· The master images in a historical archive need to be adapted for particular types of displays. The most memorable displays place photographs in context of a story.
· The basic capabilities needed to modify historical images for display are available in both consumer level software such as Photoshop Elements and professional software such as Photoshop CS.
· Restoring tone and color is relatively safe, but any efforts to correct dust, scratches, and other defects must be done carefully and can compromise the historical integrity of the image.
· The steps that may be needed to prepare an image for display are listed in order below, but not all steps are used with all images.
o Make a copy of the master file
o Implement grain and noise reduction
o Convert to grayscale
o Make tone and color adjustments
o Correct dust, scratches, and other defects
o Set output color space
o Change image size and resolution
o Sharpen the image
o Set output format
o Set output file name and location
· Adjustments to images can be done more quickly and reliably using layers.
The ultimate goal in developing a historical archive is to share historical information. This chapter describes some strategies for displaying historical photographs and documents, and the steps needed to prepare the master images for different types of display.
Historical photographs and documents are most meaningful and memorable when they are presented as part of a story. Isolated photographs can be interesting, but a story is more interesting. The documentation for each photograph as described in Chapter 4 is the basic means to archive the story along with the photograph. The optimal display of a photograph will include information about the context or story of the photograph. The story could be about a person’s life, a couple, a period in a person’s life, or a particular event, building, or piece of land. Challenges and difficulties are part of a good story, including when the difficulty does not have an ideal outcome.
Photographs and documents can be displayed as prints on paper, as electronic still images on computers, and as slideshows on computers or TVs. Slideshows can provide the most meaningful and effective method for telling a historical story and are described in Chapter 8. The present chapter focuses on preparing and displaying prints on paper and images on computers. First, basic strategies for presentation are summarized. Then the steps for preparing the master images for presentation are described.
Restoration of a historical photograph or document must be done very carefully. It is easy to degrade the historical integrity of an item. In my early days working with historical photographs I once “restored” a photograph with a significant tear. I was pleased that the tear was not obvious in my “restoration” and rather proudly showed it to a friend. She knew more about the Amish-Mennonite community where the photograph had been taken and quickly commented “I doubt that the woman’s dress was actually that short.” That was an area that I had “restored.” Later we found the original negative, which verified that her comment was correct. Even after I became much more cautious, subtle but significant historical distortions sometimes occurred. I once unintentionally removed a mole from a man’s face while removing dust spots. In fact, the man had the mole removed sometime after the photograph was taken. As noted by his daughter, my alteration of the photograph misrepresented the actual timing of events.
This emphasis on historical integrity is a very different approach from the artistic improvements that are often considered standard procedure when working with Photoshop. I have regretted doing too much to restore historical photographs more often than I have regretted doing too little.
Careful efforts to restore tone and color in an image and to sharpen the image usually bring out detail and rarely compromise historical integrity. However, efforts to correct defects such as dust, scratches, fingerprints, stains, and tears can easily compromise the historical integrity of a photograph.
People expect historical photographs to have defects and tend to not notice the defects. This is particularly true when a photograph is presented in context of a story. Minimal restoration is not only acceptable; it is the safest practice.
At the same time, some defects are distracting and obviously were not part of the original photograph. When these can be easily corrected without threatening the historical integrity of the photograph, it may be appropriate to make the corrections. Defects tend to be more noticeable in photographs that are presented alone rather than as part of a story. When there is uncertainty about whether a spot is actually a defect or about whether a defect can be corrected without compromising the photograph, the best practice is to leave it alone.
The automatic enhancement functions in image-editing software such as Photoshop may produce results that are adequate for casual display of photographs. However, the steps described below generally produce better results for more formal displays. For example, the Shadows/Highlights filter produces effects that are not matched by the automatic enhancement functions. Also, in my experience the automatic enhancements rarely work adequately with documents.
Prints on paper are the preferred method for viewing historical items for many people, particularly older people. Public display of photographs has historically been based on prints. Gifts of individual photographs or collections generally have been paper prints. In addition, prints can also be an effective means to organize and review the current status of a developing historical archive.
A simple three ring notebook with sheet protectors can present historical information in a way that is flexible, easily developed, easily updated, inexpensive, and relatively rugged. Notebooks are an effective way for people to review the current status of an archive and can easily be revised when images are added or documentation is updated. I often print photographs on 8.5 by 11 inch photo paper and put them in a notebook with the documentation on the adjoining page. The images could be printed smaller than 8.5 by 11 inches to reduce the number of pages, but the large photographs and easily read documentation are very popular, particularly with older people. A notebook targeted to a particular topic can be quickly assembled for a special occasion, such as a reunion or funeral. Last minute revisions, or even revisions during the event, are no problem.
Traditional photograph albums and scrapbooks may be used when revisions will be rare. These methods typically are not designed for frequent updates and provide less information about each photograph or document than is recommended here. However, an album or scrapbook could be adapted to provide more information about the items. The more complete information is particularly appropriate for albums and scrapbooks that are intended to be passed down to future generations rather than serving to hold memories for one person.
Mounted and framed photographs and documents are the most formal type of display. Here too, if the mounted items are intended to be passed to future generations, information about the items should be maintained with the items. If display of written information is not wanted, the documentation could be attached to the back of the frame or mount.
Archival materials for a paper print are best if the print is planned to be used for many years. The paper and ink for the print should be archival quality. For example, some photo papers and ink made for Canon and Epson printers are intended for long-term use, while other are not. In addition, the materials that hold or mount the print should be acid-free archival quality, and preferably meet the PAT criteria described in Chapter 1.
