Experimental Guideline; (Draft: 2000/12 revised)

!Note: Thi s draft is provided merely for TC member's disucussion, and not for public review!

Matching;

<Default>

Softcopy image vs. Hardcopy image (under same ambient illumination)

<Option>

Softcopy image vs. Hardcopy image (under different ambient illumination)

Softcopy image vs. Softcopy image (under different ambient illumination)

Image/Patch;

<Default>

Pictorial Images - more than 3 different images
(most of colors in the image should be within both input/output devices' gamut)
Should include "party" and "picnic".

<Option>

Uniform Color Patches

Image size: 9-15cm (4-6 inch) x 12-20cm (5-8 inch)

Proximal Field: 100% device white of 5 mm (1/4 inch) wide,

Back Ground: 20% device white (or 50% in L*) with a reference 100% device white field

Monitor's White Points;

<Default>

9300K or 6500K

<Option>

5000K and other CCT's

Ambient Light;

Note; Whatever illuminants are used, spectral data should be measured and reported.

<Default>

F11(4000K) or F10(5000K)

<Option>

CWF with other CCTs, F2/F6(4000K) or A illuminant

Luminance of Monitor's White Point vs. Paper White

Note; Hardcopy luminance (from ambient light) should be measured at 45 degree, and should be reported.

<Default>

80-100 cd/m2 vs. 80-100 cd/m2 (similar luminance)

<Option>

80-100 cd/m2 vs. 200 cd/m2

80-100 cd/m2 vs. 300 cd/m2

Observers

<Default>

Number of the observer should be more than 15,
Before the experiments, observer has to take a color deficient test, such as Ishihara Test or ColorytePlates by Budapest University. Those who are color deficient can participate in the experiments,but experimenter must report that the observer was color deficient .

Matching Method;

<Default>

SMB (Simultaneous Binocular Method)

<Option>

MEM (Memory Matching Method)
SCB (Successuve Binocular Method)
SMH (Simultaneous Haploscopic Method)

Paired comprison is most common procedure for pictorial image comparison.

Observers' Adaptation

Before the experiment, observers are given approximately 3-5 minutes to adapt to the viewing conditions of the experimental set-ups. The observer is instructed to sit approximately 50-60 cm from the screen. The observer can move the pair of images anywhere he/she desires, but not on the screen next to the softcopy image, so that the observer have to move his eyes at some distances for the image comparisons. No time restriction is placed on the observers (- will not be fixed to single state).

Accuracy of Device Color Characterizations

Accurate device characterization is of great importance in cross-media comparison experiments. All the device must be stable enough, and characterization accuracy should be within certain tolerances. As a rule of thumb, CRT monitor could be characterized as good as deltaE*ab of 2~3, and Printer could be characterized as good as deltaE*ab of around 5.