Scan a black-and-white print confident the image on the monitor will look like the print that was scanned. Then print it with an inkjet printer assured the print will look exactly like the screen image with any ink and paper combination. If there is no intermediate manipulation, the inkjet print will be a nearly perfect density-for-density match to the original silver print. Scan once, print once, with no wasted time, paper, or ink, and no aggravation.
Developing the ability to do this with inexpensive scanners and printers was our first goal as we entered the digital black-and-white arena. These articles explain our procedures. In Part I we described the goals in greater detail and discussed monitor calibration tactics to apply to ensure your screen shows the necessary tonal information. Our procedures require the use of a densitometer, but we also explained how you can make your scanner function as a surrogate for a densitometer. In this article we’ll focus on how to scan a black-and-white print into Photoshop and obtain a screen image that is visually faithful to the original print on the first try.
Tones in the original print are easy to characterize: Measure their densities with a densitometer, if you have one, or measure their Log ED values with your flatbed scanner (as explained in Part I). After scanning the print, tones are even easier to measure. Move Photoshop’s Eyedropper Tool over the tone and note which of the 256 possible Brightness Values has been assigned to it. If the correct Brightness Values are assigned to each tone, your eye accepts the screen image as “looking like” the original print. What determines which Brightness Value is assigned to each tone? You do. Adjustments made to the scanner’s Highlight, Shadow and Gamma controls before you hit the Scan button determine the rules for converting densities in the print to Brightness Values on the screen.
However, the print image and screen image never match perfectly. One is a reflected light image, the other is a transmitted light image. It’s like comparing a black-and-white print to a black-and-white transparency. But if you get the viewing conditions right, and map densities to Brightness Values correctly, you can get to where your eyes accept the print and its screen image as interchangeable. This is where we’re going. The first step is getting the viewing conditions set up correctly. Scanner adjustments come second.
 |
 |
Fine-tuning the workstation
The monitor should be in dim surroundings, situated so there is no glare, flare or extraneous reflection on the monitor’s surface. A good way to test for adequate dimness is to create (in Photoshop) a large square that is 100% black and place a somewhat smaller square that is only 95% black inside this. The smaller square should be discernible (though not by much) within the larger darker square. If it isn’t, you will need to check the monitor’s calibration or further reduce the ambient light level surrounding the monitor. This test may be easier to do if you use the Tab key to hide Photoshop’s palettes and toolbar, and strike the keyboard’s F key a couple times to cycle to a black surround.
However, at the same time the monitor is in these dim surroundings, the original print should be viewed in rather strong light—strong enough to discriminate small tonal differences in the extreme shadow end of the print’s tonal scale. The print also should be near the monitor screen, so your eyes can move quickly between the two. A gooseneck lamp (or something similar)— perhaps with a homemade baffle—can provide the brightness for print inspection without spilling light on the adjacent monitor (you may want to construct a baffle around the monitor). Once you’re satisfied with the viewing arrangements, the scanner settings become fair game.
To illustrate, we’ll refer to the operation of our Epson Perfection 1640 SU Photo Scanners. Many other scanners have comparable controls. To scan a print into Photoshop with this unit select “File: Import: Epson Twain 5” from the Photoshop Main Menu. This brings up the opening screen (Figure 1). Choosing the indicated button opens the Image Controls screen (Figure 2). The key controls—Highlight, Shadow and Gamma—are flagged in Figure 2. (Ignore the “Exposure” slider because it apparently serves only to duplicate the function of the “Highlight” slider.) For the Highlight, Shadow and Gamma, you can enter specific numeric values in the boxes or let the software assign values. We will identify values to enter manually, but it is fair to ask what is wrong with letting the software make the assignments.
Let the scanner decide
To illustrate the problems that can occur, you’ll need an old silver test strip that has uniform stripes in a variety of densities. The strip we used is shown in Figure 3. The blackest step on this strip had a density of 2.15. We scanned the strip into Photoshop, letting the scanner assign values for Highlight, Gamma and Shadow. Then, we covered the D=2.15 step with a piece of white paper so that the darkest visible step had a density of 1.75. We repeated the scan.
This process was repeated several more times, essentially scanning (with the scanner’s choice of settings) a series of strips of ever-decreasing maximum density. This series of scans mimics a real world array that might include a toned glossy print with a rich Dmax, a high-key glossy print lacking strong blacks, a matte-surface print with a lower Dmax, and an old photo in need of restoration due to its extremely bleached-out Dmax. We want all such images on the screen to look exactly like the originals that were scanned. If a strip with a weaker Dmax is scanned, its Dmax should appear lighter on the screen.
 |
What happens when the scanner selects the settings? Look at Figure 4. Every test strip we scanned has its own unique density- Screen Brightness curve. The scanner tried to force the darkest area of a strip to appear very black on screen, even if the darkest area was nothing more than a medium gray. This is not what we want!
However, Figure 4 contains a peek ahead to what we do want. The curve labeled “Ideal” will emerge as the one and only way we want a print density to translate to a screen brightness. This is the curve required if our eyes are to accept an on-screen image as “looking like” the print original.
As an aside: If you use scanner Log ED values in lieu of densitometer density readings, see Figure 4a. This is the Log ED counterpart to Figure 4. While the test strips were visible in the scanner’s Preview mode, we read ED values for all the visible steps with the scanner’s eyedropper tools and looked up the logarithm of each ED value. As we moved to strips with markedly larger minimum values of Log ED, the Screen Brightness Values decreased only slightly. As with Figure 4, this indicates the middle and darker tones of the prints with weaker maximum blacks were forced to look excessively dark and ponderous. This chart is included to illustrate that whether one works with density readings from a densitometer or Log ED readings from a flatbed scanner, the conclusions are the same. Throughout this article we will reference only density measurements, but this is only a matter of convenience.
