Web DevCenter    
 Published on Web DevCenter (http://www.oreillynet.com/javascript/)
 See this if you're having trouble printing code examples


The Ideal Digital Photographer's Workflow, Part 4

Steps for Working Optimally with Scanning

by Ken Milburn, author of Digital Photography: Expert Techniques
05/07/2004

Editor's note: Ken Milburn continues his quest to create ideal (and least destructive) workflows for people working in digital photography. This installment brings this series to a close, but if you missed any of the first three articles in this series, you'll find links and descriptions at the end of this piece. Whether you are new to digital photography or a seasoned pro, the tips and suggestions presented in this series will help improve your productivity in all aspects of your shooting -- from preparation to final product. As Ken suggested in part one, print out these tips and tack them on the wall where you develop your photos; next to your digital camera, they may become the handiest tools you own.

This article details how to get the most out of the time you spend converting your film library to digital. If you’re like most of us, you’ve shot thousands of pictures that pretty much document your life (or career, if you’re a pro). You’ve probably realized that film fades and scratches over time. So you’ve decided that the sooner you digitize those images, the sooner you will have ensured that they will not deteriorate any further.

This article is not about scanning for pre-press, making a quick scan in order to email a film picture to a friend, or scanning for any purpose other than to find the most efficient way to get the most picture information from your slides and negatives when you convert them to digital. For that reason, this article intentionally ignores some scanning features and commands that might prove useful when your application calls for the most immediately satisfying results, rather than for acquiring a digital negative that will give you the most options for the highest-quality interpretation of the digital image.

Note: Even if you do scan every negative and slide you’ve ever taken, don’t even think about throwing the film away. It is still valuable as backup, should a fire or theft obliterate your digital files. Furthermore, the technology of scanning may be twice as good in two years as it is now.

1. Pick the Right Scanner

Be forewarned that scanning takes time. It can take several minutes (four minutes is actually pretty fast, though some Nikon Coolscan models can scan a 4000 DPI image in one minute) to scan a single slide. If you have 1,000 slides, that’s 4,000 minutes, not including the amount of time it takes to mount the film in the scanner and to make adjustments in the scanner software before clicking OK to scan. Add at least another 4,000 minutes for all of that. That’s about three and a half weeks of work, given a 40-hour work week.

You can speed up scanning by scanning at lower resolution than the scanner is capable of, but then you’ll have to re-scan the same slide if you ever need to scan it at higher resolution. So always pick the fastest scanner with the highest resolution (at least 4,000 DPI) that you can afford.

Note: When you’re preparing to scan slides, make sure you skip the slides you know perfectly well you’ll never use. Scanning those slides not only wastes your time, but uses valuable hard drive space and makes it harder to find the slides you really do want to use. Of course, the same is true of your digital camera files.

Second, you want to capture as much image information as you can in your scan. In order to do that, you ideally want a scanner with the following features:

  1. A D-max of at least 3.6 (you’ll find the D-Max in the scanner specs)
  2. Resolution of at least 4,000 DPI for 35mm film
  3. Bit depth of at least 36 (48 would be better)
  4. Very little shadow noise (if all other specs are the same, significantly lower prices can indicate cheaper circuits, which usually indicates more noise)
  5. Software that features a histogram (see "Scanning for Maximum Image Data," below)
  6. Scanning without banding or other mechanical problems (read several hands-on reviews to determine if this is the case)

Third, get a scanner that will do batch scanning of at least six frames at a time. A few scanners will even scan a whole roll of 36-exposure film at a time. If the scanner does it right, you can pre-scan the preview frames all at once to get thumbnails. You can then make individual settings (including “delete” or “don’t scan") for each frame or slide, and then push a button and let the scanner do all the work while you either go out for dinner or use another computer to do something more productive.

2. Organize Your Files into "Job" Categories

Before you start scanning, physically organize your files into “job” categories. Physically separate the film into boxes and envelopes in this order:

Client/Event/Subject, Date, Job/Shoot Description

So professional photos might be organized something like this:

Clorox-042004-bottles

And a private or non-client photo folder label might look like this:

SuzyBday-042004-candles

You are organizing your film in this way because it will be most efficient (for a variety of reasons) to scan job categories in sequence. The two most important reasons why this is so are: 1) settings are most likely to be similar for larger numbers of frames, so once you have the settings for one frame, you can simply apply them to many others in the same shoot; and 2) it is easier to keep track of the organization of your files when you archive them and when you place them on your computer and start processing them, especially since I’m going to show you an editing workflow that demands organization.

3. Calibrate Your Whole Scanning System

There is only one way you can be assured that you’ve captured all of the data your film has to offer. That is to work with a system that uses a calibrated monitor and scanner.

