This section presents subjective assessment results for the amount of visual discomfort induced by planar motion and in-depth motion characteristics :
1) velocity of horizontal motion: the average change in the horizontal visual angle and apparent depth for the planar motion;
2) velocity of vertical motion: the average change in the vertical visual angle and apparent depth for the planar motion; and
3) velocity of in-depth motion: the average change in the angular disparity.
3.1 Visual stimulus
A set of visual stimuli was generated with various velocities and directions of object motion using a computer graphics tool.
As shown in Fig. 20, these visual stimuli consisted of a grey meteor object (chromaticity: D65, illumination: 25 cd/m2), a background region (chromaticity: D65, illumination: 50 cd/m2), and a guide for a zero parallax position , .
As shown in Table 12, a total of 49 visual stimuli were generated (21 stimuli for horizontal motion, 21 stimuli for vertical motion, and 7 stimuli for depth motion). 42 visual stimuli had horizontal and vertical motions at seven different velocities, moving at 1°crossed disparity, zero disparity, and 1°uncrossed disparity, respectively. 7 visual stimuli had depth motion at seven different velocities.
The meteor object with in-depth motion periodically moved back and forth between 1° crossed disparity and 1° uncrossed disparity (the starting point of movement for the meteor object was 1° uncrossed disparity). The size of the meteor object was 2° except for the case of depth motion , when the size of the object changed continuously while the object moved back-and-forth in the depth direction. The size of the object was set to 2° when the object arrived at the screen position. Figure 20 shows some examples of the visual stimulus.
The visual stimulus with horizontal and vertical motions contained a pair of high contrast coloured bars and the visual stimulus with depth motion contained a high contrast coloured ring. The bars and ring were at the zero disparity so as to provide a depth plane of reference for viewers.
Attributes of visual stimulus used in the experiments
(c) Depth motion. For depth motion, the meteor object periodically moves back and forth
between 1° crossed disparity and 1° uncrossed disparity.
3.2 Subjective assessment method of visual comfort
A total number of 49 visual stimuli were randomly presented to each subject. The display duration of each visual stimulus was 10 seconds and the resting time followed for 15 seconds using
a mid-grey image. During the resting time, subjects assessed the overall level of visual comfort for each visual stimulus.
A total of 40 subjects, aged between 20 and 37, participated in the subjective assessment. The subjects were recruited under approval of the KAIST IRB (Institutional Review Board). All subjects had normal or corrected vision and a minimum stereopsis of 60 arcsec (as measured in a stereo fly test).
In order to grade the degree of visual comfort, the adjectival categorical judgment method of single stimulus (SS) was used with five-grade scale as shown in Fig. 21(a)
The subjective assessment methods were not standardized when assessing the visual discomfort of the stereoscopic content. As mentioned in Recommendation ITU-R BT.1438 , evaluation methods for the assessment of particular factors of stereoscopic television systems require further study. In many studies, meanwhile, the subjective test methods described in Recommendation ITU‑R BT.500 have been applied with slight modifications -.
When a reference image is available, the double-stimulus impairment scale (DSIS) or double-stimulus continuous quality-scale (DSCQS) methods can be used. When no reference is available, single-stimulus (SS) methods or the single stimulus continuous quality evaluation (SSCQE) method can be used. These methods were also mentioned in Recommendation ITU-R BT.1438, which specifically mentions the evaluation of stereoscopic content .
In this experiment, a modified version of the SS methods using a five-point category rating scale was used. In the SS methods, a sequence of images is presented and the assessor provides an index of the entire presentation . Hence, in some studies, SS methods have been used for assessing the visual discomfort of individual short video sequences (e.g. 10 seconds) without any reference sequence  and even of long video sequences (e.g. 15 minutes) . This was the case for our experiment.
In general, three types of SS methods have been used in television assessments: the adjectival categorical judgment method, the numerical categorical judgment method, and the non-categorical judgment method . As mentioned in Recommendation ITU-R BT.500-11 , categorical scales have most often used the ITU-R five-grade scales.
4: Perceptible, but not annoying
3: Slightly annoying
1: Very annoying.
Because the categories may reflect whether or not an attribute is detected (e.g. to establish the impairment threshold).
For this reason, the categorical scales in the SS methods adapted from the ITU-R impairment scale  were applied. The semantic terms of categories (i.e. very comfortable, comfortable, mildly comfortable, uncomfortable, and extremely uncomfortable) were the same as in .
The term of “very comfortable” indicates that visual discomfort is imperceptible. The term of “comfortable” indicates that visual discomfort is perceptible but not annoying. On top of this method, an additional questionnaire was used in order to investigate physical symptoms accompanied by the perceived visual discomfort. The two score sheets used in our assessment are illustrated in Fig. 21.
The questionnaire consisted of five terms: eye strain, general discomfort, nausea, focusing difficulty, and headache . A description of each term is as follows: