Some researchers have indicated that substantial improvements in the sonar audio detection threshold may be achieved by a technique referred to as phase‐altered audio. The technique described in the literature, sometimes referred to as the masking level difference phenomenon, involves phase shifting the signal 180° and presenting this to one ear while the unphase shifted signal is presented to the other ear. The noise background is presented in phase at each ear. Alternatively, the noise to one ear may be phase shifted (180°) and the signal presented in phase at each ear. The degree of improvement is a function of the frequency of the signal and the degree of noise correlation. While it is clear from the literature that large improvements in audio detection threshold may be achieved under laboratory conditions where the signal and noise can be independently controlled, it is not clear how this phenomenon can be utilized to improve the audio presentation to sonar operators. A series of tests with a cylindrical array sonar was conducted at a US Navy Ashore Training Facility and a contractor test facility to determine if a phase‐altered approach offers any substantial improvements to sonar operators on a passive sonar. The tests showed no significant benefit from the masking level difference phenomena for any phase‐altered audio mode in the normal operating range of passive sonar half‐beam noise correlation. But there was a significant improvement insum(right half‐beam plus left half‐beam) beam audio performance for high negative half‐beam noise correlation and withdifference(right half‐beam minus left half‐beam) beam audio performance for high positive half‐beam noise correlation. Any improvement in performance at high half‐beam noise correlations was due primarily to thesumordifferencebeam gain and not the masking level difference phenomena. [Work supported by NAVSEA.]