Stereo geometry tests

Abstract

Since judgement of distance to a sound source is vague, whereas lateral localization is about 2°, stereo tests were conducted with a single row of sound sources. Live sounds were observed and recorded outdoors; reproduction indoors utilized speakers on a 25-foot array. To express the results a criterion of error was computed which ranged from 0.042 for live listening to 0.60 for 2-loudspeaker stereo without toe-in. Three-loudspeaker arrays with toe-in exhibited criterion-of-error values of 0.11 to 0.14 with no significant difference between 2 and 3 electrical channels.