An alternative strategy is to not use archival materials and simply reprint items that deteriorate. One of the major advantages of electronic archives is that duplicate prints can be easily made. In some situations, reprinting the item will be easier and less expensive than using archival materials.
Commercial printing of digital images typically uses sRGB color space and will produce poor results for images with the Adobe RGB color space. Color space conversion will be needed if the master images have Adobe RGB color space. Printing with a dot matrix photo printer can produce good results with Adobe RGB color space; however color management as described in Appendix A will normally be needed for consistently good prints. Also, optimal prints of historical photographs and documents normally require significant sharpening.
The internet is the easiest way to widely share historical images. The innumerable options for displaying images on the internet include social networking websites, websites dedicated to images, personal websites, blogs, and the many websites that catalog and display archives of historical images. Internet search engines can be used to find images with certain content.
Images optimized for the internet need small file sizes. Large files are slow to open and adversely affect the use of a website. Many websites resize and reformat images automatically. Images for the internet often have 500 to 1200 pixels on the longest side and are in the JPEG format with medium quality compression.
One increasingly common strategy is to have two copies of an image on a website. One copy is a small file optimized for internet browsing. The other copy is a larger file that is downloaded only if a higher resolution image is specifically selected. Some websites display a low-resolution image and require the higher resolution image to be purchased. This strategy is used on many historical archive websites.
Images for the internet have the sRGB color space. At present most internet browsers assume images use the sRGB color space. Other color spaces are not properly handled. Improved implementation of color management can be expected in the future; however, at present sRGB is needed for internet use. Also, optimal display of historical photographs and documents on the internet usually benefits from some sharpening (but much less than for prints on paper).
Archival master images usually need some adaptation for optimal display. The amount of adaptation depends on the characteristics of the master image, the type of display, and the degree of restoration that is considered appropriate for the project. Steps that may be needed are listed below. Most images will not need all of these steps. The steps that may be needed for any type of display are:
· Make a working copy of the master file
· Implement grain and noise reduction
· Convert to grayscale
· Make tone and color adjustments
· Correct dust, scratches, and other defects
After these adjustments have been made, the image could be saved to a working or intermediate file. This file would be the starting point for the derivation of files for different types of display. The steps that depend on the type of display are:
· Set output color space
· Change image size and resolution
· Sharpen the image
· Set output format
· Set output file name and location
Each of these steps is described below.
This chapter describes the use of Photoshop Elements and Photoshop CS for preparing images for presentation. The discussions of Photoshop Elements and Photoshop CS in Chapter 3 are prerequisites for the current chapter. In particular, the use of layers when modifying images is extremely useful. In addition to basic layers, layer masks are a more advanced method that can be useful when preparing images for presentation.
A layer mask can be used to make the effects of the layer apply to different degrees in different parts of the image. The presence of a mask is indicated in the Layers panel as a white box on the row for the layer. The mask covers the full image. Where the mask is white, the effects of the layer are fully implemented. Where the mask is black, the effects of the layer are suppressed and the lower layers apply. Where the mask is gray, the effects of the layer are partially implemented. Light gray has greater implementation than dark gray. The entire mask is white by default so the layer is fully implemented. If a layer does not have a mask, a mask can be added by clicking the icon for “Add Vector Mask” at the bottom of the Layers panel, or by using the dropdown menu Layer> Layer Mask> Reveal All. Masks are handled the same in Photoshop Elements and Photoshop CS.
The Brush Tool can be used to paint the mask to control how much the layer effects apply in different parts of the image. Set the brush foreground or top color to black by tapping the D key or clicking the small icon on the left toolbar for “Default Foreground and Background Color.” If this sets the default foreground color to white, change it to black by tapping the X key or by clicking the small icon for “Switch Foreground and Background Colors”. Then activate the Brush Tool by tapping the B key or clicking on the icon on the left toolbar. If a gray color is wanted, set the opacity of the brush to less than 100% on the options bar at the top of the screen. Set the brush size and hardness (sharp or soft edges) by clicking the icon for the “Brush Preset Picker” on the brush options bar at the top of the screen. The size can also be increased or decreased by tapping the ] or [ keys. Now simply paint on the image. The black or gray color being painted does not directly show on the image. Rather the modified effects are displayed on the image as the image is painted. To see the actual mask, click the Channels tab on the Layers panel on the right of the screen. Then click the box for the eye icon on the row for the layer mask. The black or gray painting on the mask shows in red.
The safest procedure is to copy a master image to a working folder before making modifications. This greatly reduces the possibility that the master image will be unintentionally modified. A file naming convention is a good practice for the different stages of developing display images. For example, after the basic steps that will be needed for any type of display have been implemented, the image could be saved to the working folder. The file could have the same name as the master image with a code such as _a (for adjusted) added to the file name. When this file is modified for printing, the _a could be replaced by _p. The file modified for display on a computer or the internet could have _c used in the file name.
Grain and noise can significantly detract from a photograph. A scanner that produces sharp images can bring out grain in an image. Conspicuous grain can occur with any type of original, but tends to be particularly distracting in images from black and white film. Digital cameras produce noise that manifests as a grainy texture and random pixels with incorrect color. Digital camera noise tends to be most noticeable in darker areas of an image and in images taken with higher ASA settings.
In recent years the software for reducing grain and noise has tended to focus on noise from digital cameras more than grain from scanning. Grain and noise are reduced by blurring the image. Significant effort has been spent attempting to develop algorithms that reduce the grain and noise while minimizing blurring of the edges and detail in an image.
The most common recommendation by experts is to do grain and noise reduction early in the processing. However, many people report that it makes little difference whether it is done as an early step or near the end of processing. Grain reduction should be done before the step of final sharpening. Noise reduction and sharpening software are increasingly integrated and can be done together as a later step in processing.
