These images were collected by John A. Watlington from the Media Lab. The source of the images on this page can be found here.
Here is his description of the image collection procedure:
The experimental data consists of (basically) the same scene, under three different conditions of illumination :
- Condition C0: Illuminant A, Fluorescent lighting, a 3200K camera filter
- Condition C1: Illuminant B, Halogen lighting, a 3200K camera filter
- Condition C2: Illuminant C, Fluorescent and Halogen lighting, a 5600K camera filter
Three different scenes are lighted by the test illuminants, having different overall color balances (and accidentally, some specularities.) These are labeled by the color gamut present in the scene: all colors, little blue, and little red.
The video camera used was a Sony DXC-3000, which produced a composite video signal. The video was then digitized using a Digital Equipment Co. AV321 video card.
One problem with this test set is that the different illuminations were not co-located, causing shadows to be located in different positions. Another is that the same camera (and therefore sensor responses) was used throughout.
Images which contain "all" colors:
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
Images which contain "no blue" colors:
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
Images which contain "no red" colors:
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
Charles Rosenberg Last modified: Fri Jan 14 17:06:52 EST 2000