Jaj Jacques Roufs

Text quality metrics for visual display units: I. Methodological aspects

A comparison of four experimental variables to measure the visual quality of text on visual display units (VDUs) is discussed. Scaled visual comfort is considered to be the most relevant psychological attribute. Visual performance, saccade length and fixation pauses during a search task approximating closely to reading are under the present conditions highly correlated with visual comfort if luminance contrast is the independent variable. All four variables are found to be satisfactory with respect to sensitivity, but it is argued that comfort scaling and the measurement of visual performance will probably suffice. Only small differences between subjects are found, suggesting that in the future a standard observer is feasible. The advantage of a print reference is argued.

Text quality metrics for visual display units : II. an experimental survey

This paper discusses several experiments aimed at evaluating the image quality of visual display units (VDUs). A number of quality metrics are presented, i.e. scaled preference, based on numerical ratings and paired comparisons; search velocity in a letter-search task; and two metrics derived from eye movement recordings. The former two metrics are found to be practical. The latter metrics, which are hard to determine, turn out to correspond with search velocity. With varying luminance contrast the scaled preference correlates closely with search velocity. With increasing resolution, however, this is true up to a moderate bandwidth, beyond which search velocity and the eye movement metrics saturate, while scaled preference continues to increase. Small quality differences across various commercially available VDUs are reflected by both search velocity and visual comfort. The results are consistent within subjects. Scaling differences across subjects can be understood by allowing for a subject-dependent weighting of underlying perceptual attributes like sharpness and brightness contrast. It is shown that visual comfort and performance are supplementary metrics that are both useful for modelling VDU image quality.

Prediction of thresholds and latency on the basis of experimentally determined impulse responses

As was shown before (Roufs and Blommaert 1981), temporal impluse responsses and step responses can be obtained psychophysically using a driftcorrecting perturbation technique. In this paper, experimentally determined impulse responses are given for eight subjects using different experimental conditions, i.e. a 1 deg stimulus field at background luminances of 1200Td and 100Td, and a point source superimposed on an extended background of the same luminances, which is a possibility to separate transient and sustained processing. For a large class of stimuli, predictions of threshold curves and latency of different time functions are calculated on the basis of these measured impulse responses. Predictions are tested against experimental data. It will be shown that a simple model, only consisting of a linear filter followed by a noisy peak detector, suffices for a fair quantitative description of the available data.

Instabilities in a continuous medium model for the retina

The shape of the spatial response of the retina on a small light stimulus as found by Rodieck (1965) for the cat, and proposed for the human retina among others by Korn and von Seelen (1972), resembles a Mexican hat known as a sombrero. A model presented by Röhler (1976), using a continuous medium as a description of the retina, can lead to such a “Mexican hat” response function for a specific choice of parameters. The spatial Fourier transform of this response function has a general appearance that corresponds to that calculated for stationary signals. However, such an analysis of the model for stationary signals is incomplete. Inspection of the time-dependent equations shows that it is unstable precisely for those parameter values that give the stationary response function its desired shape. Such stationary situations cannot be physically realized since the model is unstable.