The Reduce Noise filter in Photoshop Elements and Photoshop CS is widely described as being less effective at reducing grain and noise than third-party noise reduction software. That has been my experience also. This filter also focuses on digital camera noise and is less effective with grain in scanned images. However, the Reduce Noise filter may have some benefit if it is the only available option. After creating a new merged layer use the dropdown menu Filter> Noise> Reduce Noise. The controls in the dialog box are straightforward.
Adobe Camera Raw in Photoshop has noise reduction options that are sometimes described as competitive with third-party software. However, my experience has been that third-party noise reduction software that generates a noise profile for each image is significantly more effective and easier to use than Camera Raw noise reduction. In addition, Camera Raw is primarily designed for the early stages of processing raw images from digital cameras. It can be used with TIFF images, but the workflow of switching between Photoshop and Camera Raw is cumbersome and does not incorporate layers. Camera Raw is designed for an earlier stage in image processing that is not based on workflow built around layers. Other options are preferable for the later stages of preparing a scanned image for display.
Three of the most popular third-party noise reduction programs are Neat Image, Noise Ninja, and Noiseware. All three are widely described as producing good results. Neat Image was originally developed years ago when grain in scanned images was the primary problem and that legacy has been maintained as the program has been updated to handle noise from digital cameras. Neat Image works well with historical images. My experience with grain and noise reduction has primarily been with Neat Image.
Neat Image can be used as a plug-in for Photoshop Elements or Photoshop CS. After installation, Neat Image appears as an option on the Filter menu. It can be used with layers, masks, and selections. With Photoshop CS, Neat Image can also be used as a smart filter and can be included in Actions to automate processing (as described in Appendix E on batch processing). Neat Image is also available as a stand-alone program.
To use Neat Image, create a new merged layer with Ctrl-J if this is the first layer or Alt-Ctrl-Shift-E if there are other layers. In Photoshop CS, Neat Image can be run as a smart filter by tapping Ctrl-Shift-S or by selecting “Convert to Smart Object” in the menu in the upper right corner of the Layers panel. Initiate Neat Image with the dropdown menu Filters> Neat Image.
Creating the Noise Profile
The first screen that displays in Neat Image is for selecting an area of the image to develop a profile of the noise. Neat Image displays a blue box over the area that it believes is appropriate. The area should be uniform with no items, people, or detail. On the rare occasion when a better area can be selected, the box can be dragged to a new location and re-sized if needed. Click the button for “Auto Profile” to generate the noise profile. The “Quality:” field on the right displays a measure of the usefulness of the profile. If the value is above 75% the profile is considered accurate, but lower values usually work well.
Applying the Noise Profile
After the noise profile has been generated, click the tab for “Noise Filter Settings” to apply the profile. If the 15 sliders for the advanced options are not displayed, set Neat Image to advanced mode with the dropdown menu Tools> Advanced Mode. The Neat Image screen displays the image with the noise reduction implemented in accordance with the current settings. If the mouse pointer is placed over the image and the left mouse button held down, the display switches to the original or input image. This allows easy evaluation of the effects of the settings. Also, the mouse scroll wheel can be used to zoom in and out of the image and holding the spacebar down converts the mouse pointer to a hand tool that can be used to drag the zoomed image.
Usually only a few of the sliders will be adjusted. Unfortunately the workflow for the sliders is not from top to bottom. The safest strategy is to click on the pick-list near the top of the right column and select the Default option. This clears the settings from the previous use of Neat Image. In most cases the only adjustments that are needed is to set the High, Mid, Low, and Y sliders in the Noise Reduction Amounts section in the middle of the column. The default settings of 60% for the Y slider and 100% for the other 5 sliders in this section usually produce too much noise reduction for historical photographs and are more appropriate for digital camera noise.
The amount of noise reduction can be reduced by reducing the Y slider—which adjusts the amount of luminance or tone correction— and/or reducing the High, Mid, and Low sliders. As indicated by the icons, the High slider applies to fine detail and the Mid and Low sliders are for larger content in the image. Values closer to 50% for all 3 sliders are optimal in many cases, but the entire range can be used with different images. The High and Mid sliders are often more useful than the Low slider. If more noise reduction is needed, the Y slider can be increased. The Cr and Cb sliders adjust the amount of color correction and have less noticeable effects. The checkboxes for “High resolution” and “High quality” should normally be checked. The checkboxes for “Very low freq” and “Smooth edges” can be useful in certain situations, but usually are not needed and can be counterproductive.
My usual strategy for historical images is to start with the High, Mid, and Low sliders all set to 60% and the Y slider set to 0%. I adjust the Y slider upward until the desired noise reduction is achieved. For cases with severe noise, the High, Mid, and Low sliders are increased to 75% or higher. Neat Image also has options to sharpen the image based on the noise profile. I have found sharpening with Neat Image to produce less grain and noise enhancement than other sharpening methods and increasingly apply sharpening and noise reduction together as a later step in processing. The options and sliders in Neat Image for sharpening an image are described in the later section on sharpening images.
The 6 sliders in the Noise Levels section at the top of the column are used only on the rare occasion that the noise profile is not accurate. However, generating a new noise profile may be a better strategy than trying to correct the profile with these sliders.
Excessive noise reduction results in an artificially smooth image that looks plastic. When evaluating noise reduction, it is useful to look at the full image as well as zooming in to key areas. Areas with texture such as a gravel road require particular care.
Other Grain Reduction Methods
Layer masks and selections can be used to apply different amounts of noise reduction in different parts of an image. Also, the Blur Tool on the left toolbar works well for reducing grain in limited areas such as a person’s skin or an area of sky. Setting the Strength of the tool to a value in the range of 15% to 50% can give good control.
