The USAF 1951 Resolution Test Chart on PhotoCD

This page is dual purpose; to explain the USAF 1951 test chart, and to establish the real-world resolution of Kodak Photo-CD.

The Chart

I submitted a copy of the chart, mounted in a standard plastic slide mount, to my local photo lab for conversion to photocd. The original chart resolves 160 line pairs per millimeter. The entire chart consists of as many copies of this test target as will fit on the film. Each target has five groups of six elements,each with five pairs of horizontal and five pairs of vertical lines. The actual size of the line pairs doubles with every seventh element. To determine resolution, simply identify the smallest element in which you can clearly distinguish all the line pairs, and read the corresponding resolution from the chart at the right.

Here is a snippet from a test group in the center.

Element Group 0Group 1Group 2Group 3Group 4
Element 110.0 lppm22.0 lppm40.0 lppm80.0 lppm160 lppm
Element 211.2 lppm22.4 lppm44.9 lppm89.8 lppmn/a
Element 312.6 lppm25.2 lppm50.4 lppm101 lppmn/a
Element 414.1 lppm28.3 lppm56.6 lppm113 lppmn/a
Element 515.9 lppm31.7 lppm63.5 lppm127 lppmn/a
Element 617.8 lppm35.6 lppm71.3 lppm143 lppmn/a

The Chart, verses theoretical PhotoCD resolution

Standard photocd's highest resolution scan is 3072x2048 pixels. Theoretically, this ought to be able to represent on the order of 42 line pairs/mm. As I read this scan, group 2 element 1 is clearly resolved. Group 2 element 3 is not clearly resolved, but still distinguishable from a uniform gray, and in a pinch you might be able to count the lines reliably. Referring to the chart, this corresponds to a resolution of between 40 and 50 line pairs per millimeter. So, at least by this lab on this day, the scan captured as much resolution from the film as it could reasonably be expected to do. Good!

A few other notes

This copy of the test target was taken from the center of the scanned image. Resolition was just a good at the edge. Also Good!.

The colorimitry of this image is a little bit off; the actual chart is clear, with black marks. This looks decidedly yellowish, and black isn't really black. Not so good.



comments/suggestions ; My photo page; My home page