Answers are grounded in this paper and cite their sources.
Listening experiment examines differences in chord perception, indicating implications for musical composition.
Not all perceived (experienced) pitches in a chord correspond to notes. A simple masking model predicts that a typical C major triad has audible partials at pitch classes B and D (third harmonics of E and G), regardless of voicing; an A minor triad, at G, B, and D. In a listening experiment, two successive chords were heard in each trial. The chords were identical except that one partial was removed from one of them. Participants rated how much the chords differed. We tested major and minor triads in root position and inversion, augmented and diminished triads, other common trichords ([025]/CDF, [035]/DFG), and one tetrachord (major-minor 7th). All chords were in close position on C4. To simplify the stimuli and allow for testing of a simple model, chord partials had equal physical amplitude, were tuned to 12-tone equal temperament, and were included only if predicted by a simple psychoacoustic model to be audible (perceptually relevant). We predicted and demonstrated empirically that common musical trichords have 10 to 15 audible partials, some of whose pitch classes deviate from chord tones. Results imply that notes commonly added to trichords to form tetrachords (e.g. 7ths and 9ths) often correspond to spectral or virtual pitches in the original trichord. Commonly added tones also minimally increase dissonance. Results did not correlate with predicted spectral or virtual pitch salience according to Ernst Terhardt’s pitch algorithm, perhaps because it neglects complex temporal effects.
No takes yet. Share an insight, caveat, or question.
Parncutt et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: