Exploring 3D Navigation : Combining Speed-coupled Flying with Orbiting
Virtual environment encompass mare than what can display one point of view. For a good user experience, it is necessary to take care of the navigation to allow the user move intuitively in the environment. The design of navigation can be treated as a task-based model and can be subdivided into three subclass. Exploration consists of gaining a survey knowledge, search consists of locating an object or a route and inspection consists of establishing nad maintaining a certain point of view. The abstract solution can be defined in terms of travel controls, which are speed, position and orientation of the camera. In every case, the level of immersion affects the navigation tasks and the associate performance.
Four methods are described. For inspection, object manipulation and ghost copy can be used. The user drags on a object and then manipulate a copy of this object. It can select its position and orientation. When this action is finished, the user is teleported to the exact location that provide the specified orientation and scale to the object. The second technique is called inverse fog or scaling an ephemeral world and is used for searching. The user controls the radius of a sphere. Each object that is in the sphere is either make transparent (inverse fog) or scaled down (ephemeral world). The two others techniques have been elaborated by the authors. Possession and rubbernek navigation is quite simple. When a user drags on an object, he is teleported to the inertial center of this object. It can then chose the orientation and create on the fly a new camera. The last technique is called speed coupled flying without orbit and is based on the behavior of the cursor. If the user drags on an object, the system assumes he wants to examine the object. Otherwise, when the user drags on the floor or wall, the system assumes he wants to freely moves around the object. To mark the difference between the two behaviors, the cursor changes when it is on an object. Experiments show that the last technique, speed coupling without flying orbit offers the better result.
A survey of design issues in spatial input
Spatial inputs are interfaces based on free-space 3D input technologies such as camara-based. They have their issues, which are discussed in this paper. The human perception with 3D space is more about experiencing than understanding. Indeed, for a given tasks, it is highly likely that users will more try combination of 3D objects to accomplish a task than mentally visualizing the combination in order to define if that combination works. To help the experience, spatial references, which are real word objects, have to be used to help users interacting in 3D. There is also a choice to make between absolute gesture and relative gesture. The one is relative to the global frame when the second can be relative to a more logical frame, such as the user body frame. Experiments have prooved that users understand two handed interfaces in only few minutes. Hence, using those kind of interfaces can provide more natural useful interactions. Multi-sensor feedbacks, like force feedback, space exclusion or auditory feedback, can also provide useful information to the user. Head tracking is another way to provide feedback. Users have to understand the limitations of the software. This cognitive load time can be reduce or remove using physical constrains that seem natural for humans.
Designers should also be careful to not let too many degrees of freedom to users because it can be confusing and perturbing during an action if there are more possibilities available than needed. The right kind of camera metaphors have to be chosen. 4 possibilities are camera metaphor, scene in hand metaphor, flying vehicle metaphor and ray-casting metaphor. To avoid or reduce issues in dynamic acquisition, transparency can be used to provide information about occlusion or context. Sometimes, ray-casting is sufficient and can be a more convenient selection method than direct positioning. If a gross selection is enoungh, cone casting, which consists of defining an area of interest can be an other user-friendly method. Finally, designers have to be particularly concerned about clutching mechanism, who allow a user to move the spatial input device without changing the virtual environment.