We measure microwave frequency (4-40 GHz) photoresistance at low magnetic field B, in high mobility 2D electron gas samples, excited by signals applied to a transmission line fabricated on the sample surface. Oscillatory photoresistance vs B is observed. For excitation at the cyclotron resonance frequency, we find an unprecedented, giant relative photoresistance (\Delta R)/R of up to 250 percent. The photoresistance is apparently proportional to the square root of applied power, and disappears as the temperature is increased.Comment: 4 pages, 3 figure

Walking-in-place techniques for locomotion in virtual environments typically have two problems that impact their usability: system latency (particularly troublesome when starting and stopping locomotion), and the fact that the change in the user's viewpoint may not be smooth and continuous. This paper describes a new WIP interface that improves both latency and the continuity of synthesized locomotion in the virtual environment. By basing the virtual avatar motion on the speed of the user's heel motion while walking in place, we create a direct mapping from foot-motion to locomotion that is responsive, intuitive, and easy to implement. In this paper, we describe the technique, analyze its starting and stopping latency, and provide experimental results on the suppression of false steps and general usability of the system. KEYWORDS INTRODUCTIONThe quality of a virtual-environment (VE) locomotion interface has a significant impact on the level of presence a user feels in a virtual environment [1,2] and the interface affects the way a user moves [3]. Although head-tracked real walking in VEs consistently evokes user behavior most like walking in the real world, locomotion by really walking is impractical in large-scale VEs, because the tracked space must be as large as the virtual space. Scaling high-precision tracking systems to arbitrarily-large sizes is expensive; wide-area tracking systems do not provide sufficient precision for a first-person display [4]. Because of these problems, the many VEs that require locomotion in large virtual scenes employ interfaces through which users can move their avatar (and viewpoint) to anywhere in the scene while remaining essentially stationary in the real world.Although various stationary-user locomotion interfaces have been proposed, previous research demonstrated that walking-inplace (WIP) is more presence-inducing than pointing interfaces [1,2]. One of WIP's greatest strengths is its similarity to real walking: The user controls their motion by moving their legs. From experience with our own WIP systems and others described in the in the literature [4,5,6, 7], we have identified two problems that impact WIP usability: system latency (particularly troublesome when starting and stopping movement), and the fact that the change in the user's viewpoint may not be smooth and continuous. Latency in visual feedback decreases the user's ability to precisely control their speed and stopping-position. _____________________ * e-mails: {feasel, whitton, jwendt}@cs.unc.edu In some systems viewpoint-movement is implemented as one or a series of discrete increments triggered by the detection of a step. This can lead to motion that feels jerky. A more natural motionsynthesis method is to distribute the locomotion over several frames, making sure that the change is smooth and the oscillations during sustained walking are not too great.Our system -called the Low-Latency, Continuous-Motion Walking-in-Place (LLCM-WIP) System -has four design goals. The following are the goals and how ...

Many Virtual Environments require walking interfaces to explore virtual worlds much larger than available real-world tracked space. We present a model for generating virtual locomotion speeds from Walking-In-Place (WIP) inputs based on walking biomechanics. By employing gait principles, our model – called Gait-Understanding-Driven Walking-In-Place (GUD WIP) – creates output speeds which better match those evident in Real Walking, and which better respond to variations in step frequency, including realistic starting and stopping. The speeds output by our implementation demonstrate considerably less within-step fluctuation than a good current WIP system – Low-Latency, Continuous-Motion (LLCM) WIP – while still remaining responsive to changes in user input. We compared resulting speeds from Real Walking, GUD WIP, and LLCM-WIP via user study: The average output speeds for Real Walking and GUD WIP respond consistently with changing step frequency – LLCM-WIP is far less consistent. GUD WIP produces output speeds that are more locally consistent (smooth) and step-frequency-to-walk-speed consistent than LLCM-WIP.

Figure 1. Fill reduction using CC shadow volumes: On the left, standard shadow volumes are rendered in transparent yellow. The middle image shows CC shadow volumes. The final rendered scene is shown on the right. In this environment, CC shadow volumes reduced the fill requirements by up to 7 times over standard shadow volumes.

Interwell two-dimensionaltwo-dimensional (2D-2D) tunneling in a density-imbalanced double-well heterostructure is studied in an in-plane magnetic field using a gated-bridge technique which does not require independent contacts to the two 2D electron layers. Two sharp peaks in the tunneling conductance at B~1.3 T and B =5-6 T are explained using a linear-response theory which incorporates finite scattering and is based on a field-induced momentum shift of the two quantum wells' Fermi circles. A second gate is used to locally enhance tunneling on a submicrometer length scale.The past several years have seen tremendous interest in double-barrier resonant-tunneling devices in which the tunneling is from three-dimensional (3D) to twodimensional (2D). ' In the presence of an in-plane magnetic field B, the resonant-tunneling I( V) peak undergoes a voltage shift and considerable broadening. Zaslavsky et al. proposed a geometrical construction, in which the dispersion curves for the 3D emitter and 2D quantum well (QW) are plotted in (k, k, E) space, with their origins offset by a B-induced shift in transverse momentum Ak. Requiring conservation of energy and transverse momentum, the tunneling is possible only when the dispersion curves overlap. More recently, 2D -2D tunneling has become of interest. Smoliner et al. examined nonequilibrium (i.e. , at bias voltages »kT) magnetotunneling between an accumulation and an inversion layer. Eisenstein et al. examined equilibrium magnetotunnelingin a double-QW system with initially balanced electron densities, as a function of both B and the QW's density ratio. A similar geometrical model, modified to the equilibrium 2D-2D case, was used to explain their results. In these works, the two 2D electron layers were contacted independently, either using diffused contacts, or by thinning the sample and patterning both sides with gates. ' In this work, we report on our studies of equilibrium 2D-2D magnetotunneling in a double-QW GaAs heterostructure with a built-in density imbalance. We avoid the difIicult task of contacting the two electron layers independently by using a gate to deplete the top QW across a small region of a resistance bridge, and then measuring the four-terminal source-drain resistance RsD(B). We establish a formal relationship between R sD (B) and the tunneling conductance G,""(B) in terms of a transmission-line model. G,""(B) is then obtained from the measured RsD(B). Two sharp peaks are observed in G,""(B) and explained in terms of the two QW's Bdisplaced Fermi circles intersecting one another tangentially. We present a linear-response theory, incorporating finite scattering times, which shows good agreement with the data. Finally, we use a second 0.3-pm-wide gate to locally control tunneling on a submicrometer length scale.Our sample is a molecular-beam-epitaxy (MBE) -grown heterostructure consisting of two 150-A GaAs wells separated by a 65-A Alo 3Gag 7As barrier. A 7X 10" cm Si 5-doped layer lies 800 A beneath the bottom well (QW1), while a 250-A-wide 10' cm Si...