COLOR BASICS

But wait! Before you venture out in the field with your digital camera and tripod or reset your scanner to a higher bit depth, it helps to understand the principles behind the color information within pixels. An image’s bit depth determines the amount of information contained within each pixel. Bitmap images (see Figure 1) are true black-and-white graphic images and contain one color channel with one bit of information per pixel and two potential colors, literally white or black. It’s like a light switch, it’s either on (white) or off (black). Since bitmap images contain a minimum amount of information, their file sizes are relatively small. 

 

Grayscales (see Figure 2) are used for what we traditionally think of as black-and-white images.  Each pixel contains eight bits of information. To use the light switch analogy again, there are eight switches that in combination can be either on or off. The more switches that are on, the brighter the shade of gray. When all the switches are on, white is produced. Conversely, darker shades are produced by more switches being turned off, and when all the switches are off—you guessed it—black is produced. A grayscale image is composed of one channel with 256 possible shades of gray. The formula, 28=256 is derived from the potential number of gray values including white and black that eight bits of information can generate. 

 

8-bit RGB (see Figure 3) images consist of three colors—red, green, and blue—in three 8-bit color channels. These images are sometimes referred to as full color. Once again, the light switch, but this time we have three colored bulbs—a red, a green and a blue. Each bulb can produce up to 256 shades of its color. Instead of bulbs, we actually have three color channels. In Photoshop, there is also a composite RGB channel that combines the color information into a full-color image. Eight-bit RGB can produce a total of 2,563, or 16,777,216, possible colors.

 

HIGH-BIT IMAGES

Images that contain pixels with 12, 16 or 32 bits of information per pixel are sometimes referred to as high-bit color images. Few monitors can accurately display high-bit images. What’s more, no printer has the capability of printing the billions of colors that high-bit images can potentially produce, and the debate rages on over whether images photographed or scanned as high-bit images and converted to eight-bit before printing are significantly better in quality than images captured in eight-bit color from the get-go.  

 

Proponents of high-bit image technology claim that scanning or photographing in high-bit color initially captures more colors, resulting in less severe banding (see Figure 4) in graduated areas and less degradation as a result of the behind-the-scenes math applied to pixels during editing sessions. True, the histogram of an eight-bit image may show several gaps after a number of image-processing edits, (see Figure 5) and smooth gradients may tend to exhibit banding on screen. In a 16-bit image, for example, the number of visible gaps in these histograms will be reduced and banding may be less severe on screen. 

 

Figure 5: Histograms of the same image scanned in 8-bit color (left) and 16-bit color (right) after tonal manipulations have been performed.

Skeptics of high-bit image technology ask weather the print will be noticeably affected by all this voodoo and make the case that in the early part of this decade, poor scanning and digital camera technology was the primary cause of banding and posterization in eight-bit color images as a result of incorrect math in the scanner drivers. This problem resulted in the development of the high-bit workflow as a solution.

 

But over the last few years, scanner and digital camera technology has improved. Photoshop image-processing has been refined and has become more accurate to the degree that high-bit workflows are no longer essential for editing except for a small percentage of images. 

 

IMAGE CAPTURE TIPS

While the debate rages on, let’s discuss the subject of image capture. Here are a few image capture tips that will improve your final output.

 

Avoid exceeding the optical resolution of the scanner. Although a flatbed scanner may show 9,600 dpi as its maximum scan resolution, many scanners today don’t exceed 2,400 dpi as their optical capability. Exceeding the optical resolution of the scanner is the same as increasing the images file size by resampling, a practice that can reduce sharpness and contrast.

 

If you have a scanner that produces banding, posterization or dithering with either eight-bit or 16-bit images, acquire a more current scanner driver or get a new scanner.

 

Scan in 16-bit black and white or color and then duplicate (Image >Duplicate) and save the file. Perform whatever edits you can in 16 bit (some Photoshop features are not available) and convert the image to eight bit (choose Image > Color Mode > 8 Bit/channel) when you need apply an unsupported command or filter or before printing. Use the 16-bit original as a backup.

