Visual fatigue while watching 3D stimuli from different positions

Abstract

Purpose: When observers focus their stereoscopic visual system for a long time (e.g., watching a 3D movie) they may experience visual discomfort or asthenopia. We tested two types of models for predicting visual fatigue in a task in which subjects were instructed to discriminate between 3D characters. One model was based on viewing distance (focal distance, vergence distance) and another in visual direction (oculomotor imbalance).

Method: A 3D test was designed to assess binocular visual fatigue while looking at 3D stimuli located in different visual directions and viewed from two distances from the screen. The observers were tested under three conditions: (a) normal vision; (b) wearing a lens (-2 diop.); (c) wearing a base-out prism (2▿) over each eye. Sensitivity and specificity were calculated (as Signal Detection Theory parameters: SDT).

Results: An ANOVA and SDT analyses revealed that impaired visual performance were directly related to short distance and larger deviation in visual direction, particularly when the stimuli were located nearer and at more than 24° to the centre of the screen in dextroversion and beyond.

Conclusion: This results support a mixed model, combining a model based on the visual angle (related to viewing distance) and another based on the oculomotor imbalance (related to visual direction). This mixed model could help to predict the distribution of seats in the cinema room ranging from those that produce greater visual comfort to those that produce more visual discomfort. Also could be a first step to pre-diagnosis of binocular vision disorders.

Keywords: Accommodation; Acomodación; Astenopia; Asthenopia; Binocular vision; Estereopsis; Fatiga visual; Stereopsis; Vergence; Vergencia; Visión binocular; Visual fatigue.

Figure 1

Upper panel: 3D Stimuli (anaglyphs) used in the discrimination task. The top row of alphanumeric characters is seen as floating in front of the reference plane (crossed disparity), while the lower row of alphanumeric characters is perceived as being behind the reference plane (uncrossed disparity). Lower panel: An piece of a trial of the test with 22 3-D alphanumerical characters. Subjects see them through red/cyan filter glasses. Because of the large extension, in this line there are only twelve of the 24 characters presented in every line of the test.

Figure 2

Left panel: Mean sensitivity scores in the SDT paradigm as a function of the viewing distance and according to the three visual conditions for zone 4. *p < .05 normal visual condition compared to ‘Lens’ and ‘Prism’. Error bars indicate the standard errors of the mean (95% confidence level). Right panel: Mean specificity scores of the SDT paradigm for each viewing distance, depending on the zone on the screen. Error bars indicate the standard error. *p < .05 Zone-4 compared to the other zones. Error bars indicate the standard errors of the mean (95% confidence level).

Figure 3

Results of the survey to assess participants’ subjective discomfort (in Likert scores). The participant reported no other problem and/or sensation. Therefore, item 6 was removed. Error bars indicate standard error.

Figure 4

Rescaled version of the ‘mixed model’ fatigue F (by applying the rescaling plotted values range from 0 to 1), as a function of the position in the movie theatre, for different choices of the ρ parameter defined in the text. Left: ρ = 0.1, middle: ρ = 0.5, right: ρ = 0.9. As can be seen, small ρ values yield F functions similar to the visual angle itself, while large ρ values result in F forms resembling the oculomotor imbalance. Distances are in m, and angles in degrees.

Figure A.1

Optical-geometrical relations for a given observer located at (X, Z), relative to a screen in a 3D cinema. The X coordinate goes along the screen itself, Z is perpendicular to X, and the origin 0 is set at the screen centre, which is also the centre of the stereogram. Note that d = ▿obj/2, I = IPD/2. The left panel (A) shows the case of an observer who remains still but moves his/her eyes, whilst the right panel (B) shows the case of an observer moving his/her head towards the screen.

Figure A.2

Left panel: Heat maps for focal distance D_f (up panel) and vergence γ (middle panel) as functions of the observer's location in the cinema theatre. Heat map for the visual angle subtended by the target stereogram as a function of the observer's position in the cinema (down panel). Distances are in m, and angles in degrees. Right panel: upper panel: Heat map for the D_c/D_f ratio as a function of the observer's location in the cinema theatre. Lower panel: Heat map for the oculomotor imbalance |β₁ − β₂| as a function of the observer's location in the cinema of Figure 2, Figure 3, Figure 4, for a situation in which the interocular line is kept parallel to the screen (the case of cross-disparity). Distances are in m, and angles in degrees.