University of Illinois Library

1408 W. Gregory Drive

(MC-522)

Urbana, IL, 61801


Email:
dsd[at]library.uiuc.edu

About the Image Quality Calculator


What is the purpose of the Image Quality Calculator?

The Image Quality Calculator is a program that we have developed to determine the recommended resolution for scanning documents and visual materials. Based on information that you provide about your original document, the program calculates resolution in dots per inch (dpi), and indicates both the file size (in megabytes) and the pixel dimensions of the resulting uncompressed image file. The algorithms used in the calculator are based primarily on formulae developed by the Department of Preservation and Conservation at Cornell University Library for the scanning of text and manuscripts.

We developed the Image Quality Calculator in order to save time and to help us identify the best and the most appropriate image quality for a variety of uses. But, the IQC is only one of many components that comprise the digitization decision-making process. Other sites provide information that is invaluable in determining your purpose and planning your project. Before you use the IQC, we urge you to consult those suggested below, or others that you find useful:

Sitts, Maxine K. (ed.) Handbook for Digital Projects: A Management Tool for Preservation and Access.
(First Edition). Andover, MA: Northeast Document Conservation Center. URL http://www.nedcc.org/digital/dighome.htm.

Moving Theory Into Practice: Digital Imaging Tutorial. Ithaca, NY: Cornell University Library Department of Preservation and Conservation, 2000-2001. URL http://www.library.cornell.edu/preservation/tutorial/.

As you use this program, please bear in mind these recommendations:


We are always looking for ways to improve the Calculator--please send your suggestions to DSD.



What does the Image Quality Calculator do?


When using the Image Quality Calculator, you will perform the following steps:

  1. Supply height andwidth measurements in inches or centimeters for your original material. (Fractions are automatically converted to decimal equivalents.) Example: 8.5 x 11 inches OR 8 11/16 x 12.5 inches OR 20 x 30 centimeters
  2. Choose a quality index value from the list of three commonly used values that have been benchmarked. Example: High (QI = 8)
  3. Supply the height (in millimeters) of the smallest significant character (for text documents) or visual feature (for graphical materials, including photos, maps, or illustrations) that you feel must be resolved clearly in the scanned image. Example: 10 mm
  4. Choose a bit depth (bitonal (black and white, grayscale, or color). Example: Standard True Color (24-bit)



How are the results generated?


The IQC returns your input values along with calculated resolution and file size, based upon the values you have indicated for your document. Using the example below, the following results would be obtained:

Inputs: Results:
Image Width = 8.5 in
Image Height = 11 in
Quality Index = 8
Smallest Significant Character = 1 mm
Bit Depth = 24-bit
Recommended Resolution = 307 dpi
Approximate File Size = 25.212 Mb
Horizontal Pixel Dimension = 2609 Pixels
Vertical Pixel Dimension = 3377 Pixels




We use the following formulae in the IQC to compute optimal resolutions for bitonal, grayscale or color, and to compute the approximate file size (uncompressed TIFF format):


Optimal Bitonal Resolution Optimal Grayscale or Color Resolution
Resolution = (3.0 x QI) / (0.039 x h) Resolution = (1.5 x QI) / (0.039 x h)
3.0 = multiplication factor that takes into account misregistration and sampling errors for bitonal images
1.5 = multiplication factor that takes into account misregistration and sampling errors for grayscale or color images
QI = quality index
division by 0.039 (or multiplication by 25.64) = conversion factorfrom inches to millimeters
h = height of the smallest significant character or visual feature


Approximate File Size (in TIFF format)
File Size = ((height x width x bit depth x dpi 2) / 8) / 10242
height = height of the document
width = width of the document
bit depth = number of bits representing each color
dpi = optimal resolution (calculated)
division by 8 = conversion from bits to bytes
division by 10242 = conversion to megabytes