The Problem the Ear Solves ↑ menu
When a timpano vibrates, it produces a spectrum of partials — not all of which are equally useful to the ear. The auditory system’s ability to extract pitch from a complex tone depends critically on whether individual harmonics are resolved by the cochlea. Harmonics below about the 10th are typically resolvable as distinct spectral components; above that threshold, the auditory filters become too broad and adjacent harmonics fall within the same critical band, forcing the ear to rely on temporal cues instead.[1]
This distinction matters enormously for the timpanist. When a timpano is tuned to a given note — say, C3 (130.8 Hz) — the principal vibrational mode of the membrane, the (1,1) mode, does not correspond to the first position in a harmonic series. It corresponds to the second harmonic. The true fundamental of the implied harmonic series is an octave below: C2 at 65.4 Hz. That frequency is absent from the drum’s spectrum. It must be supplied by the brain.
This is the missing fundamental — not an incidental quirk of timpani acoustics, but a central and structurally necessary feature of how these instruments produce pitch.
The Duff/Benade Mode Ratios and Their Harmonic Correspondence ↑ menu
In 1973, Arthur Benade measured the frequency ratios of the vibrational modes of a timpano belonging to Cloyd Duff, principal timpanist of the Cleveland Orchestra, after the head had been cleared to Duff’s exacting standard. The ratios he reported for the first six preferred diametric modes, with (1,1) as the reference (1.0×), are:[2]
| Mode | Duff Ratio | Nearest Integer Harmonic | Deviation from Integer |
|---|---|---|---|
| (1,1) | 1.000 | 2nd | — (reference) |
| (2,1) | 1.504 | 3rd | +0.7% |
| (3,1) | 2.000 | 4th | 0.0% |
| (4,1) | 2.494 | 5th | −1.2% |
| (5,1) | 2.979 | 6th | −0.7% |
| (6,1) | 3.462 | 7th | −1.1% |
The (1,1) mode at 1.0× is the second harmonic of the implied missing fundamental. Modes (2,1) through (6,1) occupy harmonic positions 3 through 7. The near-perfect alignment of (3,1) to the 4th harmonic (exactly 2.0×) is particularly striking — an octave above the (1,1) mode, as close to perfectly harmonic as any real membrane can achieve.
All six preferred diametric modes fall within the seventh harmonic of the implied fundamental. This is not merely convenient — it is the acoustic reason the timpano can produce a clear, stable pitch at all.
Why No (0,1) Mode? The Question of the Missing Fundamental ↑ menu
In a theoretical ideal membrane, the (0,1) mode — a circular node with no diametrical nodes — would be the lowest in frequency and would correspond to the first harmonic (the fundamental). In practice, the geometry of the kettle and the constraints of the suspension system mean the (0,1) mode does not behave as a useful pitch carrier on a timpano. The (1,1) mode, with one diametrical node line, is the lowest usable mode — and it sits at the second harmonic position.
The ear’s pitch extraction system, however, has no difficulty with this arrangement. When presented with a series of low-numbered harmonics — harmonics 2 through 7 of an implied fundamental — the auditory system readily infers the missing fundamental an octave below the (1,1) mode. This is the same mechanism responsible for the missing fundamental effect in telephone speech and in Shepard tones. The brain reconstructs the octave below from the harmonic relationships of the partials that are actually present.
This point is worth dwelling on: the timpano does not produce a physical fundamental. It produces a spectrum that occupies harmonic positions 2 through 7 of an implied fundamental. The ear supplies the missing octave. In this sense, the timpano is not merely an instrument that exhibits the missing fundamental effect — it is architecturally dependent on it.
