Characterizing acoustic spaces is important for multiple applications, from architectural acoustics to augmented and virtual reality, sound design, and others. Traditionally, this process involves making impulse response measurements in an unoccupied space, which is time consuming and potentially inaccurate because the acoustics of spaces may be affected significantly by the presence of an audience or other occupants. The ability to “blindly” retrieve acoustic information about a space from recordings made during typical use scenarios would be more practical and potentially it could provide a more accurate picture of the acoustics of the space when partially or fully occupied. In this paper we discuss how an analysis of the coherence properties of the harmonic partials of naturally occurring speech and music signals recorded in a space may provide information about the impulse response of the space. A broad class of naturally occurring sounds, such as speech and music, contain sets of harmonically related overtones with mutually correlated amplitude and phase modulations, i.e., vibrato. This is a reasonable assumption based upon the physics of sound generation in many realistic sources, such as wind musical instruments or the human vocal tract. When such acoustic signals are filtered by a space, we show how the autocorrelation of individual overtones and the cross-correlation between pairs of harmonically related overtones can provide information about the impulse response of the space. In this paper we explore this hypothesis by analyzing a few simplified cases and we discuss how these methods may be extended to more complex, and realistic, scenarios.
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.