You probably know that it’s important to calibrate your monitor, but you should calibrate your scanner, as well. Calibration is a complex topic, and there are already several good books on the subject. One of the most respected is Real World Color Management, by Bruce Fraser, Fred Bunting, and Chris Murphy.

For purposes of this article, suffice it to say that you should use a colorimeter, such as the ColorVision Spyder (arguably the best compromise between price, accuracy, and ease of use), to calibrate the monitor, scan a Kodak color chart slide, and adjust the settings until what you see in the scan when you open it in Photoshop comes as close as possible to matching the colors you see in the original color chart.

Calibrating your scanner may be a bit tougher, but at the very least, do this: go into the scanner software and set it to scan a RAW file. If your scanner doesn’t have that feature, scan for the type of film that the test chart is shot on, probably Ektachrome 200. (Your scanner manufacturer should supply you with a slide that is a standard test chart; see Figure 1. The test chart will contain color charts of several ranges of primary colors and a gray scale chart.) Preview the scan and use the histogram, color, and curves features to make what you see on your monitor looks, as much as possible, like the original slide. Some scanners won’t let you remove the slide from the holder while the preview is still on screen. If that's the case, you may have to do this several times while you use your memory to compare the screen picture to the image on the slide. When you’ve come as close as possible to seeing the same image values on screen as you see in the slide, save the settings as your scanning defaults.

Screenshot.
Figure 1. A typical color test chart

Once you’re calibrated your scanner, you’re going to scan all of the images at the highest possible resolution and in 16-bit mode. Do not try to make the scan look perfect. The idea here is to capture as much information as possible. Later, there will be any number of ways to use Photoshop in 16-bit mode to make the picture look as satisfactory as possible.

4. Use a Scanner That Has Digital ICE Technology Built In

Related Reading

Digital Photography: Expert Techniques
Professional Tips for Using Photoshop & Related Tools to Enhance Your Digital Photographs
By Ken Milburn

Digital ICE is a Kodak technology that is licensed to only certain scanner manufacturers (Kodak, Nikon, Minolta, BenQ, Umax, Durst, and Pacific Imaging). Digital ICE is currently the best way to automatically remove dust and scratches from the film surface. It does this by reading the image only from the protective coating that covers both sides of film. Since these surfaces aren’t light sensitive, the only image read by Digital ICE is that of dust and scratches. The scanner software then subtracts the dust and scratches image from the photo image and automatically fills in the blanks. Other software and filters for removing dust and scratches may be better than nothing, but will do the job (at least in part) by slightly blurring the image. Of course, if you already own a scanner and are reluctant to spend the money to replace it, you will at least want a feature in the scanner software that removes dust and scratches in the old-fashioned way. Otherwise, you’ll have to retouch them all by hand. Even if you’re using Photoshop’s healing brush to do that (after you’ve done your scanning), you’re going to spend 15 minutes to 30 minutes per image doing it.

By the way, for large image defects, the Photoshop 7+ healing brush is a godsend if you don’t have Digital ICE. You can use your dust-and-scratches routine, either in the scanner software or in your image editor, to remove the small stuff with minimal cost in sharpness. Then all you have left to do it to use the healing brush to remove the few big defects that are left over.

5. Orient Your Film Properly in the Scanner

If you’re loading film strips, make sure the carrier doesn’t cramp or pinch the film, causing it to buckle. Make sure you place the film in the carrier so that the emulsion (duller) side of the film is closest to the scanning lens. Most scanner software (and all image-editing software) lets you easily flip and rotate the image, but doing so will compromise image quality, because the scanner has to “read” through thick layers of gelatin to record the image in the film emulsion. On top of that, the software has to rearrange all of the pixels in the image, which may result in errors. Finally, scanning time is increased because the scanning software has to calculate the “flip and flop” interpolation.

Finally, never crop your frame in the scanner software. If you do, and a given job or exhibit frame requires different cropping, you will have to physically search for the film and re-scan it.

6. Collect As Much Data from the Film As Possible

You don’t want to start by making a different scan for every purpose. Instead, make one scan that captures as wide a dynamic range as possible, preferably at the right color balance.

If your scanner will let you, scan everything in 16-bit RAW format to start. RAW files contain every bit of data that the scanner’s sensors have recorded, without regard to compensation settings (such as white balance, or controls that alter contrast and brightness). You can then save these files to 16-bit TIF format that can be opened and adjusted in Photoshop for exactly the tonal range, color balance, and regional contrast (curves) you want to keep in the final image.

When it comes to using the scanner software, you should scan in the scanner’s highest bit mode (usually 36-bit). Since you’ve already calibrated the scanner, you shouldn’t do much with the scanner software to change the way the image looks. That is, don’t try to correct for color balance, sharpness, saturation, or to do curve adjustments that adjust the contrast of various ranges of color. You will be able to do a much better job of that in Photoshop and you don’t want to risk throwing out any data that might prove useful later.

