VisiSonics, a new technology development company, will demonstrate a startling new technology at the upcoming AES Convention in New York. The RealSpace Panoramic Audiovisual Camera combines arrays of both video cameras and microphones via advanced analysis techniques to capture and localize sound within a 3-D space. Using this technology, one can literally view the sound within the space on video, a capability that could revolutionize such fields as architectural acoustics and acoustical analysis, with applications in telepresence, security, and 3-D audio.
The RealSpace Panoramic AudioVisual Camera is comprised of an array of 64 microphones and five digital video cameras contained within a roughly head-sized sphere and connected to a computer. The included analysis software stitches the visual images into a 360-degree panorama, along with a visual representation of sonic landscape.
The RealSpace camera literally enables the operator to see the sound arriving at any given seat of an installation — every reflection from every wall, the sound from every speaker — with extreme accuracy and in real-time. An acoustician or designer could analyze the recording to locate aberrant reflections, noise sources or malfunctioning equipment instantly and develop a treatment plan on the spot. Even better, the VisiSonics RealSpace Panoramic AudioVisual Camera allows the client to literally see the problem, greatly facilitating solutions.
Tim Gulsrud, senior consultant at acoustical design firm Kirkegaard Associates, called the RealSpace panoramic audio camera “unique, since it allows a visual analysis of sound reflections that is not available from any other commercially available device. It would have significant application within architectural acoustics practice.”
RealSpace can also be used to record high order ambisonics with its high-quality 64-microphone array, allowing the user to produce 3-D mixes. Other applications include: noise source detection, surveillance and scene analysis.
The system’s microphone array offers 16- or 24-bit sampling at 48kHz. Included software handles all data analysis and data stream management, using standard USB and Firewire interfaces to connect to a provided graphics processor (GPU) enabled laptop. Algorithms that exploit the next generation parallel architecture of the GPU make audio and video processing computations blazingly fast. All cabling, a tripod, custom shipping case and software are provided; the system is ready to plug-in and use out-of-the-box.
Based in College Park, MD, VisiSonics was formed by researchers and faculty from the University of Maryland. VisiSonics is attending AES to launch their products and is seeking resellers, integrators, and OEMs who wish to utilize their new technology and cost-effective products. VisiSonics has posted brief video examples of the system in operation, and will offer a live demo in Booth 758 at AES.