Lo que dicen en la entrevista

In our study we had four different interface conditions all of which allowed interaction through video. TheOriginalcondition enabled users to observe and manipulate content through the pure camera image without any zoom capabilities provided (see figure 7.13a). This condition is our control condition during the test. In theManual Zoomcondition, users were able to zoom in using a slider on the mobile device (see figure 7.13b). The maximum digital zoom was restricted to the factor four. In our test, zooming in more was unnecessary as the maximum simulated distances to displays did not require higher zoom factors. Participants could decide on their own whether they zoomed in at all or how far they want to zoom in. In the Auto Zoomcondition, the mobile device zoomed in automatically based on the distance to the display to maintain a constant CD ratio (see figure 7.13c). In our study, the apparent size of objects shown on the mobile device was constantly 1.2 inches. In the Freeze condition, users were able to freeze the image by tapping

the freeze button (see figure 7.13d). The condition also featured automatic zoom. The frozen

image was then overlaid by computer-generated digital image of the external display’s portion the mobile device was pointed at. Tapping the button again restarted the live video. Participants were free to choose whether to use freeze.

Figure 7.13:The fourTouch Projectorinterfaces used during the study: (a) Original camera interface. (b) Manual zoom capabilities. (c) Automatic zooming. (d) Freezing the camera image with temporary overlay for precise interaction.

During the study, participants had to perform two different types of tasks. In thetargeting task

participants had to acquire targets on a distant display. As illustrated in figure 7.14, astart button

was shown on the target screen at the beginning of each trial. After tapping the start button, it disappeared and the target item appeared on the display. Participants were now able to aim at the target display and the item respectively to tap on it. If they missed the target, the error count was increased by one and participants tried to acquire the target again. Once they successfully selected the target item, it disappeared and the start button was shown again to indicate the next trial. We measured the task time from the moment the start button has been tapped until the target item has been selected successfully. For each trial, the target had one of three apparent sizes on the mobile device’s screen: 0.3 inches, 0.6 inches, or 1.2 inches. The apparent size was varied to simulate three distances to the target display as our lab did not have sufficient space. Therefore, we kept the distance to the target screen constant and instead varied the target size.

In thedragging task participants dragged an object from one display to another. On the origi- nating display, the target’s apparent size remained constant at 1.2”. As shown in figure 7.14, the setup consisted of two displays. Similarly to the targeting task, the beginning of each trail was indicated by a start button being shown on the originating screen. At the same time, the target’s

drop area(i.e., a frame with the target’s apparent size on the screen) was shown on the destina-

tion display. After tapping the start button, the trial started by showing the item to be transferred. Participants were then able to aim at the item and acquired the target with a touch-and-hold op- eration. If they acquired the wrong item, an error was logged and participants had to repeat the trial. Now they could move the mobile device while having the finger still on the display until

7.3 Evaluation of the Improvements 153

they reached the destination screen (i.e., it was visible in the mobile display’s viewfinder). While not being pointed at a screen, the mobile device showed a thumbnail of the item as described earlier. Once the destination display was visible, the item was transferred from the mobile device to the target display. Participants could now move the item to its exact location indicated by the

drop area. Lifting the finger initiated the transfer. If the item’s center was located in within the

boundaries of this area, the trial was completed successfully. If the item’s center was outside, participants had the option to correct it. Once a trial was completed, the item disappeared and the start button was shown again on the originating display.

Figure 7.14: A participant performing thetargetingtask. She first starts by tapping the start

button (a) and then selects the target item (b).

In this task, we measured task time and the item’sdocking offseton the destination display. This offset describes the percentage of the item’s area being located outside of the target area. Sim- ilar to the targeting task, we varied the apparent size of items on the destination display while the item’s apparent size on the originating screen remained constant at 1.2”. On the destination display their apparent size again had three possible values: 0.3 inches, 0.6 inches, or 1.2 inches. We further varied the angular distance between both involved external displays. This distance had one of three values: 45◦ (slightly left of the source display), 90◦ (directly left of the par- ticipant), or 180◦ (exactly behind the participant). The setup is shown in figure 7.15. In both tasks, we asked our participants to perform the task as quickly as possible while maintaining a low error rate. Furthermore, the simulation of apparent sizes caused the mobile device to zoom in automatically when in theAuto ZoomorFreezecondition.