As noted in Chapter 2, I have found the Digital GEM noise reduction that is part of the Nikon scanning software to be the optimal noise reduction method with Nikon scanners. My experience has been that the results with Digital GEM appear more natural and are less prone to looking plastic. Also, the Digital GEM noise reduction is less susceptible to being undone by later sharpening.
Master images in color may sometimes be converted to grayscale for display. In many cases the images will be displayed in color and conversion will not be needed. However, a conversion may be needed in situations such as correcting stains on grayscale images or displaying an image with other images that are all grayscale. One color image may not be appropriate in a group of black and white images.
If spots or stains are being corrected, check whether any stains or spots with color need to be corrected before conversion to grayscale. Some stains or spots may be automatically corrected during the conversion. Any stains or spots that are not corrected during the conversion and that are on significant content of the image should be identified and handled prior to conversion while the color information is available. For example, a stain in an area with writing in a document should be corrected prior to conversion. However, a stain or spot that is in a background area of a document can easily be corrected after conversion. Methods to adjust colored stains or spots are described in the later section on correcting dust, scratches and other defects.
The traditional method to convert color to grayscale when discoloration is present is to look at the amount of discoloration in the red, green, and blue color channels separately. The conversion from color to grayscale can be done using one of the three color channels or a combination of the channels. The channel with the least discoloration is used for the conversion. Other channels may also be mixed into the conversion to optimize the results.
In Photoshop Elements the conversion to grayscale is done with the Convert to Black and White adjustment. This does not have a built-in layer, so begin by making a new layer with Ctrl-J if this is the first layer above the background image, or Alt-Ctrl-Shift-E to create a new merged layer if there are already other layers. The dialog box for conversion to grayscale can be opened with Alt-Ctl-B or with the dropdown menu Enhance> Convert to Black and White. The dialog box has the sliders for each color channel red, green, and blue. Photoshop Elements provides several styles by default for converting different types of photographs to grayscale; however, these probably will not be optimal for discolored historical images.
To view each color channel separately in Photoshop Elements, set the slider to 100 for one channel and to zero for the other two. Unfortunately, precisely setting the sliders is an inefficient, trial and error process. The range for each slider is -200 to +200. Photoshop Elements does not display the numeric values for the sliders unless the mouse pointer is held over a slider for a second or two. Also, the value that is displayed may not be correct unless one of the sliders has been clicked with the mouse.
Fortunately, preset styles can be made with one color set to 100 and the other two colors set to zero. Examining the conversion for each color separately is easy using the custom styles. Vague information on adding custom styles can be found in the Photoshop Elements help information under “add custom presets for black and white conversion.” More detailed instructions are provided as one of the download options at http://archivehistory.jeksite.com/download/download.htm
After the channel with the least discoloration has been identified (often red) set it to 100. Set the other two channels to zero. It may be useful to move the slider for the channel with the most discoloration (often blue) in the negative direction to see if the discoloration reaches a minimum. The other sliders may need to be adjusted to obtain the best tone with minimum discoloration. For photographs the numeric values of the different color channels should add to about 100. For documents with a white or light background the numbers will often add to more than 100 for optimal results.
After the colors have been set to grayscale, the image will still be stored as if it had color until the mode is converted to grayscale. First, all layers should be flattened using the drop down menu Layer> Flatten Image or Ctl-Shift-E. Then set the mode to grayscale using the drop down menu Image> Mode> Grayscale.
In Photoshop CS, the Black & White adjustment is the newest and most flexible method for converting color images to grayscale. This adjustment method has options for built-in layers and has six sliders that control how different colors are converted to grayscale. The adjustment is initiated by clicking the “Black & White” icon on the Adjustments Panel on the right, or using the dropdown menu Layer> New Adjustment Layer> Black & White. For typical historical documents or photographs that have become yellow or brown, only the sliders for yellow and red have any effect. In some cases, the sliders can quickly minimize the effects of discoloration. The red slider may be particularly useful and may be optimal with a high setting.
In Photoshop CS the more traditional selection and mixing of the red, green, and blue color channels can be done with the Channel Mixer adjustment. This may be useful if the Black & White adjustment is not working well with an image. The Channel Mixer adjustment is available with a built-in layer. Click the Channel Mixer icon on the Adjustments Panel on the right of the screen, or use the dropdown menu Layer> New Adjustment Layer> Channel Mixer. The Channel Mixer has sliders for the red, green, and blue color channels. Click the checkbox for Monochrome to set the mixer for conversion to grayscale. The quick way to look at each color channel separately in Photoshop CS is to use Ctrl-3, Ctrl-4, and Ctrl-5 which show the red, green, and blue channels respectively. Alternatively, the Channel Mixer adjustment can be used to examine individual channels by setting one channel to 100 and the other two channels to zero. These settings can be saved as presets and reused. After the channels with the minimum and maximum discoloration have been identified, the same basic strategy as described above for Photoshop Elements can be applied.
After the colors in the image have been set to grayscale, the image needs to be flattened and converted to grayscale mode as described for Photoshop Elements above.
The basic methods for making color and tone adjustments using Photoshop Elements and Photoshop CS were describe in Chapter 3. Those same methods are applicable for refinements when preparing an image for presentation. More sophisticated methods can also be applied at this stage.
More Refined Color Corrections
If color problems occur only in certain areas of the image, a mask can be used to apply color adjustments to just the needed area. Selection tools such as Magic Wand can be used to select areas with certain tonal ranges or colors. These selection tools are described in the section below on correcting defects. If a selection is in effect when a new adjustment layer is made, the selected area automatically becomes a mask for the new layer.