The one adjustment tool you do want to use is the histogram to set the minimum black point, maximum white point, and (possibly) overall brightness. You should use exactly the same procedure you would use to do this in Photoshop. You’re only doing it in the scanning software in order to scan the maximum dynamic range from the original film image by eliminating the pixels that otherwise would be devoted to completely blocked highlights and shadows. If you’re not familiar with the basic histogram procedure, here’s the drill:

  1. Look for a command, tool, or icon called histogram (also known as levels). If there isn’t one, you can do this in your image editing program. The advantage in doing it in the scanner software is simply that you’ll start your image editing application (Photoshop, etc.) with the best-looking image you can get from the scanner without throwing out any data. Once the histogram palette has opened, it will look like the figure shown immediately below.

    Screenshot.
    Figure 2. A typical histogram palette, showing the sliders for black point (deepest shadow), midtone brightness, and white point (brightest highlight without detail)

  2. If either end of the histogram diagram (the thing that looks like the silhouette of a mountain) touches the baseline before reaching the nearest side of the frame, move the corresponding shadow or highlight slider to that point. If the image still seems too bright or dark overall, drag the midtone slider until you’re reasonably satisfied with the result. However, don’t do this to any extreme degree that might cause clipping. Clipping is what occurs when you’ve over-adjusted the midtone slider to the point where the pixels stack so high in the diagram that they are shown as cut off by the top of the histogram frame. If that has occurred, then you have lost image information. The figure below shows how the histogram dialog will look after the sliders have been properly adjusted.

    Screenshot.
    Figure 3. The histogram palette, showing the proper adjustment for maximum data capture.

Color balance isn’t very important if you’re scanning RAW, because you can correct it visually when you interpret the RAW file later.

It’s usually a good idea to do mild Unsharp Mask sharpening -- just enough to make the image look reasonably professional in respect to sharpness, but leaving room for further sharpening later before any halos or artifacts are created. I will soon do an entire article on sharpening that tells you which sharpening tools you should use at each stage of image editing, and to what degree. Meanwhile, you’ll find quite a bit of information in Digital Photography: Expert Techniques that will at least get you started in the right direction.

Be sure you embed the Adobe RGB color profile in every scanned image. Actually, this is to give you a reasonable amount of interpretative flexibility while conserving your time. Otherwise, you should just pick the widest possible gamma for your color profile and then convert that to whatever profile is most suited to the application used for any specific interpretation of the scan.

7. Open the Files in the Photoshop CS File Browser

If you followed steps 1-6, you have easily, quickly, and successfully converted your film to digital files that contain enough resolution and information to make prints of virtually any size. The first thing you should do is to use the File Browser’s Batch Rename command to rename your files according to the guidelines found in Digital Photography: Expert Techniques and in the second of my workflow articles on the O’Reilly site, "The Ideal Digital Photographer’s Workflow, Part 2."

Once you’ve renamed the files, use the File Browser’s Flag icon to mark all of the files you think you’re likely to want to present immediately. Once you’ve done that, use the Show Flagged command to show only the files you’ve flagged, and then duplicate and rename each of those file by adding “edit” to the end of each file name (see Figure 4, below).

Click for full-size view Figure 4. The Photoshop CS File Browser, showing the flagged and renamed scanned files. (Click for full-size view)

Next, apply the steps recommended in "The Ideal Digital Photographer’s Workflow, Part 2" to all of the images you’ve just duplicated. Note that you haven’t lost any of the data you were so careful to preserve in scanning the original files.

Summary

You have just learned that the most efficient way to scan your images is also the best way to ensure that you’ve preserved as much of the data that was in the original image as possible. The big surprise: as complex as some of today’s scanning software is, the secret to successful scanning is applying the KISS (Keep It Simple, Stupid) principle. That is, aside from the histogram (levels) command, sharpening, and the procedures for removing dust and scratches; you’re much better off (and the procedure is much faster) if you do your more refined image editing in a professional image editor such as Photoshop CS.

Finally, many of you are going to ask, "What are we supposed to do when friends, clients, or buyers want us to scan our film so that they can see our pictures digitally on a contact sheet, in an email, or on a web page?" The answer is simple: scan your image as I’ve recommended and then do the best job you know how to do using Photoshop to make the image as visually exciting as possible. Then use Photoshop or Album -- not your scanning software -- to batch convert and export your files to a directory intended for making your web pages or slide shows. Then use Photoshop or Photoshop Album to automatically make web pages or slide shows from the images in that directory.


In March 2004, O'Reilly & Associates released the Digital Photography: Expert Techniques.

Copyright © 2009 O'Reilly Media, Inc.