VisiSonics’ 5/64 Audio/Visual Camera consists of 5 video cameras and 64 microphones mounted on the surface of an aluminum sphere. The spherical architecture and associated algorithms allow the creation of an omni-directional acoustic image of the environment. The video camera array generates an overlapping omni-directional panoramic visual image of the same environment. These images are synchronized and co-registered using patented algorithms to produce a live audio-visual stream. The convergence of audio and video capture allows for the visualization of sound sources and their interaction with the environment to be monitored in real-time and recorded for detailed analysis.

With this product, users can capture, store, retrieve and process real-time integrated and synchronized audio and visual information. The captured data incorporates all aspects of the audiovisual world that an individual would experience if he or she were physically present where the Audio/Visual Camera were located. The processing software (FINDr) and storage system allows you to do this in real-time or on stored data.

It can be used it for:

  • Noise Source Identification
  • Ambisonics
  • Architectural Acoustics and Room Measurement
  • Telepresence
  • Cross-Modal Attention and Saliency
  • Psychophysics
  • Audio-Visual Tracking
  • Audio/Visual Recording and Analysis
  • Meeting Recording
  • Intelligibility
  • Noise measurement

Access to the raw data and our growing software development kit, makes it easy for you to imagine your application!

Microphone arrays consist of multiple microphones that are integrated together to capture sound. They usually have a common clock, and the sound can be processed to obtain spatial information about the sound, essentially, as an acoustic antenna. Using sound propagation principles, the principal sound sources in an environment can be spatially located (via a process called beamforming) and distinguished from each other. Beamforming is achieved by filtering and combining the individual microphone signals.

Our patent-pending microphone architecture achieves digitization at the microphone. The digitization specifications are software controlled and can be performed at a sample rate of 44.1 kHz, with 24 bit accuracy and a dynamic range of 105dB. All microphones are tested and calibrated to provide accurate measurement data with a raw data output SNR = 62dB. The gains of each microphone are matched to within 0.1 dB SPL.

Five video cameras and Sixty-four microphone generate a lot of data. The RealSpace 5/64 audio camera generates about 82MB/second. Recording times on the provided machines are user settable and you can record up to 630 seconds. Custom orders with additional recording space are available on request.

The standard configuration is shipped with 64 microphones. Our architecture allows for an arbitrary number of microphones up to 256. Custom orders are accepted. More microphones provide better sound source isolation.

We currently ship the Audio/Visual Camera in two configurations with 5 HD cameras or 15 HD cameras. Each camera generates a 1328×1048 video image used to generate our real-time high-definition stitched spherical panoramic display. Much as maps of the spherical earth are shown as a rectangular map or on a globe, our video can be displayed as a rectangular image (in Mercator projection) or as a globe (spherical scene projection).