RNP-Array Digital Negatives

[mkochsch]

I settled on the RNP-Array as a name for my alt hybrid negative method. More info and a website to follow...promise!
Part one is choosing a blocking colour for your process.
Print one or all of the attached arrays at your standard print time -- if you want a fill colour right away. If you want to use the densest colour your printer can create, extend your print time until a particular hue reveals itself as a super blocker (and don't assume it's the black square, it might not be).
The advantage to using this method is that the increased colour palette may open up the use of printers and inks for making alt-process negatives which would not normally be possible with other methods. I make no promises. But it works for me.

Red-Green-Array.jpg

Green-Blue-Array.jpg

Blue-Red-Array.jpg

Enjoy.

~m

[Ray Heath]

g'day mk

could you plz explain this concept a little more

why three arrays?

if i was to print them, i assume you mean as a neg for an alternative process or silver gelatin, what would i be looking for in the resulting positive?

why would i need to find and use the densest colour, traditional film negs do not have entirely opaque areas?

why colour arrays and not just greyscale?

what substrate should i print the arrays onto?

[Joe Lipka]

Works for me, too. mkochsch sent me the red/green array when I got stuck working with Dan Burkholder's method. The discovery point was that inks and pigments absorb UV light differently. Dan's color was for the Epson 2200, IIRC, which gave different results when I used it with the Epson R2400. Finding the right blocking color from the red/green array gave me good results.

[Helen Bach]

Don't you also need to know the weakest blocker so that you can use it to smooth out the printer dot pattern?

Best,
Helen

[mkochsch]

g'day mk

could you plz explain this concept a little more

why three arrays?

if i was to print them, i assume you mean as a neg for an alternative process or silver gelatin, what would i be looking for in the resulting positive?

why would i need to find and use the densest colour, traditional film negs do not have entirely opaque areas?

why colour arrays and not just greyscale?

what substrate should i print the arrays onto?

I wanted to cover more colour combinations in the RGB model. Remembering that RGB is additive and that you will eventually be printing using a subtractive colour model (CMY-K) means R needs two dots (Magenta-Yellow), G also needs two dots (Yellow-Cyan) and Blue (Cyan-Magenta). Where these channels combinations overlap you start to get oranges, turquoise and purple colours as well.

You print them as they are do not invert them, although that might be interesting, you will probably get the least saturated colours in the RGB model, the colours that tend toward clear white. When you print the arrays out you may start to see that some colours block UV light (or visible for that matter) better than others. Some squares will print paper white, some very dark and some will be somewhere in between. Paths may emerge, where a column starts dark and then turns white. These are paths of interest.

Why find the densest colour? Good question. And it's debatable for sure. My original hypothesis was that if colour was indeed responsible for blocking UV (i.e. spectral density) why not utilise that fact to minimise ink usage. Find that colour (the superblocker) and just desaturate it until you achieve the correct density for your process. Also, when inkjet printers start "pouring on the ink" to create density that creates other problems -- like banding. But initially the idea was "economy" of ink.

Why colours and not greyscale? Another good question, again debatable. Some of which I just covered in the last paragraph. One thing I wanted was to know that all the inkjets were firing and that my printer's maximum DPI resolution was being utilized. For example, if you use Epson's "Black Only" option you essentially lower the printer's resolution because black dithering isn't going to be the same as the dithering produced when all the heads are firing to produce the tone. However, when the ink combinations overlap you also get colours moving toward the CMY black colour. At some point the printer may also introduce actual black ink too. So, in short, better resolution and possibly less ink drained from the printer. n.b. I routinely use "grey" to make digital negatives but it's the grey created by combining CcMcY and maybe a little K ink.

Print the arrays onto Pictorico OHP for alt processes, or Pictorico High Gloss White Film for Black and White. Although other substrates may also work.

~m

[mkochsch]

Works for me, too. mkochsch sent me the red/green array when I got stuck working with Dan Burkholder's method. The discovery point was that inks and pigments absorb UV light differently. Dan's color was for the Epson 2200, IIRC, which gave different results when I used it with the Epson R2400. Finding the right blocking color from the red/green array gave me good results.

...and the beautiful print pt/pd print you sent me (in thanks) hangs near the main entrance of my home and always receives glowing comments. Thanks again Joe!

~m

[mkochsch]

Don't you also need to know the weakest blocker so that you can use it to smooth out the printer dot pattern?

Best,
Helen

There are two methods here and people need to understand that. The first method is similar to other systems where you find a negative fill colour that effectively sets the density (PDN). These arrays just give you more possibilities.

The second method, if it applies to your situation and needs, selects the hue that blocks the most UV light. On my printer, with my inks for example, it happened to be [RGB 176,176,0]. On some printers this colour might not work they may need to use something closer to black ink. When I found this colour I created a second stepwedge (in black & white) and then filled it (using a new layer set to "screen" mode) with this colour, printed to my process to find a less saturated version of that colour which set the density for my process. (Hint use the HSB option in the info palette to examine the example stepwedges, pay attention to the "Saturation" numbers). That colour becomes my negative fill colour. For example, a colour like RGB [204,204,90] (aka Hue:60, Saturation:56 Brightness:80) may be the correct density for cyanotype on my workflow. Doing things this way is an extra step, the trade off for possibly not having to change ink carts as often. Also all your negatives will be the same colour, which could be good or bad.