Resolved Harmonics and Pitch Identification ↑ menu
A 2025 study by Albera and colleagues at the University of Turin investigated how accurately listeners can identify musical pitch from harmonic complexes containing only two or four consecutive harmonics, with no physical fundamental present. Using synthetic tones presented to 30 amateur musicians with normal hearing, they found that correct identification rates ranged from 8% to 100%, depending on which harmonics were presented. Crucially, identification was most reliable when the harmonics were low-numbered and close to the fundamental — correct identification reached 88% to 100% with harmonics 2 through 5, and 82% to 96% with harmonics 3 through 6. Identification also dropped sharply when the highest presented harmonic exceeded roughly 1500 to 2000 Hz.[3]
This finding reinforces a foundational principle in psychoacoustics: pitch salience from harmonic complexes drops dramatically when harmonics below the 10th are removed.[4] The auditory system has fewer reliable spectral cues to work with. For the timpanist, this means that the presence and relative strength of the preferred diametric modes — all of them falling within the seventh harmonic — is not merely a matter of timbre. It is a direct determinant of whether the drum produces a clear, stable pitch percept.
Two Modes of Pitch Perception: Spectral and Virtual ↑ menu
Contemporary psychoacoustic research distinguishes between two perceptual strategies for pitch extraction from complex tones:
Spectral pitch (sometimes called analytical listening) arises when the ear resolves individual harmonic components and matches them to spectral templates. Each resolved harmonic contributes its own pitchlike sensation, and the perceived pitch results from a weighting or higher-level inference across the resolved partials.[5]
Virtual pitch (sometimes called synthetic listening or periodicity pitch) arises when the ear detects the repetition rate of the waveform — even in the absence of a physical component at that frequency — and infers a fundamental pitch. This is the modern descendant of Seebeck’s periodicity pitch.[6]
A line of research from the Heidelberg group, including Schneider, Seither-Preisler, and colleagues, has shown that individual listeners exhibit stable, measurable preferences for one strategy or the other, and that this trait is reflected in distinct anatomical and cortical patterns. Listeners who favor spectral (overtone-based) listening tend to show rightward hemispheric dominance and stronger responses in right-hemisphere auditory regions; listeners who favor virtual pitch (fundamental-based) tend to show leftward dominance and stronger early responses in left-hemisphere regions. These preferences appear to be stable traits established early in development and are not strongly predicted by musical training or expertise.[7]
A 2025 MEG study by Saus, Seither-Preisler, and Schneider added a finer-grained oscillatory dimension to this picture. The two listening modes appear to differ in their characteristic cortical oscillations: spectral listeners tend to show elevated theta-band power and reduced gamma-band power, while fundamental listeners show the opposite pattern. The two effects (anatomical lateralization and oscillatory signature) appear to be related but distinct markers of the same underlying perceptual tendency.[7]
What This Means for the Preferred Diametric Modes ↑ menu
The Duff/Benade data give us a precise picture of why the first six diametric modes are perceptually preferred:
All six are low-numbered harmonics. Modes (1,1) through (6,1) occupy harmonic positions 2 through 7 of the implied missing fundamental. All fall comfortably within the ~10th-harmonic resolution boundary, meaning each can be individually resolved by the cochlea. There are no high-numbered, poorly resolved modes competing in this range.
The (3,1) mode is perfectly harmonic. At exactly 2.0× the (1,1) frequency, (3,1) is an octave above the principal. This single coincidence does more perceptual work than any other: it gives the ear a clear, unambiguous octave relationship within the spectrum itself, anchoring the virtual pitch percept.
The (2,1) and (4,1) modes are nearly perfect 3rd and 5th harmonics. At 1.504× and 2.494× respectively, they correspond to ratios of 3/2 (a perfect fifth) and 5/4 (a major third above the octave) in the underlying harmonic series. These are the most consonant intervals in music — and they are present in the timpano’s spectrum as structural features of the Duff-cleared drum.
The (5,1) and (6,1) modes extend the series through the 6th and 7th harmonics, adding further spectral coherence to the pitch percept while remaining within the resolvable range.
This harmonic architecture — modes at the 2nd through 7th harmonics of an implied missing fundamental — is not accidental. It is the acoustic result of the Duff clearing process. Duff’s technique of precise subsidiary tension adjustment persuades the partials to align with these near-integer ratios. Benade observed that on Duff’s properly cleared drum, “the frequency alignment of the principal tone, fifth, octave, and tenth” — Modes (1,1), (2,1), (3,1), and (4,1) — was “remarkably close to harmonic.”[2] When the head is cleared, the timpano’s spectrum approximates the harmonic series that the ear is wired to expect.