Color casts can be evaluated by using the Info panel on the right. Click the Info tab. If the Info tab is not displayed, select the dropdown menu item Window> Info. The Info tab shows that actual numeric values for the red, green, and blue color channels at the location of the mouse pointer on the image. For an image with a bit depth of 8 bits, pure white has 255 for each of the three color channels and pure black has 0 for each channel. Gray colors have equal values for all three color channels. As the mouse pointer is moved over an area of the image that should be white or light gray, a color cast manifests as one or more color channels with numeric values that are different than the other channels. This can be used to identify which colors need to be changed.
Using Layers in Photoshop Elements
After tone and color have been adjusted, the further adjustments described in sections below normally can be done with 8 bits per channel without noticeable differences compared to 16 bits per channel. This point is important because Photoshop Elements can use layers only when an image has 8 bits per channel. My perspective is that the benefits of using layers are greater than the benefits of using 16 bits per channel for the adjustments described in sections below. Therefore, when using Photoshop Elements, converting the image to 8 bits per channel for use with layers is the preferred practice after tone and color have been set. However, the methods described as filters in the following sections could be applied with 16 bits per channel if applied directly to the image rather than to a layer. The methods described as tools work only with 8 bits per channel.
The methods in this section actually change the content of an image, not just alter the tone and color. They must be used very carefully with historical items to assure that significant content is not unintentionally altered. These methods should never be applied to a master image. They are most appropriate for unambiguous defects or in situations that give higher priority to aesthetic qualities than to historical integrity.
The methods described here apply to both Photoshop Elements and Photoshop CS. A few minor differences and enhancements with Photoshop CS are noted. Unless differences are noted, the descriptions apply for both.
Dust and Scratch Filter
The Dust & Scratch filter is useful when an image has many small dust spots or scratches to be removed. When used, it is normally the first step in correcting defects. If not used carefully, it can significantly soften and reduce detail in an image. The settings described here minimize these adverse effects.
Begin by creating a new merged layer with Ctrl-J if this is the first layer or Alt-Ctrl-Shift-E if there are other layers. If the dust and scratches on the image are white, set the layer blend mode to Darken. If the dust and scratches are black, set the layer mode to Lighten. These blend modes are important in minimizing adverse impacts on images. In the less frequent case that both white and black dust and scratches are present, leave the layer blend mode as Normal.
Use the dropdown menu Filters> Noise> Dust & Scratches to open the dialog box for the Dust & Scratches filter. The Radius setting determines the size of a spot that is removed. A value of 2 to 5 pixels would be typical for the initial filter. Small values are less likely to have adverse effects on the image. The effects of the setting can easily be seen because spots are removed from the image as the radius is adjusted. The Threshold setting determines how much a spot must stand out from the adjacent area before it will be removed. Use the highest value that eliminates the spots. For conspicuous black or white spots, values of 25 or higher work. For spots that are less conspicuous, values of 15 or 12 may be needed. High values are less likely to have adverse effects. Values below 10 can result in significant degradation of the image. The adverse effects are reduced if the layer blend mode is set to Darken or Lighten.
A layer mask can be used to limit the filter to certain areas of the image and to protect other areas. Also, a second filter can be applied with higher settings and a mask used to apply the settings only to selected areas. A mask can be added by clicking the Add Layer Mask icon at the bottom of the Layers panel. Make the mask black with Ctrl-I. This turns off the effect of the layer. Then set the foreground color to white (using relevant icons on the left toolbar) and use the Brush Tool to paint in the areas where the filter should apply. The spots will disappear as an area is painted white.
In Photoshop CS, the Dust & Scratch filter can be set as a smart filter or smart object. After the layer is created for the filter, tap Ctrl-Shift-S or use the list menu in the upper right corner of the Layers panel to select “Convert to Smart Object.” The dust and scratch settings can be revised if this is a smart filter. Also a layer mask it automatically created.
Whether or not the Dust & Scratch filter is applied, some defects typically need to be corrected manually. Three tools that use brushes to make corrections are useful. For all three tools, the size of the brush can be decreased by tapping the [ key and increased by tapping ]. The hardness of the brush determines whether the edges of the effect are sharp or soft and can also be adjusted. However, for correction of overt defects, hard edges are usually best. Softer edges may be appropriate when the tools are used for more subtle enhancements of images rather than for correction of defects. The properties of the brush can be set with the tool options bar at the top of the display.
These tools can be used with a blank layer that is created with Ctrl-Shift-N or with the dropdown menu Layer> New> Layer. A blank layer holds only the changes made with the tools and displays these changes on the image from the layer below. Changes to the layers below are propagated up through a blank layer. For all three tools the checkbox for Sample All Layers in the tool options bar at the top of the display must be checked for the blank layer to work properly. With all of these tools, a selection can be made to limit the area that is affected by the tool. The selection is equivalent to applying a layer mask and is particularly useful when edges in the image need to be protected.
When an attempted correction does not work as desired with these tools, it is better to undo the correction than to try to brush over the resulting problem. Compounding corrections can soften and degrade the final image. In Photoshop Elements Ctrl-Z will undo the last command and can be repeatedly applied to step backwards. However, in Photoshop CS, repeated use of Ctrl-Z toggles the last command on and off. Alt-Ctrl-Z is used to step backwards to undo more than just the last command.
