The Biophysics of Musical Harmony: Plomp-Levelt Sensory Consonance & Timbre Landscapes
Welcome to the Interactive Dissonance Viewer, a state-of-the-art computational psychoacoustics laboratory engineered to simulate sensory consonance and dissonance landscapes. Based on the foundational empirical models developed by Reinier Plomp and Willem Levelt (1965) and extended into generalized spectral theory by William Sethares, this application provides an interactive window into why traditional musical scales emerged and how the physical overtone structures (timbres) of musical instruments govern auditory roughness within the human cochlea.
Overview: Cochlear Micromechanics and Critical Bandwidth
When sound waves strike the tympanic membrane, vibrations are transferred across the ossicular chain of the middle ear to the fluid-filled cochlea. Inside the cochlea, the basilar membrane behaves as a biological spectral frequency analyzer. High-frequency tones produce maximal displacement at the narrow, stiff base near the oval window, while lower-frequency tones travel to the flexible apex. Sensory hair cells situated along the Organ of Corti translate these localized mechanical displacements into neural action potentials dispatched along the auditory nerve.
In psychoacoustics, when two pure sinusoidal frequencies ($f_1$ and $f_2$) enter the ear concurrently, their response envelopes on the basilar membrane overlap if the frequency separation $|f_2 - f_1|$ is small. If the difference is below approximately 12–15 Hz, human listeners perceive smooth, periodic loudness modulations known as acoustic beats:
$$f_{\text{beat}} = |f_2 - f_1|$$
However, when the frequency difference exceeds the beating threshold but remains within the ear's critical bandwidth ($CBW$), the overlapping basilar membrane excitation causes rapid, discordant neural firing. The brain interprets this sensory clash as harshness or sensory roughness (dissonance). As the frequency separation widens beyond the critical band, the two excitation patterns decouple into distinct spatial neural channels, eliminating roughness and producing sensory consonance.
Mathematical Derivation of the Plomp-Levelt Metric
The sensory dissonance between two pure sinusoidal partials at frequencies $f_1$ and $f_2$ ($f_1 \le f_2$) with normalized amplitudes $a_1$ and $a_2$ is governed by the Plomp-Levelt empirical transfer function scaled by the Sethares factor $C \approx 5.75$:
$$d(f_1, f_2) = C \cdot \left( e^{-b_1 \cdot s \cdot (f_2 - f_1)} - e^{-b_2 \cdot s \cdot (f_2 - f_1)} \right)$$
Where the physiological constants derived from human auditory masking tests are $b_1 = 3.51$ and $b_2 = 5.75$. The parameter $s$ is an inverse critical-bandwidth scaling factor that normalizes frequency differences relative to the auditory filter bandwidth at the lower frequency $f_1$:
$$s = \frac{x^*}{CBW(f_1)} = \frac{0.24}{0.021 \cdot f_1 + 19}$$
Here, $x^* = 0.24$ represents the critical point of maximal sensory roughness as a fraction of critical bandwidth. The denominator $CBW(f_1) = 0.021 \cdot f_1 + 19$ models the empirical widening of human cochlear critical filters across the audible spectrum (approximately 25–30 Hz wide at low frequencies and expanding to several hundred Hertz above 1 kHz).
For complex instruments comprising multiple partials $S_1 = \{(f_{1,i}, a_{1,i})\}_{i=1}^M$ and $S_2 = \{(f_{2,j}, a_{2,j})\}_{j=1}^N$, total sensory dissonance $D_{\text{total}}$ is obtained by evaluating pairwise roughness across all intersecting overtone combinations:
$$D_{\text{total}}(S_1, S_2) = \sum_{i=1}^{M} \sum_{j=1}^{N} \min(a_{1,i}, a_{2,j}) \cdot d(f_{1,i}, f_{2,j})$$
Using the amplitude weighting $\min(a_{1,i}, a_{2,j})$ (introduced by Sethares) accounts for the acoustic principle that the amplitude of beating interference between two partials is constrained by the quieter component.
Consonance Valleys and Why Musical Scales Exist
When an instrument generates an integer harmonic series—such as a bowed violin string or an open flute pipe where $f_n = n \cdot f_1$ ($n = 1, 2, 3, 4, \dots$)—the total dissonance curve displays sharp, localized minima termed consonance valleys. At simple integer frequency ratios, partials from the two notes fall onto identical frequencies:
- Octave (2:1): Partials of the second tone ($2f_1, 4f_1, 6f_1, \dots$) align with even harmonics of the fundamental, leaving zero unaligned partials inside adjacent critical bands.
