Assessment of desirable reflections and control of undesirable reflections in rooms are best accomplished if the reflecting surfaces are properly localized. Several measurement techniques exist to identify the incident direction of reflected sound, including the useful polar energy time curve (Polar ETC), which requires six cardioid impulse response measurements along the Cartesian axes. The purpose of this investigation is to quantify the incidence angle estimation error introduced into the Polar ETC by non-cardioid microphone directivities. The results demonstrate that errors may be minimized with a cardioid-family microphone possessing a certain range of directivities and by maximizing the measurement signal-to-noise ratio.

In room acoustics, directional measurements have the ability to more fully detail room effects than common nondirectional measurements. The extraction of directional information has been a subject of considerable interest for years and has resulted in the development of several interesting approaches to the problem. However, past methods have often suffered from inaccuracies, difficult implementation, lengthy procedures, extensive computational requirements, or the need for large amounts of equipment. A practical new method will be presented that extracts directional information of sound arrivals through a single multichannel measurement and a new cross-correlation technique. A direct comparison of numerical simulations will demonstrate that the method is more accurate than other well-established methods. Advantages and limitations of the new approach will be discussed.

No abstract

An acoustician can identify and treat problematic surfaces to reduce or eliminate unwanted reflections only if he knows the origins of those reflections. Several measurement techniques exist for the purpose of identifying these origins, including the Polar ETC method, which requires six cardioid impulse response measurements along Cartesian axes. This presentation will explore two implementations of the method using either a microphone positioner and six sequential cardioid measurements (as originally intended) or four simultaneous measurements from a tetrahedral subcardioid microphone array (originally intended for Ambisonic recordings, but also usable to synthesize the six cardioid measurements). It will compare the two approaches and investigate typical errors introduced by nonideal cardioid directivity patterns. The presentation will also discuss the capabilities of a new method, based on a Cartesian array of seven omnidirectional microphones, and explore the effects of nonideal omnidirectional patterns.