Overriding the default
If the scanner’s choice of settings do not yield a screen image that looks like what was scanned, how do you find the correct manual settings? The clues are in Figure 5, which shows the influence on the Density–Brightness Value relation when changing the settings individually. These charts show which print tones each slider controls. Changing the Highlight control has major impact on the Brightness Values assigned to the lightest print tones, no impact on the darkest tones, and progressively less impact as tones vary from lighter to darker. The opposite happens with changes to the Shadow control. The darkest tones change on the screen, becoming lighter or darker, with no impact on the lightest tones. If you move only the Gamma control, the lightest tones are unchanged, the darkest tones are almost unchanged, but the middle tones are altered profoundly. Understanding the shapes of these charts makes it easy to dial in optimum settings for each control.
We suggest you adjust the Highlight setting first. Select a print with the brightest, whitest whites you expect to encounter in your work. Do a Preview Scan. Leave the Shadow and Gamma controls where the scanner puts them, but vary the Highlight value, completing a scan at several different settings. With Photoshop’s Eyedropper Tool, read the Brightness Value assigned to the print’s Dmin (perhaps a portion of the print’s border) for each scan. You want a setting that forces this—your whitest-ever Dmin—to have a Brightness Value in the range of 250 to 255. We locked on to a Highlight setting of 245. This setting forces a Dmin value of 0.05 to appear on screen as a Brightness Value of 255. Darker Dmin values will produce slightly lower Brightness Values—provided we always scan all black-and-white prints with the Highlight set to our personal optimum value of 245.
Finding your optimum position for the Shadow control is similar. Select a print with the darkest, blackest, richest Dmax you expect to encounter. Perhaps this is a selenium-toned fiber print, or perhaps it’s a chromogenic one. Proceed as above, using the optimum Highlight setting you just determined, the scanner’s choice for Gamma, and a variety of values for the Shadow control. The goal is to have a Brightness Value in the range of 0 to 10 for this blackest-ever tone. We identified a value of 12 for the Shadow setting; this forced a print Dmax of 2.5 (from a chromogenic print) to yield a Screen Brightness of 0.
 |
This leaves the Gamma control, and this is where the monitor and print viewing conditions become crucial. Preview Scan the print, set the Highlight and Shadow controls to the values determined, then complete several scans using different Gamma values. View these scans on the monitor alongside the print original, and select the screen image your eyes tell you is closest in overall appearance to the original print. Do this with a variety of prints, seeking the Gamma value that most often looks best. In our case, we consistently selected Gamma values in the range of 1.8 to 2.2, tightly clustered at 2.0. Accordingly, we chose the Gamma value of 2.0 as our standard for all black-and-white print scanning.
From Figure 5, recall that the choice of Gamma had a very slight influence on how dark tones reproduced. You may want to repeat your Shadow control setting test, using your values for Highlight and Gamma. In our case, this prompted us to tweak our preferred Shadow setting from 12 to 14.
To summarize
That’s it! You now have a set of values for Highlight, Shadow and Gamma (ours are Highlight = 245, Shadow = 14, Gamma = 2.0). Use these values to scan all black-and-white silver prints—every print, without exception—to get consistently excellent correlation between the appearance of the print in strong light and the monitor’s image in dim surroundings.
Figure 4 showed how density translated to Screen Brightness when scanning strips of varying Dmax, allowing the scanner to use its default settings. Figure 6 shows the same scanning experiment using our manually entered 245/14/2.0 settings. Note the points derived from the test strips now define a common curve, i.e., density is translating to Screen Brightness Value using the same rules for all the strips. This common curve is exactly like the one described as “Ideal” in Figure 4. If you open these scans next to each other on a common Photoshop screen, the appearance of each is visually faithful to the original strip it represents.
 |
Figure 7 takes this several steps further. In this figure we prepared and scanned step tablets (essentially test strip exposures with many small increments) on a variety of silver and chromogenic papers representing a range of manufacturers and surfaces. This provided the 150 data points plotted, which confirm that as long as we scan with our optimum 245/14/2.0 settings, any particular print density from any silver original always translates to the same brightness on screen. The translation process has been “rigged” to guarantee that any Brightness Value on screen “looks like”—is visually faithful to—its counterpart density in the print.
Important note: These tactics work only when scanning silver or chromogenic prints. They do not work when scanning inkjet prints. However, if we use values of 245/19/2.0 when scanning inkjet prints, everything again falls into place. We have scanned test strips made on 16 different inkjet papers using a Shadow setting of 19 (rather than 14), and the results match Figure 7 perfectly.
Next steps
You now have a way to make any tone in a print original look correct on-screen. How do you make that on-screen tone look correct when it gets printed? For example, if an original print density of 1.00 always looks best on-screen as a Brightness Value of 98, why not force a Screen Brightness of 98—whenever it is encountered—to print on an inkjet printer as a density of 1.00? Part III will explain how to do this, essentially making the screen-to-printer conversion the mirror image of the scanner-to-screen conversion. It explains how to do this with any printer, ink or paper.
Dick Dickerson and Silvia Zawadzki are retired Kodak black-and-white product builders. Their PT articles include “The Genesis of Xtol” (Sept/Oct ’99), and “Testing Kodak’s (Remanufactured) Black-and-White Films” (March/April ’03). Through their company, Dick & Silvia’s B/W Toolbox, Inc., they offer consulting and teaching services in black-and-white photography.
