Opens in a new tab

AES E-Library

← Back to search

Convention Paper

Elicitation of Descriptors for Comparing Binaural and Multichannel Speaker Playback

Authors: Ikeda, Tsubasa; Kamekawa, Toru; Marui, Atsushi

Convention Paper · Paper 10238 · October 2025

Abstract

Binaural and multichannel loudspeaker playback were compared using a controlled listening experiment based on Projective Mapping (Napping). Four playback conditions (27.0ch and three downmixes) and three program types (organ, band, and field) were tested. From 11 trained participants, we elicited 93 descriptive terms that were consolidated into four main categories: Spatial (e.g., envelopment, reverb, localization), Frequency (e.g., much low-frequency, less high-frequency), Timbre (e.g., clear, bright), and Intensity (e.g., powerful, piercing). The intensity category, in particular, has rarely been discussed in prior comparisons. MFA and HCPC showed that organ and band materials produced mixed clusters across playback conditions, whereas the field recording separated clearly along the binaural-loudspeaker axis. UFP linked these structures to descriptor vectors, with organ and band driven mainly by spatial, timbral, and intensity terms, while the field recording aligned with frequency and partly intensity terms. An auxiliary acoustic analysis (event density, short-term loudness variance, spectral centroid, spectral flux) was consistent with these perceptual structures. Overall, both playback mode and program material shape listeners impressions, and "intensity" emerges as a useful descriptor family when comparing binaural with loudspeakers.

Details

AES Convention
159
Paper number
10238
Publication date
October 14, 2025
Affiliation
Tokyo Univercity of the arts; Tokyo Univercity of the arts; Tokyo Univercity of the arts (See document for exact affiliation information.)
Type
Convention Paper