Can't Believe the Results

[billschwab]

I've been laboriously building my own set of curves for making digital negatives... testing, tweaking, retesting, etc. I do believe I am finally there and have to admit being stunned by the results. My work has a lot of smooth gradations in the tones and it has been a nightmare trying to smooth the steps I was getting in these areas. I was giving up hope that it could be done but, having had a breakthrough of sorts, I am so far pleased. I now have a curve that seems to fit just about any image I throw at it coupled with my sensitizer formula and burning time. Sweet consistency. Admitting it took me a bit of work to get here, but can it really be this easy?

Also... has anyone else had such flat results with Burkholder's curves "out of the box"?? I am now nowhere near the starting points on his platinum curve and ended up abandoning them entirely. That set me back a good deal of time right there. Just curious as I thought there must have been something I was doing wrong.

Bill

[Michael Slade]

I'm using my own curves, and they're image dependant, not 'set' in stone. I've never used the same curve twice.

Tell me more about your testing and results. Sounds very interesting.

[Kerik]

So, what was the breakthrough?

[billschwab]

So, what was the breakthrough?

Some of the breakthrough was overcoming my own stupidity, but much of it was dumping the Burkholder method and sticking to what I had learned about building curves and profiles for different papers and ink mixes in the septone process with my Epson 7600. Hard to teach an old dog new tricks. Another was abandoning the carbon pigments and the 7600 and printing with epson pigs on the 2200. The quadtone pigs I am using on the 7600 wasn't giving me the dmax I wanted and I was getting fogging in the borders and highlights. I may switch back to the epson pigs on the 7600 if I decide to make really large negs.

Once I finally started getting a good dmax in the negs, I was able to start at the beginning again and build a good, consistent workflow for my testing.

Tell me more about your testing and results. Sounds very interesting.

Well I'm no rocket scientist Michael, so if what I am saying is old news, forgive me. I am also not willing to say my results with multiple images are "set in stone" as I am sure once I have gotten over my success, I will be looking with a more critical eye. It would seem that some tweaking between images would have to be made simply because I can't really believe that this could be that easy. However, I will say that I have now had fairly consistent results using several different types of photographs.

Basically I started by figuring my dmax using a consistent mixture of sensitizer, paper and light source printed through the base + fog of the OHP material. My test was done with 30 second intervals of light and reflective sensitometry readings after the paper was completely dried and flattened. With my light source, I found I hit the maximum density at 9 min 30 sec. Once I had that, I set-up my inversion layer by doing a simple straight-line inversion of my file and step tablet and then outputting the negative. Once a print was made and dried using the exposure for dmax, I plotted the step tablet with the sensitometer. Very small adjustments were then made to the curve and the file output and printed once again. This was repeated many times until I was able to make my print approximate what I saw on my monitor.. I now had a curve I was happy with while using the least challenging image used for my tests. I then moved on to an image that was more difficult having a much wider range of tones, applied my curve, output the file and printed the neg. Whereas the less challenging image looked great with the curve, the more challenging displayed some faults in the smoothness between tones. There were places where there was an obvious step in tones where there should have been a smooth transition as well as some clipping on the high and low end where shadow and highlight detail was being lost. I once again began the fine tuning process in these areas until I was satisfied that I had smoothed the steps between tones and got the most out of the shadows and highlights. I once again went back to the original, less challenging image I began my tests with and applied the new, adjusted curve and found the print to be virtually unchanged from the first successful test of that image. I have now run that same curve on 10 different images and am getting what appear to be consistently good results across the board. As long as I have an acceptable image on my monitor, I can come very close to reproducing that image on paper using this same curve.

All pretty elementary I would think, so forgive me for going on at length. As I said though. I am reserving my celebration until my excitement passes. We'll see what I think then.

Bill

[Michael Slade]

No, it sounds great! Your testing is much more scientific than mine is. I enjoy reading how others arrive at their successes, it gives me ideas how I can be more analytical with my own printing.

[Joe Lipka]

My problem has been with the Epson R2400 printer. Good highlights, good shadows but dead flat between the 30 and 70% steps. PM me so I can get into more boring details on where I am stuck and how you did the breakthrough. I think the problem might be that the R2400 pigments behave differently than the printers you use.

