. I understand the approach to increase ink density - is it implied that I still need a correction curve in PS to linearize?
I did not see that in the list of options so I want to ensure I understand.
Thank you
Sean

 

[sanking]

Sandy - as others have stated - thank you for the succinct summary.
One follow up question -

As someone who has moved away from PDN and just completed a successful calibration of QTR on a 2200 - I am intrigued by your statement on the 3800. I understand the approach to increase ink density - is it implied that I still need a correction curve in PS to linearize?
I did not see that in the list of options so I want to ensure I understand.
Thank you
Sean

You will need to apply a correction curve in PS for the Epson 3800 if you use it with the Epson driver.

The only time you will not need a curve is when you use QTR RIP or another RIP and linearize with the RIP.

You can create the curve with the tonal palette in PDN just as you would for the color negatives.

Sandy King

 

[MVNelson]

Great summary Sandy...just so happens I am doing calibration and curves with PDN using the 17"Canon iPf5000 even as I read your post...so far so good...

Miles

 

[clay]

I think Sandy's summary is pretty good. I would throw in one more factor about the QTR approach that I just stumbled on: it may be possible because of the individual control of each ink to minimize the infamous 'roller blind defect' (or banding) that seems to pop up on these printers after a few years. I have a four year old 2200 that was starting to show a little of this problem with the green negatives. But by using a custom QTR profile I was able to use mostly black and light black inks with some other helper colors and get a nice smooth negative again. But as Sandy mentioned, the learning curve on QTR is not for the faint-of-heart. But it is an ideal system to take someone else's hard work and just drop their curve into your system. :^)

 

[Kerik]

Just to supplement Sandy's comment about the 3800 w/Epson drivers (the approach I am using)... I do apply a correction curve to linearize the output, but the curve I derived is very "gentle" compared to many curves I've used previously (Burkholder/Schreiber) and curves students have brought to workshops using other methods such as PDN.

I do plan to pursue the QTR approach, but what I'm doing now is working so well in terms of print quality and tonal smooooooothness, I'm taking the approach to not fix something that ain't broken.

 

[Colin Graham]

3. HP 9180 and print in either composite black or in color with PDN.

Sandy King

Sandy, was curious about the composite black method- is this the same as desaturating in RGB mode and printing in color? Thanks.

 

[sanking]

Sandy, was curious about the composite black method- is this the same as desaturating in RGB mode and printing in color? Thanks.

Yes.

Sandy

 

[timeUnit]

About the Epson R1800,

Would it not be possible to get really high UV densities with the "usual" metod, i.e. colorized negs? Or is the Ultrachrome HiGloss inkset substandard to the K3 inkset, in terms of UV density.

And: the new Epson R1900 has an orange ink. It might be very good for UV blocking, am I right?

Thanks!