Stereo imaging, as studied here, is called natural in depending only upon supplying to each ear exactly the same signal as would be presented naturally in hearing a real, discrete source. The specification of loudspeaker signals for such imaging must take into account the alteration of the signals as they propagate around an acoustic obstacle, the human head, to combine at the ears. Available, realistic, head-diffraction data are inadequate for the inversion of this "acoustic matrix," except for imaging at center front, center back, or center side, because of omitting phase data in free-field form. Computer generation and inversion of the acoustic matrix for a spherical-head model, adjusted to fit real-head interaural delays, as well as real-head, center-front, and center-side imaging specifications, was used. The spherical-head model indicates trends pending a;availability of real-head data. At center front, both models require departure from flat frequency response for natural imaging. Bifontal (standard stereo), trifontal, and quadrafontal imaging several cases of each, are studied. Some trifontal cases related to possible image stabilization using a "center fill" loudspeaker. Data are presented in curves plotting levels and phases versus frequency.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.