Zooming in GLIMPSE360, Spitzer’s Infrared Milky Way image (540,000 × 15,000 pixels), on WILDER, a touch-sensitive ultra-wall at Inria Saclay, France, that features a resolution of 14,400 × 4,800 pixels.Įarly work on ultra-walls has mainly focused on the technical aspects of how to make such platforms: 10 how to display complex graphics, how to stream data across the nodes of the computer clusters that drive them. We describe FITS-OW’s architecture and our solution to the specific challenges that this application raised: the generation of FITS tile pyramids and their multi-scale rendering the computation of sky coordinates queries to sky catalogs the dynamic adjustment of scale, color mapping and graphics compositing settings, and the framework for managing user interaction with various input sources. Indeed, ultra-walls are often driven by clusters of computers (the abovementioned WILD platform uses 32+1 graphics processing units in 16+1 computers) and involve heterogeneous input devices, which causes problems of data sharing, graphics rendering, and handling multiple user input channels. We also explain how we addressed related technical challenges. We explain how we enable astronomers to perform these operations using interaction techniques that were designed specifically for wall displays, using direct manipulation and gestures performed on the wall’s surface or on handheld tablets: adjusting the scale and color mapping used to render the raw, high-dynamic-range FITS data overlaying and manually compositing multiple bands measuring the brightness and colors of astronomical sources, displaying their light curve making queries and adjusting object filters. The very high pixel density of wall displays means that detailed information can be shown for multiple sources simultaneously, including multiple measurements as well as research articles retrieved dynamically through links found in relevant databases. Additionally, FITS-OW lets astronomers query databases such as SIMBAD 8 servers, and visualize the results of such queries in-place, right next to the corresponding source in the image. They can pan and zoom in images that are several hundred thousand pixels in both width and height, overlay the results of data analyses, fetch and display additional images of a specific object or region in the sky, showing observations in different ranges of the electromagnetic spectrum or made at different times. 5– 7 In this paper, we introduce FITS-OW, an application that enables astronomers to visualize and interact with very large FITS images and collections thereof. Wall displays also offer good support for collaborative work, enabling multiple users to simultaneously visualize and interact with the displayed data. They can represent the data with a high level of detail while at the same time retaining context: users can transition from an overview of the data to a detailed view simply by physically moving 4 in front of the display. ![]() ![]() Ultra-walls have properties that make them well-suited to the visualization of very large datasets. Other, even larger, platforms include Stallion 2 (328 megapixels) and the Reality Deck 3 (1,500 megapixels in an immersive configuration). ![]() For instance, WILD, the first wall display we set up in our laboratory, has a total resolution of 20480 × 6400 = 131 megapixels for a surface area of 5.5m × 1.8m. Ultra-high-resolution wall-sized displays, also called ultra-walls 1 for short, feature a very high pixel density over a large physical surface.
0 Comments
Leave a Reply. |