ABSTRACT
Existing near-eye displays (NEDs) have trade-offs related to size, weight, computational resources, battery life, and body temperature. A recent paradigm, beaming display, addresses these trade-offs by separating the NED into a steering projector (SP) for image presentation and a passive headset worn by the user. However, the beaming display has issues with the projection area of a single SP and has severe limitations on the head orientation and pose that the user can move. In this study, we distribute dual steering projectors in the scene to extend the head orientation and pose of the beaming display by coordinating the dual projections on a passive headset. For cooperative control of each SP, we define a geometric model of the SPs and propose a calibration and projection control method designed for multiple projectors. We present implementations of the system along with evaluations showing that the precision and delay are 1.8 â¼ 5.7 mm and 14.46 ms, respectively, at a distance of about 1 m from the SPs. From this result, our prototype with multiple SPs can project images in the projection area ($20\ \text{mm} \times 30\ \text{mm}$) of the passive headset while extending the projectable head orientation. Furthermore, as applications of cooperative control by multiple SPs, we show the possibility of multiple users, improving dynamic range and binocular presentation.
ABSTRACT
Cu/Zn-superoxide dismutase (SOD1) is a homodimer with two identical subunits, each of which binds a copper and zinc ion in the native state. In contrast to such a text book case, SOD1 proteins purified in vitro or even in vivo have been often reported to bind a non-stoichiometric amount of the metal ions. Nonetheless, it is difficult to probe how those metal ions are distributed in the two identical subunits. By utilizing native mass spectrometry, we showed here that addition of a sub-stoichiometric copper/zinc ion to SOD1 led to the formation of a homodimer with a stochastic combination of the subunits binding 0, 1, and even 2 metal ions. We also found that the homodimer was able to bind four copper or four zinc ions, implying the binding of a copper and zinc ion at the canonical zinc and copper site, respectively. Such ambiguity in the metal quota and selectivity could be avoided when an intra-subunit disulfide bond in SOD1 was reduced before addition of the metal ions. Apo-SOD1 in the disulfide-reduced state was monomeric and was found to bind only one zinc ion per monomer. By binding a zinc ion, the disulfide-reduced SOD1 became conformationally compact and acquired the ability to dimerize. Based upon the results in vitro, we describe the pathway in vivo enabling SOD1 to bind copper and zinc ions with high accuracy in their quota and selectivity. A failure of correct metallation in SOD1 will also be discussed in relation to amyotrophic lateral sclerosis.
