Comparing mobile phone liquid chrystal display optical measurements with end-user evaluations in bright outdoor illuminations

Abstract

While specifying the features of a mobile phone display, the properties are specified as objectively measurable parameters. Optical measurement of displays allows comparisons between different manufacturers and also aids quality control in the manufacturing process. It is important to know how well these measurements correlate with the end-user experience so that the specifications specify the correct parameters and values.

The goal was to verify if the methods used in the study are suitable when comparing optical measurements with subjective evaluations and if there are suggestions for improvements to either the measurements or the evaluations based on the results of the study.

Three displays were chosen for the study. Of the three displays, only one had both a decent reflective color gamut and a decent reflectivity. Of the two other displays, one had a decent reflective color gamut and the other had a decent reflectivity.

The images in the test were chosen to be representative of the use scenarios that are likely to be important outdoors. Four images were used to test the visibility of the image and two images were used to test the colors in the image. The difference in how to score the two groups of images was explained to the observers before the test.

The optical measurements were carried out using a Display Measuring System DMS-703 measurement device. Luminance and reflectivity were measured, along with color gamut and contrast both in transmissive and reflective mode.

The illumination conditions in the end-user tests were created in the SimOne lighting simulator, which is an experimental laboratory used to simulate different lighting conditions in mobile use. Three different illuminations in the 10,500 - 47,000 lx range were used for testing the displays and the images.

The optical measurements predicted the evaluation results quite nicely. The display with both a decent reflective color gamut and a decent reflectivity was generally the best in this study, even if it was only second best in the test cases for both visibility and colors.

Dividing the end-user test images into those that were used for testing the visibility and those that were used for testing the colors was quite successful. The study also showed how careful one has to be with selecting the correct questions and images for end-user tests. Quite small changes in the test instructions yielded very different results.