The visual sort and rate method for perceptual evaluation in listening tests.

Area of Science:

• Acoustics
• Psychoacoustics
• Human-Computer Interaction

Background:

• Perceptual rating of sound stimuli is crucial in various fields.
• Existing methods like visual analogue scales (VAS) have limitations in comparing similar stimuli and ensuring reliable rank ordering.
• There is a need for more effective methods for audio perception evaluation.

Purpose of the Study:

• To introduce and evaluate the Visual Sort and Rate (VSR) method for perceptual rating of sound stimuli.
• To compare the reliability and effectiveness of VSR against computer-based VAS (C-VAS) and paper-based VAS (P-VAS).

Main Methods:

• The study compared three methods: Visual Sort and Rate (VSR), computer-based visual analogue scale (C-VAS), and paper-based visual analogue scale (P-VAS).
• Two sets of stimuli were used: synthetic vowel /a/ with varying spectral tilts (single parameter) and naturally spoken voices (multi-parameter).
• Subjects rated stimuli using each of the three methods.

Main Results:

• The VSR method showed significantly higher reliability in single-parameter tests, evidenced by higher Pearson and Spearman correlation coefficients.
• For multi-parameter, intra-subject tests, VSR also yielded significantly higher Pearson correlation coefficients compared to P-VAS.
• VSR facilitates easier rank ordering of similar sound stimuli.

Conclusions:

• The Visual Sort and Rate (VSR) method offers superior reliability and accuracy for perceptual rating of sound stimuli compared to traditional VAS methods.
• VSR is particularly beneficial for rating similar stimuli and in multi-parameter evaluations.
• The findings suggest VSR as a valuable tool for audio perception research and applications.