No description available.

Coherence, which gives the similarity of signals received at two microphone locations, can be a powerful tool for calculating acoustic quantities, particularly active acoustic intensity. To calculate active acoustic intensity, a multi-microphone probe is often used, and therefore coherence between all microphone pairs on the probe can be obtained. The phase and amplitude gradient estimator (PAGE) method can be used to calculate intensity, and is well suited for many situations. There are limitations to this method—such as multiple sources or contaminating noise in the sound field—which can cause significant error. When there are multiple sources or contaminating noise present, the coherence between microphone pairs will be reduced. A coherence-based approach to the PAGE method, called the CPAGE method, is advantageous.Coherence is useful in phase unwrapping. For the PAGE method to be used at frequencies where the probe microphone spacing is larger than half a wavelength (above the spatial Nyquist frequency), the phase of transfer functions between microphone pairs must be unwrapped. Phase differences are limited to a 2π radian interval, so unwrapping—adding integer multiples of 2π radians to create a continuous phase relation across frequency—is necessary to allow computation of phase gradients. Using coherence in phase unwrapping can improve phase gradient calculation, which in turn leads to improved intensity calculation.Because phase unwrapping is necessary above the spatial Nyquist frequency, the PAGE method is best suited to dealing with broadband signals. For narrowband signals, which lack coherent phase information at many frequencies, the PAGE method can give erroneous intensity results. One way to improve calculation is with low-level additive broadband noise, which provides coherent phase information that can improve phase unwrapping, and thereby improve intensity calculation. There are limitations to this approach, as additive noise can have a negative impact on intensity calculation with the PAGE method. The CPAGE method, fortunately, can account for contaminating noise in some situations. A magnitude adjustment—which arises naturally from investigation of the bias errors of the PAGE method—can account for the additional pressure amplitude caused by the contaminating noise, improving pressure magnitude calculations. A phase gradient adjustment—using a coherence-weighted least squares algorithm—can likewise improve phase gradient calculations. Both adjustments depend upon probe microphone coherence values. Though not immune to contaminating noise, this method can better account for contaminating noise. Further experimental work can verify the effectiveness of the CPAGE method.