In many sound reinforcement and reproduction scenarios, the desired audience sound coverage may only be achieved by using multiple electro-acoustic transducers emitting coherent signals at equal or nearly equal sound power levels. When transducers are not part of an acoustically coupled array, any difference in path-length from a listening position to two or more loudspeakers will result in a relative phase difference between the received signals. The summation of such signals will result in a frequency response that is dependent on the path-length difference, and there is cancellation of frequencies where phase difference equates to 180 degrees. High interchannel correlation can also lead to lack of apparent source width in multichannel reproduction and a lack of externalization with headphone reproduction. This work examines a time-variant, real-time decorrelation algorithm for the reduction of interchannel correlation that is also capable of reducing correlation between direct sound and early reflections. The focus is on minimizing wide-area low-frequency magnitude response variation in sound reinforcement scenarios but is applicable to a wide range of sound reproduction applications. Key variables that control the balance between decorrelation and processing artifacts, such as transient smearing, are described and evaluated using a MUSHRA test. Parameter values that render the processing transparent while still providing decorrelation are discussed. Additionally, the benefit of transient preservation is investigated and shown to increase transparency while not significantly degrading performance.
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.