Spot Healing Brush
The Spot Healing Brush Tool is the easiest method to manually remove individual spots or lines on an image. When this brush is clicked or dragged over a spot or line, it modifies the spot or line to be consistent with the surrounding area. On the options bar for the tool, the Type should be set to Proximity or to Content Aware. Proximity is the original type for this tool and often creates artifacts when a spot is near or crosses an edge in the image. Content Aware reduces this problem but artifacts can still occur. Different results can be obtained with this brush with different directions of movement or different directions of approach to a spot. The brush can be clicked on an individual spot or it can be dragged across multiple spots or along a line. The Spot Healing tool is activated with the icon on the left vertical toolbar or by tapping the J key. Other tools share the icon and hotkey. If another tool is activated, change tools by right-clicking the icon or by tapping J again in Photoshop Elements or Shift-J in Photoshop CS.
The Clone Stamp Tool provides more control over corrections. The user specifies an area of the image that is to be used as the source for replacing another area of the image. As the clone stamp brush is clicked or dragged across another point in the image, the pixels under the brush are replaced with pixels from the source area. Not only can areas with spots and lines be replaced near edges, but damaged edges can be rebuilt by setting the source at a nearby area that has good edges. The Clone Stamp Tool is activated with the icon on the left toolbar or by tapping the S key. Other tools share the icon and hotkey. If another tool is activated, change tools by right-clicking the icon or by tapping S again in Photoshop Elements or Shift-S in Photoshop CS. For repairing defects, the checkbox for Aligned in the tool options bar will normally be cleared and Opacity will be 100%. The source for the tool is set by holding down the Alt key and clicking on a point in the image. Then move the mouse cursor to the area where these pixels should be pasted. When the brush is clicked or dragged, the corresponding pixels relative to the source point are pasted.
Defects that can be identified with color can be corrected with the Clone Stamp Tool using selections based on color. The effects of the Clone Stamp Tool are limited to only areas that are within the selection and that are specifically modified with the Clone Stamp Tool. For example, a stain on a document in an area with writing can be difficult to correct. In Photoshop Elements, the Magic Wand Selection tool is the best method to make selections by color. Clicking on a point on the image creates a selection of an area that has similar color and tone as the point clicked. If the icon for “Add to Selection” near the left end of the tool options bar is selected, clicking additional points adds the colors at those points to the selection. Another icon can be used to subtract colors from the selection. The Tolerance setting determines the range of colors that are included in the selection. Good control can be obtained with a relatively low tolerance setting such as 15 and the addition of other colors by clicking relevant points on the image. The Contiguous checkbox can be cleared. Having extra areas selected is not a problem because the actual parts of the selection that are modified depend on where the Clone Stamp tool is applied. In Photoshop Elements the Selection Brush tool on the left tool bar can be used to modify or fine-tune the selection. Painting with this tool expands the selected area, and holding down the Alt key while painting reduces the selected area.
In Photoshop CS, Color Range Selection provides a more precise selection by color. This is activated with the dropdown menu Selection> Color Range. The Fuzziness setting determines the range of colors that are included. Good control can be obtained with a fuzziness setting of 10 to 20. The small image in the dialog box can be set to Selection, and the Selection Preview set to “Quick Mask,” which displays the non-selected areas in red as in Quick Mask editing. If the Quick Mask setting makes it difficult to identify the colors that need to be selected, the Selection Preview can be changed to None, which shows the selected areas in white on the small image in the dialog box. The eyedroppers in the dialog box are used to control whether clicking a point on the image sets the initial color for the selection or adds or subtracts colors. After the color range dialog box is closed, the selection can be easily modified in Quick Mask mode. Tap the Q key or click the icon for “Edit in Quick Mask Mode” on the left toolbar. Use the Brush Tool to modify the selection. Painting white adds to the selection and painting black removes from the selection. The X key switches between white and black for the paint brush. Tapping Q or clicking the Quick Mask icon again returns to the main display with the modified selection. Precise selections can be made relatively quickly with the Brush Tool activated and frequent use of the Q, X, and spacebar keys.
The Healing Brush Tool functions similarly to the Clone Stamp Tool, except it blends the texture of the source area with the tone and color of the destination area. This blending of the source and destination areas eliminates defects without the precise matching required by the Clone Stamp Tool. The Healing Brush Tool is particularly effective with streaks or diffuse defects in large uniform or gradually changing areas. The Healing Bush Tool and Spot Healing Tool share the left toolbar icon and J hotkey. The source point for the Healing Brush is specified by holding down the Alt key and clicking on a point on the image. For repairing defects, the checkbox for Aligned in the tool options bar will normally be cleared and the Source will be set to Sampled. The Healing Brush Tool is also used for “enhancements” such as softening or removing the lines in a person’s face in situations when aesthetic qualities have priority over historical accuracy. Make a layer for applying the Healing Brush Tool and use the layer opacity setting to adjust the amount of the effect for such enhancements.
If the Adobe RGB color space is used for the master images, the color space for the display image will often need to be converted to sRGB for display on computer screens or for commercial printing. In Photoshop Elements, the conversion is done with the dropdown menu Image> Convert Color Profile> Convert to sRGB Profile. In Photoshop CS, this conversion is done with the dropdown menu Edit> Convert to Profile. In the dialog box, change the field for “Destination Space for RGB” to a sRGB profile in the pick-list, such as “sRGB IEC61966-2.1”.
Images used for printing from the hard drive can have the same size and resolution as the master images or working images. The print size will be set in the software making the print.
Images used for the internet will normally have the size and resolution modified to reduce the file size. For historical images the smallest size that would be used on the internet is typically about 600 pixels on the longest size. Larger sizes such as 800, 1000, or 1200 pixels on the longest side can be used. About 800 pixels on the longest size is often a good compromise for photographs. Images of documents may need 1000 pixels or higher for the writing to readable.
