Wider fields of view from the sky to the microscope
Rubin Observatory's COSMOS image and a 48-sensor computational microscope offer two different arrangements for combining broad fields with fine visual detail.
Science··Midday
A first deep image of the COSMOS field
In the release reported by Phys.org, the NSF–DOE Vera C. Rubin Observatory opened its first LSST Camera image and accompanying catalogue of the COSMOS field. The new view of this patch of sky, previously studied by Hubble and James Webb, contains more than 500,000 galaxies and more than 50,000 stars. The 3.2-gigapixel LSST Camera sits on the 8.4-metre Simonyi Survey Telescope in Chile. Observations included in the release were collected between April 2025 and January 2026 within a second early-data package spanning roughly 3,000 square degrees. Observatory director Bob Blum says repeated visits to the region will show the survey design's capabilities. Access to the wider package currently remains limited to researchers in the United States and Chile and authorised international data-rights holders, followed by a two-year proprietary period before public release.[1]
Turning forty-eight sensors into one field
A computational microscope built by Kevin C. Zhou and colleagues at the University of California, Berkeley, treats a grid of 48 sensors on a credit-card-sized circuit board as one large sensor. In the study published by Nature Photonics on 28 July, the array supplies 48 times the pixels of a single sensor. A diffractive optical element in the Fourier plane addresses the physical gaps: light is spread into a distributed multi-spot point-spread function, and compressive-sensing algorithms reconstruct the missing information. The report attributes a further 5.4-fold throughput gain to this process. The system can reach about 3-micrometre resolution across more than 5.2 square centimetres at up to 120 frames per second, equivalent to 25.2 billion pixels each second. Reconstruction carries an assumption that the object is sparse in some domain. The team demonstrated the system in dark-field imaging and while tracking calcium activity across freely moving Caenorhabditis elegans worms.[2]
Two ways to build breadth
The two imaging arrangements combine breadth and detail at very different scales. Rubin's single 3.2-gigapixel camera places a defined sky field inside a large survey; the COSMOS catalogue makes hundreds of thousands of galaxies available in one release, while future visits to the same field add change over time to the survey design. The computational microscope divides its field among 48 smaller sensors, reconstructs light spread across physical gaps and joins a broad view to a high frame rate. In one case, coverage is assembled through the telescope, camera and repeated sky visits; in the other, through a sensor array, diffractive optics and compressive reconstruction. The Rubin release also carries a timetable governing who can access the wider data and when. The microscope paper explicitly states the sparsity assumption used by reconstruction. These details show a wide field as an arrangement of hardware, data processing and access conditions alongside the headline pixel count.[1], [2]
