There is a fundamental difference between two types of light mixtures that complicates the reproduction of images on monitors or in print. 

Our sight distinguishes millions of colors from two types of light mixtures: additive or subtractive. The former involves adding different parts of the light spectrum. In contrast, the latter consists of the subtraction of parts of the spectrum, allowing the transmission of the remaining portions. Computer monitors use an additive system, while print is subtractive. This difference complicates image reproduction.

Capturing Digital Images

Dynamic range is the span of difference between the image’s lightest light and darkest dark. The higher the dynamic range, the more shades are represented. However, the dynamic range does not automatically correlate to the number of tones reproduced. Therefore, dynamic range also describes a system’s ability to produce tonal information, which may be the most critical aspect of image quality.

Color depth is also known as sample depth or bit depth because a binary value represents digital color values. The number of bits representing each pixel, or the number of bits recording the value of each sample, determines how many colors can appear in a digital image. Higher color depth gives a broader color range. However, the more color information captured and stored, the higher the file size. 

8-bit grayscale is recommended for capturing and representing printed text, music manuscripts, hand- or typewritten papers, photographs, and graphic arts where color is unimportant. Grayscale images of this range create 256 shades of gray, and the human eye cannot detect any tone changes. 

24-bit color is recommended for archival-quality color images. A 24-bit color image can be created from a palette of 16 million colors, capturing a spectrum beyond human perception. In addition, a 24-bit color display is currently the highest bit depth available on affordable monitors. 

Resolution for Digital Images

Resolution refers to the quantity of detail described in pixel dimensions, such as 4000 by 5000, with the horizontal dimension provided before the vertical. DPI and PPI are standard terms used to express resolution. Dots per inch (dpi) measure the number of dots of ink a printer can produce. Pixels per inch (ppi) measure the number of pixels in an image within a linear one-inch space. 

From a digital project perspective, the most important difference between chemically based and digitally produced images is that analog images contain continuous tones. In contrast, discrete, measurable, and reproducible elements create digital images. Unlike analog photographs that vary constantly, the grid approximates digital images’ fine details and smooth curves, as continuous tonal gradients are broken into steps. For example, in physical photographs, enlargement reveals more detail but a grainier picture. Users cannot enlarge digital images beyond a certain point because they have limited spatial and tonal resolution; they show their microstructure as their pixels become prominent. Therefore, images should be captured and digitized at the highest quality practical. Transforming a high-resolution image into a lower-resolution image is easy, but doing the reverse without pixelation is impossible. 

Selecting the quality level depends on analyzing the images’ desired uses in the long term. For instance, undergraduates would find low-resolution images suitable for classroom use. In contrast, conservators assessing details would require high-resolution files. 

Depending on their type and quality, monitors display images between 72 and 100 ppi, with top models displaying 200 ppi, in which the pixels present a smooth appearance. Low resolution is suitable for preparing images to be displayed on the screen, such as on a website. Low-resolution images online prevent the usage of the images for commercial printing. However, printed images require 300 dpi to replicate the sharpness of their originals. Most inkjet printers can produce 300 to 1200 dots per inch. 

Masters and Derivatives

Digital files require master and derivative versions. Master files are of the highest quality available. The scans are rich enough to use without rescanning. From these, copies of the images create lower-quality, more easily distributed files. Master files should be digitized at 600 to 300 ppi and saved as TIFF images. Thumbnails and smaller-scale images, such as those suitable for online delivery, can be produced from these master files as 72 ppi JPEGs.

Suppose the digital image is resized without increasing or decreasing the number of pixels, the resolution of the image changes. For example, a 900 by 900-pixel image three inches wide has a resolution of 300 ppi. Each square inch contains 300 by 300 or 90,000 pixels. If the image is resized to six inches wide, without increasing the number of pixels, the 900 by 900-pixel image now has a resolution of 150 ppi as the 900 pixels are stretched over 6 inches instead of 3. Each square inch contains 150 by 150 or 22,500 pixels. Conversely, the resolution increases if someone decreases the image size without changing the number of pixels within the image. 

One and Done

Materials should be scanned once, creating a high-resolution image from which to derive smaller images to ensure that the originals have as little handling and light exposure as possible. If archivists follow best practices related to dynamic range, resolution, and formats, they can be confident that they will only need to perform digitization of their holdings once.