About the Technology

Updated: 9/08
Many people have expessed an interest in the technology behind my artwork. Here are all my secrets for anyone who wants to take a shot at it themselves.

I've learned (and re-learned) some things in the past year - a lot has changed.

Cameras and Film

Most of the images on this site were taken with an 8x10 camera. It's a Century Universal, wooden, made back in 1910-1920. The 12" Goerz Dagor has a unique character all its own. The film was either Tri-X or FP4 (not FP4 Plus). These films scan well and delicate and sensitive prints almost make themselves.

I have done a great deal of research trying to get negatives as good as the ones I shot back in the day. It turns out that in the 1990's they changed film. I have no idea how they got together on this, but every major film manufacturer has now shifted their priorities and processes for making film. Contrary to my first assumptions, it isn't all about the amount of silver in film that caused this change. In fact, much of the issue is that they got better at making film emulsions. They also started focusing on sharper grains, due to the popularity of films like Tech Pan. Finally, they changed the response curves, creating a bump in the center of the curve. I don't know why they did this, except it was probably to match their response better to available papers.

All of this resulted in film that is less sensitive. This is arguable, of course, in fact its being argued about all the time. A lot of it depends on the definition of the word sensitive. One theory posits that the problem comes from the fact that emulsions used to consist of many sizes of film grain. They would react to differing amounts of light coming thru the lens, based upon their surface area. Now that they have gotten better at making emulsion the grains are more often all the same size. Therefore, less sensitive to "all light" in the same area. The other theory is that they simply changed the response curve and it doesn't look like it used to. I think it is a little bit of both.

Whatever it is, photographers who grew up in traditional processes prior to 1990 often find it hard to figure out what is wrong with film. It certainly doesn't look like it used to. It takes a lot, often many years, to be able to look at a color landscape and see in black and white. It is a process of looking at a scene, going thru development and printing and looking at the final print - a lot of times. If at the end of the process, they change everything on you its very difficult to adjust.

The good news is that these days I am very happy with my results. I achieved a new balance by going back to my roots. I had hoped that a scanner capable of top resolution would be able to get as much "juice" out of a 4x5 negative as it would out of an 8x10. There are limtis to resolution here and there and the data pointed in that direction. This was not the case.

The illustration of this conclusion comes down to the following example. A 4x5 camera that takes a normal landscape photo with a telephone pole in it might use 1/4 or an inch of film to describe the pole itself. The same image, framed the same way, using an 8x10 camera would use a full inch of film to describe the same telephone pole. Regardless of lens resolution, or anything else, this is a substantial difference. The pole, taken with the 8x10, would look far more 3 dimensional in comparison, in fact it would come off the page.

I decided to purchase a new camera, as the Century Universal 8x10 was hitting over 100 years old. After much deliberation I chose a Canham lightwieght. Keith Canham makes cameras that are simply works of art. It is an inspiration to use such a finely crafted camera. I also picked up a trio of modern Nikon lenses so that the response curve would match more closely to these modern emulsions. I also suspected some flare in my system and decided to replace the old 12" Dagor. I am still looking for a Rodenstock 300 Sironar S.

The film I am currently using is Efke 25. It's a little slow, but the quality is excellent. I am developing it in D-23, the same developer I used in the 1970's. As smooth as it gets. Research is still continuing with the new 8x10 so keep checking this area for the latest.

The 4x5 Ebony now stays home to do commercial work and fine art reproduction. It's as sweet a camera as I have seen, titanium gears and all. Like butter.

Scanner and Scanner Software

The scanner is an 8,000 dpi Aztek Premier drum scanner, the finest scanner ever made. The key number, however, is the optical resolution, which is also near to 8,000. I Have a lot of choices of different apertures, which allows me to match each scan to the film's grain.

My Arithmetic Lesson:
A few years ago I purchased a Mamiya 645 fitted with a Kodak DCS ProBack, which gave me 16 megapixels. I was pretty impressed with this gadget but sold it after learning some basic arithmetic regarding numbers of pixels available for printing.

The 16 megapixel camera had a 4,000 by 4,000 chip. Turns out, if we can remember back to Junior High Math, that if you multiply a number with 3 0's by another number with 3 0's you get a number with 6 0's - or something in the millions. Few of us are rich enough to know how to deal with a concept like millions, so it just seems like a huge amount. We can hear Carl Sagan saying, "Billions and billions of stars."

Back to two-dimensional reality. Along a single dimension at the top of a print, that's 4,000 pixels for the whole image. If you want to make a 20 inch wide print you divide the 4,000 by 20. You end up sending 200 dpi to the printer. In other words, awful. My tests show that the "sweet spot' for most of the top inkjet printers is 720. There are a lot of people who would disagree with me about this. 360 is generally accepted as terrific, and I would agree.

Of course, it's fine if you never want to make more than a 10 inch print. Most commercial photographers wll never exceed that - their target is a magazine and almost all of them are 8 1/2 by 11 inches. Many artists, however, will want to print larger. My prints start at 16x20 and go up to 32x40. If you print at the 16x20 size or more a digital camera will not supply the quality you need. Every darkroom worker will tell you that if it isn't there in the negative, it's not going to be there in the print. This basic rule still applies.