The Duff Clearing Process as Perceptual Engineering ↑ menu
Duff described the process as “clearing” — persuading the partials to more closely match the ideal. Benade, quoting Duff directly, explained it this way: “It is not sufficient merely to get the overall skin tension correct for the desired pitch of the kettledrum, one must also make small additional changes in the tension produced by the various screws around the periphery of the drum.” These small adjustments compensate for the inherent irregularity of the membrane and for eccentricity in the kettle rim. When everything is in perfect adjustment, the drum “is said to have been cleared.” Benade described hearing Duff clear a well-tuned drum as “a revelation” — the tone “ringing with smoothness and clarity.”[2]
What Duff was doing, in acoustic terms, was maximizing the harmonic coherence of the timpano’s spectrum. He was bringing the (2,1), (3,1), (4,1), (5,1), and (6,1) modes into precise alignment with the harmonic positions 3 through 7 of the implied missing fundamental — so that the auditory system’s pitch extraction mechanisms encounter exactly the spectral structure they are designed to process.
This explains why a well-cleared timpano produces a pitch that is perceptually as clear and stable as that of any string or wind instrument — despite being built on a missing fundamental. The ear receives a nearly perfect harmonic series from the 2nd through 7th harmonics and reconstructs the missing octave below (the 1st harmonic) automatically and effortlessly.
The Individual Listener Factor ↑ menu
Perhaps the most significant finding from recent psychoacoustic research is that pitch perception strategy — spectral versus virtual — varies across individuals and appears to be a stable trait. A 2026 MEG study by Andermann and colleagues found that listeners with stronger f0-based pitch processing (virtual pitch preference) showed measurably worse behavioral performance on pitch discrimination tasks when the fundamental was absent. Critically, this individual difference was not predicted by musical training or expertise.[8]
This has a practical implication: two listeners in the same hall hearing the same drum may be extracting pitch in genuinely different ways. One may be hearing the virtual pitch implied by the harmonic pattern; another with a stronger spectral processing tendency may be tracking the resolved partials individually. Both arrive at a pitch percept, but via different auditory routes. This explains why players sometimes disagree about whether a particular drum is “in tune” — and why a drum that sounds perfectly in tune to one ear may feel slightly uncertain to another.
Summary ↑ menu
The preferred diametric modes of the timpano — (1,1), (2,1), (3,1), (4,1), (5,1), and (6,1) — occupy harmonic positions 2 through 7 of an implied missing fundamental. This is not a coincidence. It is the acoustic goal of the Duff clearing process, which aligns the membrane’s vibrational modes with the near-integer ratios the ear requires for reliable pitch extraction. All six modes fall within the ~10th harmonic resolution threshold, meaning each is individually resolvable by the cochlea. The (3,1) mode provides a perfectly harmonic octave above the (1,1) mode; the (2,1) and (4,1) modes provide near-perfect fifth and major-third relationships within the harmonic series. The result is a spectrum that the auditory system processes with the same ease it brings to string and wind instruments — despite the structural missing fundamental at its foundation. Modern psychoacoustic research on resolved harmonics, individual pitch processing strategies, and the neural correlates of spectral versus virtual pitch perception has confirmed what Duff and Benade demonstrated empirically decades ago: timpani pitch is not approximate or second-best to the ear. It is a precisely engineered harmonic spectrum, waiting to be heard.
References ↑ menu
- [1] Resolution threshold for low-numbered harmonics (cochlear filter bandwidth and critical-band arguments). See ref. [4] for the primary modern statement.