Normal Stepwedge:
StepWedgeBnW.jpg

Stepwedge after colour is applied:
StepWedgeFilled.jpg

Part of the reason for me releasing this information is to get more feedback for people who want an alternative method for making digital negatives or who think they can't because they can't achieve the required density with their current hardware or workflow. Also, I've been e-mailing it out to a lot of people in the last month and I'm tired of doing that so I'm just going to post the information here....if that's ok with everybody.
~m

[sanking]

Part of the reason for me releasing this information is to get more feedback for people who want an alternative method for making digital negatives or who think they can't because they can't achieve the required density with their current hardware or workflow. Also, I've been e-mailing it out to a lot of people in the last month and I'm tired of doing that so I'm just going to post the information here....if that's ok with everybody.
~m

I appreciate the fact that you made these arrays possible, and made use of them this past evening in selecting a color that would match the exposure scale of the Ilford papers I tested. for the Epson 1400. That part was very smooth, especially since I use a transmission densitometer and all that was necessary to chose a good match was to read the density of squares on the array and determine which one produced the desired density. I chose a process ES of 1.6 and there were a number of matching squares on all three arrays.


However, I was quite surprised at how different some of the squares of eaqual density printed, both on the OHP substrate, when you visually compared them with a loupe, and when printing. One square right next to another would be perfectly smooth and visually show little grain, while the one next to it, with almost equal reading, would show severe dithering and grain. I don't know why this, whether a problem of the printer driver or the mechanism that Photoshop uses to send data to the printer, but it surprised me a lot. However, resolving why this happens would be important. Of course, maybe this is a printer specifc issue?

Bottom line is that in order to find out what square of same density actually printed the best (smoothest, least grain) involved a series of trial and error testing. I believe tht in the end I got to where I needed to go, but for printers that have sufficient density for the PDN combinations it is very easy to tell which color will print smoother. Course, using the PDN colors was not an option with the 1400 at all for alternative printing since the UV blocking of the Claria inks is very low.

Sandy

[clay]

Michael,

I think this is a great addition to the inkjet negative mix. I agree with you that it is the most comprehensive approach to color determination I have yet seen. I am in the middle of a calibration exercise with a new Epson 7800, and its inks are very different than the 2200, so these grids may come in very handy.

A question:Have you ever loaded up a printer profile for the pictorico and done a soft proof with this grids? I did that a while ago with the ternary diagram approach, and I was surprised to find out that some of the colors were out of gamut for the printer I was using at the time. It may be a moot point, since it won't print what it can't print, but it is interesting to see that some RGB combinations cannot be printed...

One approach I am also fiddling around with right now is using the Imageprint RIP (that I bought for color printing, mostly) with this printer for making diginegs. Their profile library is incredibly accurate and they have a profile for Pictorico glossy white film. The reason this interests me is that their RIP has their own rasterization/dithering algorithm that I think may be an improvement over the Epson driver, and it also has the benefit of being a true 16bit RIP, at least according to their marketing blurb. As you can imagine, the ability to keep your file in 16 bits all the way through the printing stage ought to have some advantages, at least in theory, over the Epson driver's conversion of 16 bit data to 8 bit data for output to the printer. I can't say right now if there actually is any advantage, but I aim to find out. I am going to print your grids tonight with the Imageprint RIP and see what I get.

As an aside, the other advantage that the Imageprint RIP offers to those of us using roll media is that you can tile many images in the same print job and really optimize the media usage with no additional hassle. It is also about 3X faster than the Epson driver....

Clay

[mkochsch]

Michael,

I think this is a great addition to the inkjet negative mix. I agree with you that it is the most comprehensive approach to color determination I have yet seen. I am in the middle of a calibration exercise with a new Epson 7800, and its inks are very different than the 2200, so these grids may come in very handy.

A question:Have you ever loaded up a printer profile for the pictorico and done a soft proof with this grids? I did that a while ago with the ternary diagram approach, and I was surprised to find out that some of the colors were out of gamut for the printer I was using at the time. It may be a moot point, since it won't print what it can't print, but it is interesting to see that some RGB combinations cannot be printed...
Clay

Thanks everybody. No Clay I haven't created a printer profile for the arrays. I don't currently possess a calibration system that can accurately generate a profile. Somewhat moot, perhaps. Photoshop may show many of these colours as being out of gamut in the CMYK space but it still prints something. I'm finding that knowing actual density is nice but it is becoming less relevant since all the calibration is being confined within the program. Stouffer wedges are still pretty handy for keeping me on track though (i.e. detecting flaws in chemistry like fogging, poor clearing etc.)