[donbga]

My problem has been with the Epson R2400 printer. Good highlights, good shadows but dead flat between the 30 and 70% steps. PM me so I can get into more boring details on where I am stuck and how you did the breakthrough. I think the problem might be that the R2400 pigments behave differently than the printers you use.

Joe, Bill, and Michael,

It's not clear from your posts if ya'll are printing black only negatives or colorized negatives. Colorized negatives (regardless of whose method you use; Burkholder, Schrieber, Nelson or Harmon ) will require less radical process adjustment curves.

However to build good curves requires some iterative testing that demands consistent processing technique as Bill points out. The Epson 2400 is easily capable of making good negatives. I normally use a 2200 or 1280 but I've been testing with a friend's 2400 and it seems to work fine though I've not made a good negative from it yet (this is simply due to the logisitcs of driving to my bud's home and begging him to make a negative for me ).

In the end you have to pay your dues testing.

My 2 cents,

Don Bryant

[Michael Koch-Schulte]

Did you do a "pre-selection" of your blocking colour? Does your curve start and end in the corners or are you "lifting" the toe to compensate for too much blocking ink? The reason I'm asking is that once I realized (my own breakthough) that picking a blocking colour that "just" printed white and not more density than I needed, my curves became much easier to calculate because they started and ended in the corners (when I picked the right blocking colour which effectively set the density). Because I use a scanner and not a densitometer it became a matter of scanning my printer's density wedge and setting the white point and black points in PS to zero and 100 respectively. The curve doesn't touch the black or white points . The curve's soul purpose becomes to adjust the relationship between tones. Just my two cents.

[Joe Lipka]

Don -
I am using a color layer that came with Dan's inkjet negative companion. I think that the color layer (orange-ish color) is competing with (and beating the daylights out of ) the density of the black inks and causing the big flat spot in the middle of the curve. I have a string of tests (about 20) that have convinced me that the color layer is the problem. I have read that the PDN method of digital negs mention a green cast to their negatives. Maybe that color doesn't fight the black in the negative as much as the orange cast does. Just interested to see how others have made the breakthrough.

[Michael Koch-Schulte]

Joe,
You must find a better blocking colour. I've been experimenting with different blocking colours on different printers, and what I've found is that properties of negative density as it relates to colour is largely dependant on the type of inks the printer is using. Specifically, dyes inks act more like colour gel filters and pigmented inks act more as opaque blockers. Dyes inks are more true to "spectral density" theory and pigmented inks are more prone to block when black ink enters into the equation.
What I did was print up a stepwedge which ran a broad sample of all red and green colours through it. I call it the Red Green Wedge (I ignore blue because I'm filtering UV light and it has very little effect as UV is actually ultra blue.)

Red Green Wedge Example
http://www.mondotrasho.ca/pages/alt-.../Red-Green.jpg
Basically, it shows you what pure (saturated) red, pure green and everything in between on your printer does. Print a stepwedge similar to this to your negative material, contact print to your emulsion and it will show you what colour is the most effective UV blocker your printer can produce. On my Epson dye printer I noticed a distinct "spot" of density around R176,G176,B0 (strangely a "pyro" stain colour, go figure). Others have reported that their pigmented printers trend toward the black ink area of the wedge. Once you've established where your printer is operating at its most efficient colour you can find the colour that will print to the density you need by making another stepwedge which steps "down" the Saturation and Steps "up" the Brightness. I switch my Info box in PS to HSL mode and it becomes more apparent.

Graduated Stepwedge Example
http://www.mondotrasho.ca/pages/alt-...R176R176B0.jpg
similar to this. This is the wedge which pinpoints which colour I need to "just" have a dense enough colour to print white based on my emulsion, exposure and paper.
Once I've established that colour it then becomes my "overlay" colour for the process, say VDB for example. I fill a separate layer in "Screen" mode and print my curve stepwedge.

VDB Curve Calculator Stepwedge Example.
http://www.mondotrasho.ca/pages/alt-...s/VDBWedge.jpg
If my exposure time is correct and my colour choice is correct I'm pretty close already now all I have to do is create a curve based on the differences between tones instead of fighting with the highlights, which is what I think is happening to you. You're using too much of your curve trying to control the highlights and it produces a long "flat line" in much of the midtones of the curve. Anyway if anyone's interested in the finer points of this workflow or the original wedges send me a note and I'll try to clarify in more detail how I make my negatives.

~m