Using Kodak PhotoCD Technology for Preservation and Access:
A Guide for Librarians, Archivists, and Curators
Anne R. Kenney and Oya Y Rieger
Department of Preservation and Conservation, Cornell University Library for
New York State Education Department, Program for the Conservation and Preservation of Library Material
Don Brown's Recommendations for Kodak Photo CD Scanning Process
First, there are some very basic scanning practices all Photo CD service providers should follow. I would like to think these do not need to be enumerated by customers negotiating with a Photo CD scanning service, however I will go ahead and list them:
I also feel the need to add a few words about different Photo CD scanners. The system supports two scanners: the PCD Film Scanner 2000, which is a 35mm-only scanner, and a multi-format professional scanner, the Professional PCD Film Scanner 4050. The predecessor to the 4050 scanner is the 4045 scanner, and it is still used by some labs. The 4050 scanner has improved internal electronics resulting in better dynamic range. An upgraded 4045 scanner is called a 4045 Enhanced scanner and is equivalent to the 4050. For scans of negatives, the improved dynamic range of the 4050 is not needed, so there is very little difference between negative film scans on the 2000 and 4050 scanners. The primary benefit of the 4050 is in scanning chromes. One additional benefit of the professional scanners is that higher resolution scanning is possible. These scanners can produce Image Pacs with a maximum resolution of 64BASE (6K x 4K). For a scan of 35 mm images, a 64BASE Image Pac really does not provide any more image detail than a normal 16BASE (3K x 2K) Image Pac. The benefit of higher resolution is appreciated more in scans of mid and large format films. So, the conclusion is that very good results can be achieved with scans of 35 mm negative film on the 2000 scanner. As a happy coincidence, this is also the most cost effective option due to the productivity of the 2000 scanner.
Now, when we did our scanning, we tried to assure the Kodak Gray Scale was captured correctly. This is based on the belief I've mentioned before: if you get the neutrals correct, most of your work is done. So, a technique was derived to do a first-approximation calibration to the Kodak Gray Scale to improve the quality and consistency of the scans. The technique assumes the scale is perfectly neutral, the copystand lighting is perfectly uniform, and the frame to frame exposure variability of the film is negligible. All of these are reasonable assumptions if care was taken during the photography. Scan to scan variability of the scanner is known to be very small.
Individual scans can be custom-adjusted with this same technique, but the desire is to create a standard correction for all images at the beginning of the scanning session and automatically apply it to all subsequent images. The PIW software accommodates such a workflow, and after the initial setup, the scanning process can be very efficient.
The technique consists of three basic steps:
1. Create Aim Photo YCC Values
The assumption was made that the Kodak Gray Scale is perfectly neutral throughout the range, not a bad supposition, and that its location in an image is representative of the lighting in the rest of the image. Given that, aim Photo YCC values can be calculated.
Photo YCC is a luminance-chrominance color space, and therefore a perfect neutral will have no chrominance, or color, component. Photo YCC encodes this neutral point with chrominance values of C1=156 and C2=137. The luminance, or Y channel, records the variation from black to white, low numbers dark and high numbers light.
Given the assumptions above, the C1 and C2 values are set to 156 and 137 respectively for all the gray scale steps. The Y values can be calculated from the reflection density of the patches of the actual Kodak Gray Scale. Ideally, the same scale used in the photography should be measured. The measurement is made on a reflection densitometer, reading visual density. (In our case, we measured one of our Kodak Gray Scales and made a leap of faith that it was identical to the one on the image.)
The conversion from reflection density to Y luminance value is:
Y = 181.9(1.099*10-0.45D - 0.099)
where D = reflection density
It should be noted that this conversion is only valid for perfect neutrals, and not for colors.
The Kodak Gray Scale has nominal density values for each of the patches, ranging from 0.0 to 1.9 in 0.1 increments. The literature included with the scale states, "Density increments may vary slightly above and below the stated 0.10 value. Neutrality and uniformity of the patches are carefully controlled". With this in mind, it is suggested the specific scale used in the photography be measured, and aims derived from those measurements. In the interest of providing an example, and if measurement is not possible, the Photo YCC aim values for the nominal Kodak Gray Scale densities are given below.
A 0.0 182 156 137 1 0.1 162 156 137 2 0.2 144 156 137 3 0.3 128 156 137 4 0.4 114 156 137 5 0.5 101 156 137 6 0.6 89 156 137 M 0.7 79 156 137 8 0.8 69 156 137 9 0.9 61 156 137 10 1.0 53 156 137 11 1.1 46 156 137 12 1.2 40 156 137 13 1.3 34 156 137 14 1.4 29 156 137 15 1.5 24 156 137 B 1.6 20 156 137 17 1.7 16 156 137 18 1.8 13 156 137 19 1.9 10 156 137
2. Disable the Auto-balance
Color negative film scanned on PIW systems is automatically analyzed for color and density. The analysis algorithm otherwise known as the Scene Balance Algorithm or SBA, produces color balance values which normally are applied to the scanned image. The operator can then make additional manual adjustments based on the preview image on the monitor.
Assuming each of the images were captured with identical lighting and exposure, a single set of manual adjustments will correct for each frame in the series. The auto-balance algorithm will only confound this technique and therefore needs to be disabled.
The method to disable the algorithm is somewhat arcane. It requires creating a new film term based on the normal film term for the particular film type being scanned. Service providers are familiar with creating custom film terms, and the process is well documented in the PIW manuals. The esoteric part is the value which must be changed within the film term to disable the auto-balance. Without going into the details, the new film term must set CMD=3. This step is critical, and not difficult to perform. It may, however, be an unusual request.
The default film term for color transparencies has the auto-balance disabled, so this step in not necessary for scans of chromes.
3. Determine Color and Tone Adjustments
The image adjustments are derived during the preview part of the scanning operation. A preview scan is made of the first image in the series using the newly created film term. Manual color and tone adjustments are made on this image to drive the sampled Photo YCC values of the patches of the Kodak Gray Scale to the calculated aim values.
Two sets of adjustment tools are available to the operator: contrast and balance image adjustments, and the Tone Scale Adjustment tool. Both are manipulating the same parameters, and both can achieve the same results. For this type of adjustment, however, the Tone Scale Adjustment tool is easier to use.
This is an iterative process, but the following guidelines should help. The Tone Scale Adjustment tool allows the operator to pick a highlight and a shadow point on the image, and set RGB aim values for them. The highlight point should be on patch A or 1 on the Kodak Gray Scale, and the shadow point on 15 or B. The RGB aims should be adjusted, with the constraint that R = G = B to preserve neutrality, until the Photo YCC aims for that patch are achieved. After the highlight and shadow patch aims are attained, the Photo YCC values of the rest of the steps in the scale should be sampled to make sure they are neutral and near the correct aim values. Selecting different highlight and shadow patches may improve the overall tone scale neutrality. When the best compromise is determined, the adjustments are held for the remaining scans.
This process is essentially a straight line adjustment to what may be a nonlinear variation. In other words, there may be some color and density variation within the tone scale that cannot be fully corrected. If properly done, however, this technique can correct for the majority of color and tone problems and give very good results. The success of this process is somewhat dependent on the skill of the operator, but the aim values make this task much more manageable.
Tone scale patches on a few subsequent images should be examined to assure the image on which the corrections were based is representative of the entire series. Once this confidence is achieved, the rest of the scanning can be done with no operator adjustments. If anything changes within the a series of images, such as photographic lighting difference, film emulsion change, even film processed at a different time, it is a good idea to repeat the correction for each sub-series within the entire series of images.