Can't Believe the Results

[donbga]

I never said a blue'ish neg couldn't be use. Maybe doing Salt Prints might be out of its range, but I'm sure many other processes could be made to fit. The blue square on your 2200 wasn't the same blue square you chose on the 1280, I hazard to guess, though was it? And blue won't fill any better than any other colour in the "shadow" part of the curve where the curve does most of the work.

Sorry if I miss understood about not using the Blue-Red, Blue-Green ink branch but you did say or suggest that the PDN system tends to favor the emarld green colored negatives. So my point is that the PDN CDRP when printed reveals the color density that will work with a specific process.

And of course the blue square made by the 2200 was different than the 1280. FWIW, the 1280 dye inks seems to have a higher atinic <sp?> filtering effect than the pigments do, for some processes. It has been very interesting discovering other anomolies between the two types of ink and how they react so to speak depending on the process being tested.

For example yellow pigments used for gum printing don't exhibit a reverse toe at the base of the process adjustment curve which was an unexpected discovery. IOW, no uplift. This curve shape has been observed by two different people using two different yellow water color pigments using two differnt types of BL tubes and using two differnt ink color branches verifying that the PDN system is outcome based and shows what works.

What the fill color does is to set the density range and the toe...while exposure sets the shoulder. So the curve has less work to do. The curve adjusts everything in between the black point and the white point.

I hope that clarifies everything.

[Michael Koch-Schulte]

Sandy,
I've not forgotten about you! Many things are happening up here, alas it's spring, read: I've better things to be doing than sitting at the computer -- I need a T-shirt that says "I'd rather be somewhere else...". Please don't take my rebutt as an attack on you personally or PDN, which in truth, I have both much respect for. I'm also very thankful that you take the time to engage in healthy debate on the issue of making negatives which I do realize shouldn't be an end unto itself....

There are two issues involved in finding a standard printing color for your process.

1. Does the system provide a Color Density that is sufficient to match the exposure scale of the process with which you are working ?

2. Does the Color Density chosen also allow for all 256 tones of an image to print? This is true for the PDN system but not necessarily true for other methods.

The answer to 1. is Yes. With regards to photographic tonality, 256 is a rather arbitrary number, one which binary-based math has imposed upon our world thanks to computers. (It's also a system which doesn't r-e-e-e-aly understand 50 % grey, 127 or 128, is the eternal question). Since the human eye can only detect tonal changes in the order of about 100 tones anything past 100 that is over sampling as far as what we can perceive. So, the arguement would have to be proven either way that PDN actually provides 256 tones, and this is needed, and that other methods routinely produce much less than 100 tones causing them to fall short of what is required.

Black Ink is not being printed with any of the colors in the PDN CDRP, but this is just a matter of the colors used—not a matter of removing the black cartridge, which would not allow most printers to print anyway. In fact, black ink is used in the areas around the color squares of the PDN CDRP, so "resolution" of the dithering software of the driver is not changed.

The question is: do the black ink nozzles of head fire, how much for a particular colour and is this good or bad. I say black ink is good, and necessary. Given that Epson's current head is made up of 180 nozzles per colour, including black, isn't a system that goes out of its way to exclude black ink in effect excluding up to a 1/6th of the head's ability to produce dot pattern, hence resolution? The fact is Black may or may not be present depending on the RGB colour chosen and the method the printer's driver uses to create/interpret RGB colour using a CcYMmK inking system. How about the differences in gamut between the two models (RGB-CMKY)? Is there physical proof to support the claim that small amounts of black ink cause contrast problems when present on a digital neg? The answer would be to perhaps use a utility like "SSC Utility" to count the number of times an individual colour/cartridge on the head fires while producing a given colour.

1. This is correct in that some colors that people use are too dense and block too much light. If you use one of these colors you have to adjust the curve to compensate. If you do this the curve becomes a destructive one to the number of tones in the image. In fact, you could just blend black ink only and compensate for the excessive blocking with a very severe curve, but this will reduce the number of tones in the image.

Point well taken. I can't say at this point in my own tests if I have seen an overly dense colour causing a contrast problem. Those who find themselves compensating on the toe of the curve are definately spending more time fiddling to get things right. Trying to guess where the white point begins is no fun, and has definately potential for creating bad results or wonky tonality.

3. There is no inherent value whatsover in first finding the color that "has the most 'potential" for blocking UV and then finding a subset that then matches the "target highlight density". This can be quickly and easily done by printng the PDN Color Density Range Palette along with the 31 step transmission wedge when you establish the Baseline Printing Time. Mark Nelson suggests to new users that they do this because they can see the direction the density moves with their printer/ink combination and the alt process they are calibrating as the printing time is reduced to the Standard Printing Time.

The inherent value is that an alternative system (free and open source) might work just as well as a commercial system that people pay $75 for, plus $100/hr. in support for (admittedly a bargain as far as support goes these days). Also, if I'm using the principle of creating image density from choosing a colour which specifically blocks UV and not the principle of creating density by shooting more ink onto the page (I've heard stories of some users on occasion having to dial up more ink using the "Ink Configuration" dialogue, probably the exception and not the rule) I'm saving on a trip to the office supply store to spend more money on ink -- in theory. People who want a supported system should stick with PDN. There's nothing wrong with paying for support and service, especially for something that works. I think in the end, support is the only way Nelson's system will remain economically viable, for him.

Much of the above would be vey obvious to pesons who have actually prited a PDN CDRP? Michael, have you actually printed a PDN CDRP according to Nelson's directions?

Those of us who wish to print out and test the CDRP need look no further than these very forums at APUG, or the Internet. There are examples which people have posted of the CDRP out there. It's no great secret. People can easily make up their own test wedges by creating six step wedges, two for each colour primary colour combination ie. Solid Red 255 with Green stepping from 0 to 255, Red 255 Blue stepping from 0 to 255, then two more iterations for green and blue etc..
I've even made my own variant some time ago with some slight improvements:
http://tinyurl.com/kn8vg
I changed the steps from ten down to eight. More accurate and I find the math is more sensible to work with in an 8-bit world. I'll profess my ignorance with regard to PDN using steps of 10, maybe someone can clarify why this is superior than my tweak to 8 units per step (finer) or perhaps a coarser scale or 16 units per step.
And I'll throw than one right back at you. Have you tried Harmon's method? Or followed the steps I've already outlined in this and other posts? My resulting negs produced with my faux version of PDN were fine, but no better than my system.
Epilog
...and there's still more ways to make negatives out there. For what it's worth Sandy, I'm also working on another way of producing digital negatives, one that if it proves viable, will completely eliminate the need for ever having to apply a curve to a negative. I feel passionately that people should keep thinking and talking about this subject if it's ever going to advance beyond the point where it's at.

~m