Back to arithmetic for just a moment... I typically scan a 4x5 at 4,000 dpi. With a drum scanner, that's 4,000 pixels for every inch, so the 5 inch part of a 4x5 yields 20,000 pixels. If you make a 16x20 inch print, you divide the 20,000 by 20 and you get 1,000 dpi. Even at 40 inches, one am still at 500 dpi. They are sharp, grainless, smooth, and stunning. Oh, yeah, the 16,000 x 20,000 pixels of my 4x5 scans multiply out to 320 megapixels. I do my 8x10's at 2666 dpi and end up with 568 megapixels. Which would you want - 22mp or 568mp? Do you think there might be a difference in a large print?

I would be remiss if I ddn't mention that there is some real magic in the scanner software, Digital PhotoLab, by Aztek. It has a patented feature called CMS, which essentially goes one step further than other scanning software.

The process of scanning begins by mounting the negative on the drum. The scanner then previews the image. One uses curves and/or levels to modify the image to look like the desired result. A "CMS" file is created with the changes you have made and a host of other specifics. The scanner uploads this file to hardware registers in the scanner before the scan occurs and the scanner uses this file to manage what occurs in the scanning process. It is very different from scanning a RAW image and then applying curves to it after the fact. It effectively creates a different scanner for each image.

There are standard CMS files for many different types of film, but I generally use this feature to create a different CMS for each image - at least in black and white, where you can really see the difference.

Editor

There is only one fully featured editor these days - Photoshop. For those of you who are interested, I use a Mac G5 Quad, with 13 Gigs of RAM for doing the editing. I recently built a SATA RAID array and now have 7 Terabytes of space. (I'm not running out for a little while!)

The scanner software and the RIP software both work on a PC I built, a 3 Ghz Pentium 4. Despite the fact that comparing a Mac to a PC is like comparing a Maserati to a jackass, it's actually good to have the scanning and printing software on a different computer. The editing process, especially in 16 bit, takes a tremendous amount of processor resources. You wouldn't want to try and do all three operations on the same computer, Mac or PC.

The files sizes are fairly large. The 8 x10 images come off the scanner in 16 bit RGB, at about 3 gigabytes apiece. When taken to Grayscale, after initial editing, they are about a Gig. Add a couple of adjustment layers and they can get large.

Printer, Inkset, Paper and RIP

I have two printers, now both 54 inch wide Roland Pro 2's, one for color and one dedicated entirely to Black and White. I am now an Epson-free shop! They are exceptional pieces of machinery. They are made of metal, and have much better consistency than the plastic, smaller printers.

I run a d'Vinci system for color, the widest gamut, most accurate and smoothest color of any printer on the market.

For Black & White I start with the Cone Piezotone inkset and formulate my own custom dilutions. The new 12 cartridge Roland allows me to put a 6-dilution set of warm tone on one side and a 6-dilution set of cold tone on the other. I mix the two to create a beautiful neutral. I can also split tone any channel, in any direction and create custom sets for any black and white photographer.

Six dilutions of black give the prints a very long tonal range, the equal of which is found only in platinum prints - long known as the king of Black and White processes.

The inks are fully archival. They have a pure carbon base.

My favorite paper is Hahnemuhle PhotoRag 308. It has a superb density range, and creates very 3 dimensional prints. It is 100% Cotton fiber, so it is fully archival. Very nice paper, feels wonderful in your hand.

I use the new d'Vinci software, by Ergosoft, for printing. It is an upgrade from StudioPrint/PosterPrint that can control all of the 12 slots in the Roland individually, or fire them all together.

Calibration

I use a Gretag EyeOne spectrophotomer coupled with an iO Table. The Gretag EyeOne is spectacular, on target, accurate, everything you'd want from such a device. The iO Table allows me to create and read profiles with ease.

I use ErgoSoft's ColorGPS software for profiling. I am read 2500 or more patches to make an accurate profile.

In Summary

When I started changing over to the digital medium I had 5 people utter the words' "Oh, you won't succeed unless you do the following....." The difficult part was that they all disagreed about what followed. There are obviously different paths to the same result, so I'll refrain from suggesting that my way is the only way to succeed.

In addition, you will find that there are titans of the industry - those who have extended the inkset, or created a particular piece of fantastic software - and while there is much to learn from them they all disagree as well. Further, they are all nuts - and I say this with great affection. They apparently had to be very opinionated to succeed in their creations, they had to make some assumptions and stick by them. However, some of them don't even speak English, its some gibberish about gamma or arguments about color spaces and targets or micron slices. The long and the short of it is that we all have to do our own testing and trust our results.

It has taken a long time to learn each of the different parts. There were a lot of new concepts to understand and embrace. If you are just starting out, give yourself some time. You have a lot of patches to print (and read) before you can print anything at a consistent, professional level.

However, now that I have made it this far, I can say without hesitation that I am thrilled with the results.

Lenny Eiger
eiger@eigerphoto.com