A user-friendly tuning system is needed to specify, control, and adapt the temporal behavior and other quality parameters of audio/visual multimedia. The goal of this project is to explore the tuning service with its visualization techniques for quality of service (QoS); specification of bidirectional translation between different views of QoS with emphasis on translations for compressed media and reverse translation from the system to the user; signaling between the tuning service, the application, and the network in case of adaptive behavior; application-driven ne gotiation protocols to incorporate heterogeneous configurations of multimedia devices and different loads at the end-points; and distributed timing control.

As the presence of audio/visual support in multimedia distributed applications increases, the user needs a friendly tuning system to specify, control, and adapt the temporal behavior and other media quality parameters. The goal for this project is to explore the visualization of quality parameters, specification of bi-directional translation between different views of quality parameters at the user and system layers, control of media quality and signaling between the tuning service, and the underlying system in case of quality change.

New transducer technologies now make it possible to sonically and visually immerse a user in data. From a sensory perspective, the computer need not be simply the medium of interaction between user and data; instead, users can interact directly with the data. We are exploring techniques that immerse users in a virtual world based on dynamic data that describe the behavior of massively parallel computer systems and the World-Wide Web (WWW). Through a head-mounted display, three-dimensional sound cues, and direct data manipulation, the user can explore, viscerally experience, and modify the dynamic behavior of a high-performance parallel system or a WWW server.