 

Shoot in Camera Raw Large whenever possible and change the bit depth to 16 bit in the Camera Raw dialog box.

 

When you work in the 16-bit workflow, do so during the initial stages of your Photoshop work session. Apply tonal adjustments, such as levels, curves and color balance to the image as this is where the severest image deterioration may take place. Then convert to eight-bit before printing.

 

HDR IMAGES

Dynamic range describes the ratio between the maximum (white) and minimum (black) intensity of light of a specific device. The higher the dynamic range of a device, the greater the range of tonality it can capture or print. High Dynamic Range (HDR) images combine multiple exposures to improve the tonal range of your pictures. HDR is a group of techniques that produce an increased range of luminance within light and dark areas of an image. The intention of HDR is to accurately represent the wide range of intensity levels found in real scenes ranging from direct sunlight to shadows. The process of tonal mapping and image alignment in combination with bracketing exposures of normal digital images can result in a high, sometimes even exaggerated dynamic range. In Photoshop, HDR images are achieved by combining a series of bracketed exposures using a Photoshop automation—the Merge to HDR Pro command found under the File > Automate menu. The latest release of this automation in Photoshop CS5 has big improvements that make it more accurate and more user-friendly.

 

Figure 6: Three bracketed images displaying an emphasis on the highlight areas (left), midtones (center) and shadows (right). (Photo courtesy of Southwest Interior © 2009 Ken Barr)

PHOTOGRAPHING HDR IMAGES

First of all, let’s talk about the best way to capture images that will be combined to produce HDR. Two or more bracketed exposures of the same scene are necessary (see Figure 6). When shooting the exposures, the camera is mounted on a tripod and a cable release is used to assure stability. Any number of bracketed images is acceptable but best results are achieved when areas of the highlight, midtones and shadows are emphasized separately on each of three of the bracketed images with additional images to function as in between tonalities. The HDR Pro automation combines the bracketed photographs into a single image with enhanced detail in the three tonal ranges. 

 

Vary the shutter speed when making the bracketed exposures. Do not vary the aperture or the depth of field or it will be difficult or impossible to align and blend the images. Maintain the same ISO setting for each exposure to reduce variations in shadow noise. Bracket with 1 or 2 exposure value step intervals. Smaller steps will produce more images than are necessary.

 

MERGING THE IMAGES

Combining the various exposures to into a single HDR image is performed with the automation in Photoshop (File > Automate > Merge to HDR Pro), or with Adobe Bridge (Tools > Photoshop > Merge to HDR Pro).  You can composite separate images or, if you’ve composited all of your exposures to a single document on multiple layers you can pre-align selected layers using Photoshop’s Auto Align Layers feature and then apply the Merge to HDR Pro automation.

 

Figure 7: The Photoshop CS5 HDR Pro interface combining seven exposures into a single image with balance in all three tonal ranges. (Photo courtesy of Southwest Interior © 2009 Ken Barr)

The HDR Pro interface (see Figure 7) offers a control panel that enables color and tonal manipulation. Choose a mode; 8, 16, or 32-bit color. The higher the bit depth, the more pixel information is combined and thus, the larger the file, but be aware that there are many features in Photoshop that do not support 32-bit images. Edit the image with sliders found in the Edge Glow, Tone, Detail and Color fields. As you make adjustments, the results preview in the image window of the interface. When the image is at its best, click the OK button to implement the HDR process.

 

High-bit and HDR images can potentially produce better-looking images. Your decision to use these features will depend on the image itself. Does it need enhanced dynamic range or higher bit-depth to accentuate the contrast in the highlights and shadows? The target printer is also a consideration. Is it capable of printing that extra detail? If the answer to these questions is yes, then consider incorporating high-bit and HDR image technology into your workflow.