- [2] Benade, A. H. (1973). Measurements on Duff’s timpano, reported in “Defining a Standard,” The Quantum World of Timpani Pitch, Pauken, 2021. https://quantum.pauken.org/defining-a-standard/. The primary measurement record is also reproduced in WTT Chapter 3 (Missing Fundamental): https://wtt.pauken.org/chapter-3/missing-fundamental. The Duff “It is not sufficient merely to get the overall skin tension…” and “a revelation to listen to an expert such as Duff clearing a good drum” passages are Benade’s paraphrases of Duff’s teaching, captured in Benade’s published account of the 1973 measurement visit.
- [3] Albera, R., Urbanelli, A., Lucisano, S., Aprigliano, A., Morando, L., Amoroso, A., Alexeev, M., & Albera, A. (2025). “Musical note recognition based on the upper adjacent harmonics without the presence of the fundamental frequency.” PMC, 2025. PMCID: PMC12022334; PMID: 40274804. https://pmc.ncbi.nlm.nih.gov/articles/PMC12022334/
- [4] Preisler, A. (1993). “The influence of spectral composition of complex tones and of musical experience on the perceptibility of virtual pitch.” Perception & Psychophysics, 54, 589–603. Also discussed in Schneider et al. (2005) and in Saus et al. (2025).
- [5] Terhardt, E. (1974). “Pitch, consonance, and harmony.” Journal of the Acoustical Society of America, 55(5), 1061–1069. https://doi.org/10.1121/1.1914648
- [6] Seebeck, A. (1841). “Beobachtungen über einige Bedingungen der Entstehung von Tönen” [Observations on some conditions of the generation of tones]. Annalen der Physik und Chemie, 2nd ser., liii, 423–436. The dispute with Ohm over periodicity pitch versus spectral pitch ran in Annalen from 1841 to 1863 and was ultimately resolved in favor of Seebeck by later experiments. For modern summary, see WTT Chapter 4-2: https://wtt.pauken.org/chapter-4-2/chapter-4
- [7] Saus, W., Seither-Preisler, A., & Schneider, P. (2025). “Harmonic vowels and neural dynamics: MEG evidence for auditory resonance integration in singing.” Frontiers in Neuroscience, 19, 1625403. https://doi.org/10.3389/fnins.2025.1625403. The individual-difference findings about fundamental-versus-overtone listening modes in this paper build on earlier work by the same Heidelberg group: Schneider et al. (2005), Seither-Preisler et al. (2007), and Schneider et al. (2023). Note: this paper’s primary topic is vowel resonance perception in singing; the spectral-versus-virtual-pitch findings appear as a secondary observation about individual differences in pitch processing mode.
- [8] Andermann, M., Reineke, A. L., Riedel, H., & Rupp, A. (2026). “The role of (missing) fundamentals, active listening, and musical expertise in cortical and subcortical correlates of consonance/dissonance.” European Journal of Neuroscience, 63(6), e70483. https://doi.org/10.1111/ejn.70483
Editor’s note on this revision (click to expand)
This is the 2026-06-24 citation-corrected version of the 2026-04-06 article. The substantive argument is unchanged. The following citation accuracy problems in the prior version have been corrected:
- Reference 1 (was “McPherson et al. 2025”) is actually Albera, Urbanelli, Lucisano, Aprigliano, Morando, Amoroso, Alexeev, and Albera (2025) at the University of Turin. McPherson is a real pitch perception researcher but is not on this paper.
- Reference 3 (which cited Reference 1) is updated accordingly.
- Reference 4 (Preisler 1993) is now cited directly rather than as “cited in Heinrich 2025.”
- Reference 7 (was “Heinrich, Reineke, Rupp, Andermann 2025”) is actually Saus, Seither-Preisler, and Schneider (2025). The prior citation also mischaracterized the paper’s primary topic (vowel resonance in singing) as being about pitch perception strategy. The current version notes both points explicitly.
- The in-text description of cortical oscillatory findings in Section 5 has been corrected. The prior version had the gamma/theta trends backwards.
The Benade 1973 reference (2) remains as cited, with the source chain (Defining a Standard, 2021) noted. The Terhardt 1974 (5), Seebeck 1841 (6), and Andermann 2026 (8) references are verified and unchanged.
<<Previous — Pitch Perception and the Auditory System — Next>>