A psychoacoustic experiment and analysis, proposed to ascertain permissible levels of phase distortion in various audio signals, addressed the significance of the audibility of phase distortion. A valid experimental design was justified by the application of the Kwalwasser-Dykema Music Tests format. The use of a computer enhanced the accuracy in implementing the digital all-pass filter. The use of studio-quality audio equipment also aided in enhancing the accuracy of the experimental results. The statistical analysis method used in this thesis research of equalizing type 1 and 2 error provided for a fair study. Finally, permissible levels of phase distortion audibility within the confines of the experimental design were established.
These permissible levels may be beneficial in the design and application of audio equipment. It is concluded that phase distortion is an extremely subtle effect, complicated by reverberant listening or original recording conditions. The audibility of phase distortion in audio signals seems to be highly dependent upon individual ability. It is also suggested that since audible only in extreme situations, the audibility of phase distortion may be of concern only when primary factors (i.e., reverberant listening environments) are accounted for.
Although in agreement for the most part with previous research, the phase distortion audibility results in this thesis research did not seem to be as significant as demonstrated by Lipshitz et al. . Furthermore, this test does not support the results of Hansen and Madsen  since their research revealed increased phase sensitivity with a loudspeaker in reverberant environments as compared to headphone listening tests.
The selection of the all-pass center frequency of 3.5 kHz where the Robinson-Dadson curve displays the lowest threshold of audibility may suggest that for broad-band signals (such as the impulse), mid-range phase distortion is most audible. Since phase-locking of the auditory fibers is lost above 5 kHz, this may be a contributing factor for the 10 kHz sawtooth wave having negligible results in phase distortion audibility. Although phase-locking of the auditory nerve fibers is present even at low intensities, the lack of phase distortion audibility in music test signals implies this theory is suspect to explain phase distortion audibility for complex audio signals and/or another masking mechanism may be present.
Considerations regarding this research include the fact that phase distortion inherent in the transducers used for the listening tests such as the microphones, headphones, and loudspeakers, were not investigated. The amplifiers used by the headphones and loudspeakers were assumed to be phase linear. Test subjects were assumed to have normal hearing, but were not tested.
The human auditory system was found to be extremely tolerant of even gross
phase distortion effects. Although the impulse test signals were very revealing
of the presence of phase distortion, more refined research utilizing an improved
selection, broader frequency range, and various all-pass filtering implementations
(wider range of phase distortion levels and Q) of test signals are necessary
to ascertain more accurate permissible levels of audible phase distortion. Improved
irradiation methods, such as the use of phase-equalized loudspeakers in an anechoic
environment, may also aid in ascertaining more accurate permissible levels.