Efficacy of a new spatial ear training program for “Ensemble width” and “Individual source width”

The authors have devised a matching-based training paradigm that assists listeners in evaluating spatial width of a reproduced sound field with increased sensitivity and memory. The program uses a learning paradigm of repeated comparison and match to references. All sound sources were horizontally spread using a pairwise constant power panning law over front five loudspeakers located at + -60, + -30, and 0 degree. Trainees were asked to adjust a parameter controlling the panning intervals of five-channel sound sources until it matched to the perceived width of a given reference. The program provided visual feedback for isomorphic mapping. To make the training equipment more accessible, the headphone version of the program was also developed using a simulation of listener and sound sources positions in a room. Thirty participants were divided into three groups: a group with loudspeaker-based training, a group with headphone-based training, and a group without training. Participants in the trained groups took 6 days training program with 3 to 5 sessions in a day. The difference between reported and presented intervals before and after the training were measured. The analyzed results showed that regardless of training method (loudspeaker- or headphone-based), trained groups showed improvement in their matching precision.