To modify the size and resolution of an image, tap Alt-Ctrl-I, or use the dropdown menu. In Photoshop Elements the dropdown menu is Image> Resize> Image Size, and in Photoshop CS it is Image> Image Size. In the image size dialog box make sure the checkbox for “Resample Image” is checked. The default resample method of “Bicubic (Best for smooth gradients)” is appropriate for historical images used for internet display. One way to set the size and resolution is to start by setting the resolution value to 200 ppi. Then set the value of either the width or height, whichever is longer. Use the value 3 inches to get 600 pixels on the longest side, 4 inches to get 800 pixels, 5 inches to get 1000 pixels, and 6 inches to get 1200 pixels. In Photoshop Elements, an image with 16 bits per channel can be resized.
The size of the output image can also be set with the option Save for Web & Devices that is described in the later section on Set Output Format.
Virtually all digital images of historical photographs and documents benefit from proper sharpening. Sharpening is needed to compensate for the softening of the image that occurs with digitization. Extensive adjustments to a digital image also tend to soften the image. In addition, further sharpening enhances the display of the image, particularly when the image is printed on paper.
Sharpening an image for printing is challenging because the sharpness of the image displayed on a computer screen does not accurately represent the sharpness of the image that will be printed. An image that has optimal sharpening for printing will appear to be significantly over sharpened on a computer screen because the computer screen has much lower resolution. The best method to determine the optimal degree of sharpening for printing is to compare actual prints with different degrees of sharpening. For important prints, I often make two or three prints with different degrees of sharpening and save the file with the settings that produced the best result. Having different degrees of sharpening on different layers in Photoshop makes the process of finding and saving the optimal result reliable and efficient. Layers that are not needed can be deleted and the remaining layers can be flattened for future use of the file.
The usual practice for developing archival master images is to apply sharpening only when an image is being prepared for display. Some scanning software attempts to sharpen a digital image as it is initially created, and some writers suggest that this initial sharpening is appropriate. This initial sharpening may be appropriate when a scanned image is intended to be directly displayed, but is not optimal for an archival master image that may be used for many different purposes over an extended period of time. Although not optimal, some initial sharpening of a master image is not a major concern in situations when it cannot be avoided.
To the extent possible, sharpening should enhance the features and detail in the image without enhancing the grain and noise. Different methods of sharpening are needed for different images. In particular, images with much grain and noise must be handled differently than images with little grain and noise. Non-optimal sharpening can easily negate the effects of grain and noise reduction. Three methods of sharpening and the conditions for their use are described below. These methods apply with both Photoshop Elements and Photoshop CS unless noted otherwise. If color shifts occur during sharpening, the color shifts can be reduced by setting the layer blending mode to Luminosity.
High Pass Filter Sharpening
High Pass filter sharpening is the method I use most frequently when third-party software based on a noise profile cannot be used or when additional sharpening is needed. High pass sharpening does not work well with images with high grain or when a mask is needed to limit the area of application.
Begin high pass sharpening by creating a new merged layer with Ctrl-J if this is the first layer or Alt-Ctrl-Shift-E if there are other layers. In Photoshop CS, the High Pass filter can be run as a smart filter by tapping Ctrl-Shift-S or by selecting “Convert to Smart Object” in the menu in the upper right corner of the Layers panel. Initiate the High Pass filter with the dropdown menu Filters> Other> High Pass. The image becomes displayed as gray. The only control on the dialog box is the radius of the filter. As the radius slider is moved to the right, the edges in the image are increasingly displayed. The optimum value for sharpening is usually in the range of 1 to 5, with the lower end of the range being safest for historical items. Move the slider until the edges in the image can be seen with relatively little texture in non-edge areas. Click Ok after the slider is set.
Then set the layer blend mode to Overlay, which displays the sharpened image. The Overlay blend mode prevents the use of layer masks to limit the area of sharpening. High Pass sharpening can be handled as an action with a hotkey using the methods described in Appendix E.
The sharpening effect can be increased by duplicating the layer with Ctrl-J. The sharpening effect can be reduced by setting the layer blend mode to Soft Light or by setting the layer opacity to less than 100%. My experience has been that one layer of high pass sharpening is often appropriate for display on a computer screen and adding one or two additional duplicate layers is often appropriate for prints.
Smart sharpening is the most flexible type of sharpening, but also can require more time to adjust and experiment with the different options. Begin by creating a new merged layer as described above and make it a smart filter if Photoshop CS is being used. In Photoshop Elements smart sharpening is initiated with the dropdown menu Enhance> Adjust Sharpness. In Photoshop CS, it is initiated with the dropdown menu Filter> Sharpen> Smart Sharpen. Set the Remove field to “Lens Blur.” Check the box for “More Refined” in Photoshop Elements and for “More Accurate” in Photoshop CS. The main controls are the Amount and Radius sliders. Images with more grain need larger values for the Radius slider and lower values for the Amount slider.
For historical images with high grain, the radius is often in the range of 10 to 20 and the amount is 10 to 40. For images with low to moderate grain, the radius is often in the range of 4 to 10 and the amount is 40 to 150. For images with little grain, the radius is often .5 to 2 and the amount is 75 to 200. Photoshop CS has additional options that are displayed when the Advanced button is clicked. The tabs for Shadow and Highlight bring up sliders that allow the amount of sharpening in the shadows and in the highlights to be reduced or faded. The range of tone that is included as shadows or highlights can be adjusted with the sliders for Tonal Width and Radius.
Neat Image Sharpening
Sharpening with Neat Image usually produces less grain and noise enhancement than the methods described above. The noise profile that is generated for reducing noise is used to minimize enhancement of noise during sharpening. It is particularly useful for images with very high grain or noise. I normally use Neat Image as the first step in sharpening.
