Vision Research, introduces the newest addition to its family of High-Speed Machine Vision Cameras, the Phantom S710. The Phantom S710 brings unprecedented frame rates to machine vision applications, providing up to 7Gpx/sec (87.5 Gbps) throughput. This capability translates into over 7,200 fps at full 1280 x 800 resolution and over 700,000 fps at reduced resolutions. The S710’s streaming architecture allows continuous data flow for unlimited processing and recording, bringing new capabilities to Machine Vision applications.
The Phantom S710 draws from the extremely popular VEO710 high-speed camera and uses the same proven Vision Research CMOS sensor. With a 20micron pixel size, the S710 offers an ISO rating of 6400 (D) for monochrome cameras and 2000 (D) for colour cameras. To support high quality imaging, it also has 12-bits, a dynamic range of 59.6dB and a noise level of 29.0e-. Minimum exposure can be as low as 1 microsecond, and even down to 300 nanoseconds with the export-controlled FAST option.
The S710 16 CXP6 connections are arranged in four banks of four connections to stream varying levels of throughput. For example, one bank of four connections can stream up to 2Gpx/ sec (25 Gbps) to achieve just 2,145 fps in 8-bit mode, or 1,445 fps in 12-bit mode at full 1Mpx resolution. Two banks can stream up to 4 Gpx/ Sec (50 Gbps), achieving 4,290 in 8-bit mode, and four banks can stream the camera’s full framerate of 7275 fps in 8-bit or 5,750 fps in 12-bit.
In multi-bank operation, the image is separated and streamed by rows to be stitched back together once it is received by the frame grabber. The S710 is compatible with any PCIe3 CXP6 frame grabber; however, one frame grabber manufacturer has made the stitching function easier by incorporating it into the software of their Euresys 8-port Octo board. The camera also features a general-purpose input/output (GPIO) for fast, flexible signalling and synchronisation. It includes signals beneficial in standard high-speed applications, such as Time Code In and Out, as well as other signals commonly found in streaming applications.