A colour Vision Test – Comparing results between computer screens

Abstract

Aims/Purpose: A colour vision test was developed to run on web browser without specific hardware. It can estimate observers' colour vision thresholds, response times and eye movements with a webcam. The aim of this work was to compare the chromatic thresholds (CT) obtained for normal colour vision observers when estimated on calibrated and uncalibrated computer screens.Methods: Four squares with random luminance noise were presented on a uniform grey background. One square was randomly selected to present a colour with a particular hue and saturation. Twenty‐six hues were tested on or near the dichromatic confusion lines. CT for each hue were estimated after a staircase procedure, and by computing the colour difference between the observers' CT and the reference white of the screen in use. The observers' task was to point and select the coloured square with the computer mouse. Two conditions were considered: a condition where the uncalibrated screen of the participants' laptop was used (L), and a condition using a screen calibrated in colour and luminance (D). The setup for D condition was equal across observers ensuring the same stimulus size and colour, observers' distance to screen and dark environment. The setup for L condition was uncontrolled but observers were guided towards providing information about pixel size, using a dark environment, disabling screen night light mode, using a laptop screen, and ensuring that screen brightness and zoom were adequate.Results: 48 normal colour vision observers performed this test on D condition, while a subset of 45 observers repeated this test on L condition. CT were averaged across observers for both conditions. It was found that CT are comparable across different screens and correlatable with traditional colour vision tests. It was also found that hues at degrees 157°,162°,167°,179°,185° and 347° had statistical differences (p < 0.05) when comparing conditions L and D.Conclusions: These results seem to indicate that the assumptions made to run the colour vision test on unknown computer screens with the instructions given, may be sufficient to obtain CT comparable to traditional colour vision tests, even if the test was done in non‐controlled conditions.