Neat Image is activated as described above for noise reduction. Making it a smart filter in Photoshop CS is useful because the settings can be revised. After the noise profile has been generated and the “Noise Filter Settings” tab has been clicked, sharpening is activated by clicking the Y box in the Sharpening section at the bottom of the column. Normally the boxes for Cr and Cb are left unchecked and the box for Conservative is checked. The amount of sharpening is controlled by the High, Mid, and Low sliders in the sharpening section. My usual practice is to start with all three sliders at 0% and increase them to the lowest value that gives adequate sharpening or gives the best possible sharpening. The High slider is adjusted first and usually has optimal values in the range of 90% to 160%. The Mid slider is adjusted next and usually has settings in the range of 20% to 60%. The Low slider can make the output image have excess contrast and appear plastic. Settings in the range of 0% to 25% are usually best.
The settings for Neat Image can be explored by applying some of the built-in settings under the Advanced option—such as “Sharpen strongly out of focus image.”
Excessive sharpening, particularly with the Mid and Low sliders, can create a plastic appearance. If additional sharpening is needed, High Pass sharpening or Smart Sharpening can be applied after sharpening with Neat Image. Smart Sharpening is best for additional sharpening of an image with high grain or noise. Sharpening with Neat Image can be done as a separate step from noise reduction, but is usually easier if sharpening and noise reduction are done together. Other third-party noise reduction programs have similar sharpening utilizing a noise profile.
For prints on paper, TIFF is the optimal format for output files because it allows additional adjustments to be made if needed. However, high quality JPEG is also acceptable if no further adjustments are made or if another JPEG file is generated from the source TIFF file if adjustments are needed. A high quality JPEG image is not noticeably different from a TIFF image. However, modification of the JPEG file causes additional degradation and should be avoided if the file is being saved for long-term historical use. In Photoshop Elements and Photoshop CS, conversion to high quality JPEG is done with the dropdown menu File> Save As. Set the Format to JPEG. After the Save button is clicked, another dialog box comes up for the quality of the output file. For a high quality print file that may be enlarged or examined with significant enlargement, set the quality to 8 or higher.
The appropriate format for internet browsing is medium quality JPEG. Save the file in JPEG format as described above, but set the quality to 5 or 6. Typical file sizes are less than 100 KB when the image resolution and output format are set.
When saving files for long-term use, checking the boxes for “ICC Profile” and “User Lower Case Extensions” is appropriate on the first Save As dialog box. The ICC profile setting includes the color space profile in the output file, which will be useful as color management is increasingly implemented in the future. On the dialog box with JPEG quality, checking the box for “Baseline (Standard)” is safest.
Save for Web Option
Photoshop Elements and Photoshop CS both have an option Files> Save for Web & Devices that allows the user to explore different image size and format settings for files to be used on the internet. Although the save for web option can be useful, my experience has been that it is not intuitive and not well documented. It has fields to convert the output file to JPEG format, to set the number of pixels on each dimension of the output file, and to set the degree of compression for JPEG. The resulting file size and time to download at different internet connection speeds are displayed for the selected settings. The settings implemented with the save for web option apply only to the output image from the save for web dialog box. The main image is not changed.
The converted image is also displayed in the save for web dialog box but in a non-intuitive manner. The “original” image is displayed as a baseline for comparison; however this is the image displayed without a working color space applied, which is not expected in practice. An “optimized” image is also be displayed and is set to one of four conditions for comparison. The four options are: “Monitor Color” which is the same as the “original” image display, “Legacy Macintosh (No Color Management)” which displays the image as it will appear on a typical Apple computer without color management, “Internet Standard RGB (No Color Management)” which displays the image as it will appear on a typical Windows computer without color management, and “Use Document Profile” which is how the image will appear when color management uses the correct working color space. The last three options are useful for evaluating how the image will appear on the internet. In Photoshop CS the condition for optimized display is set with the Preview field and pick-list in the right column of the dialog box. Photoshop Elements has the same four options but they have the labels “Uncompensated Color,” “Standard Macintosh Color,” “Standard Windows Color,” and “Use Document Color Profile.” The option for display is selected in Photoshop Elements by clicking on the small triangle button icon at the upper right corner of the image display.
In Photoshop CS, but not Photoshop Elements, the save for web dialog box includes a checkbox to convert the output file to sRGB color space, which is optimal for the internet. With Photoshop Elements the image must be converted to sRGB before opening save for web. A checkbox that embeds the profile for the working color space in the output file is included for both Photoshop CS (“Embed Color Profile”) and Photoshop Elements (“ICC Profile”). In order to minimize the output file size for internet use, both Photoshop CS and Photoshop Elements do not write any EXIF documentation to the output file.
Intermediate and output files that are created for display should have file names and locations that clearly indicate the purpose of the files. As noted earlier, file names can have codes added such as _p to indicate a file for printing. Placing the print files and internet files in separate folders can be convenient if they are often copied or moved in groups. Adding codes to the file names is valuable for reducing confusion even if the files are kept in separate folders.
Andrews, Philip, 2011. Adobe Photoshop Elements 9 for Photographers. Published by Focal Press in Burlington, MA.
Evening, Martin, 2010. Adobe Photoshop CS5 for Photographers. Published by Focal Press in Burlington, MA.
Neat Image team, 2011. Neat Image plug-in for Photoshop (Win) - User guide. Accessed August 20, 2011 at http://www.neatimage.com/win/photoshop/userguide.html.
[Version of 5/2/2014]
Chapters in the Book
5. Preparation and Display of Historical Photographs and Documents
This website and book were developed by Jim Kennedy.Email: email@example.com
© 2012, 2014 James E. Kennedy