- Perfect Fifth (3:2): Every third harmonic of the lower tone locks with every second harmonic of the upper tone ($3f_1 = 2f_2$, $6f_1 = 4f_2$), forming a deep consonance trough.
- Perfect Fourth (4:3) & Major Third (5:4): Partials lock at $4:3$ and $5:4$, producing distinct localized valleys.
- Semitone Roughness Peak (~16:15 or 1.06): At a minor second, almost every overtone pair falls directly into the maximal roughness zone ($0.24 \cdot CBW$), triggering a massive spike in dissonance.
Crucially, as demonstrated by William Sethares in Tuning, Timbre, Spectrum, Scale, consonance is not an intrinsic property of frequency ratios alone; rather, it is an interaction between the instrument's timbre spectrum and the scale. When you select an inharmonic instrument like a bell ($1, 1.19, 1.56, 2.76, \dots$) or a circular drum membrane, the consonance valleys shift completely away from Western 12-tone equal temperament! Non-Western musical traditions (such as Indonesian Gamelan ensembles with metallic slendro and pelog metallophones) naturally evolved scales aligned with their instruments' unique physical inharmonic partials.
Extension to Triads (3-Tone Chords)
When three complex tones are sounded together ($S_1, S_2, S_3$), the auditory system processes acoustic interference across all three mutual pairs:
$$D_{\text{triad}}(S_1, S_2, S_3) = D_{\text{total}}(S_1, S_2) + D_{\text{total}}(S_1, S_3) + D_{\text{total}}(S_2, S_3)$$
In our Triad Evaluation Mode, this two-dimensional dissonance field is mapped into an interactive chromatic surface where the horizontal axis represents $f_2/f_1$ and the vertical axis represents $f_3/f_1$. Major triads ($4:5:6 \rightarrow 1 : 1.25 : 1.50$) and minor triads ($10:12:15 \rightarrow 1 : 1.20 : 1.50$) manifest as deep indigo and cyan consonance basins, while tone clusters appear as fiery orange and yellow ridges.
Real-Time Parametric Audio Synthesis Engine
The interactive audio engine utilizes the Web Audio API with a persistent, pre-warmed voice pool. Instead of creating and destroying temporary oscillator nodes—which introduces garbage collection latency and audible popping—the engine maintains active sine oscillators whose frequencies and gains are updated instantaneously via exponential parameter ramps:
$$\text{gain.setTargetAtTime}(g_{\text{target}}, t_{\text{now}}, 0.025), \quad \text{freq.setTargetAtTime}(f_{\text{target}}, t_{\text{now}}, 0.025)$$
Furthermore, an integrated 2nd-order lowpass Butterworth filter (~1,400 Hz) eliminates high-frequency aliasing and harsh buzz, providing an acoustic listening experience that matches the visual curve.
Explore Related Interactive Laboratories on BioniChaos
⚖️ Legal Notices, Nominative Fair Use & Professional Disclaimers
• Editorial Independence & Non-Affiliation: This software laboratory and psychoacoustic visualization is an independent educational and mathematical project developed by BioniChaos. It is not affiliated with, maintained by, sponsored by, or endorsed by any commercial entity, academic publisher, or instrument manufacturer referenced herein.
• Trademark Attribution: "Plomp-Levelt", "Sethares Consonance Metric", and associated terms are referenced strictly for nominative identification, scientific attribution, technological critique, and fair educational commentary under applicable international intellectual property laws.
• Not Audiological or Clinical Diagnostic Advice: All algorithms, spectral curves, auditory bandwidths, and synthesized sound frequencies are mathematical approximations intended exclusively for research, training, and conceptual illustration. This visualizer does not provide medical, audiological, surgical, or certified acoustic engineering advice and must not be used for clinical diagnosis, hearing impairment assessments, or hearing aid tuning.
• Computational Modeling Disclaimer: Cochlear bandwidths and roughness curves represent standardized statistical approximations of human auditory perception and do not account for individual sensorineural pathology or cognitive conditioning.
Open Access License: This interactive educational module is released under
CC BY-NC 4.0 (Attribution-NonCommercial)
for non-commercial research, academic study, and psychoacoustic education.
Commercial & Enterprise Licensing: For white-labeling, proprietary LMS/course embedding, hardware audio DSP integration, or custom feature engineering, secure a commercial license at
BioniCloud.com or contact
Dr. Yuri Beno.