Virtual Pitch & Timpani

Ernst Terhardt developed an influential theory of virtual pitch, building on earlier work on the perception of complex tones.22 Historical discussions also include August Seebeck’s work on periodicity and J. F. Schouten’s pitch of the residue. These ideas concern how a pitch can be heard even when no spectral component is present at that frequency. Tartini’s terzo suono and Rameau’s fundamental bass belong to the wider history of thinking about sound and harmony, but combination tones, harmonic theory, and missing-fundamental perception should not be treated as identical phenomena.

Virtual pitch refers to a pitch inferred from relationships among sound components, rather than simply to the pitch of one component considered in isolation. In a missing-fundamental example, listeners can hear a pitch corresponding to a fundamental frequency even when that frequency is absent from the acoustic spectrum. By contrast, spectral pitch refers to the pitch associated with an individual spectral component. A pure tone can have a clear pitch, so numerous strong overtones are not a prerequisite for hearing pitch. A complex sound may be heard as a unified tone, as several components, or in a mixture of these ways. Which pitch is most salient can depend on the components’ frequencies, relative levels and durations, as well as the listener and the listening context. For an overview, see Andrew J. Oxenham, “Pitch Perception” (2012).

The video below demonstrates missing-fundamental pitch perception. It illustrates the perceptual principle, not a measurement of timpani clearing.

Virtual Pitch Demonstration
courtesy of

Does virtual pitch exist in timpani pitch?

A timpano does not have the complete, exactly harmonic spectrum of an ideal harmonic source. Some prominent partials can nevertheless form an approximately harmonic pattern that supports a musical pitch. The tuning task is not to force every membrane mode into one exact harmonic series. It is to obtain a useful, stable tone while recognizing the instrument’s inharmonic components. See Seebeck vs. Ohm. Listeners can differ in the weight they give to individual spectral components and to an overall pitch inferred from their relationships.23 These are listening tendencies, not mutually exclusive categories or a diagnostic test of skill. Brightness and pitch can also influence one another in judgments, but a brighter sound is not necessarily sharper, and a darker sound is not necessarily flatter. In practice, timpanists can compare the overall musical note with selected partials when evaluating and tempering a head.

The principal tone used as the timpano’s tuning reference is commonly associated with mode 1,1. This is distinct from the membrane’s lowest mode, (0,1). A possible missing-fundamental pitch an octave below the principal tone is a third, separate matter: it is not automatically the frequency of mode (0,1). Some timpanists have described an indistinct sound an octave below the principal tone when using a soft mallet at particular strike positions.24 Such a report should be treated as a listening observation, not as proof that every listener hears that octave or that the head produces no acoustic energy there. Relative partial levels can change which pitch stands out. For example, prominence of mode 2,1 can alter the tonal impression; it does not establish a universal threshold at which virtual pitch is “destroyed.” See Pleading the Fifth.

Electronic tuners estimate pitch from a microphone signal using an algorithm. Their displayed note, octave, and frequency must be distinguished from both a listener’s perceived pitch and the frequencies of individual membrane modes. A note-and-cents display alone does not measure the strength of mode (1,1), identify a modal pair, or establish harmonic alignment. An octave-low reading may reflect how an algorithm interprets a complex spectrum, including an octave-selection error. By itself, it proves neither successful tempering nor perception of a missing fundamental. For example, if the principal tone is C3, approximately 130.81 Hz with A4 = 440 Hz, a C2 display corresponds to approximately 65.41 Hz. That display does not tell us whether the recorded signal contains energy at 65.41 Hz. Establishing that requires spectral analysis; establishing what listeners hear requires a listening test.

The graphics below document an octave-low tuner reading reported for this page. A notated C3 was played on a 72 cm Fiedler timpano and the Cleartune – Chromatic Tuner for iPhone & Android displayed C2, one octave below the intended C3. The observation is retained, but the interpretation is limited to what it shows: the app selected the lower octave.

The Cleartune example shows a C2 indication for an intended C3. Without the original recording and analysis of the app’s behavior, this does not distinguish an estimate inferred from upper partials from other causes of an octave-low reading. It should not be presented as proof that the head was balanced or that a listener heard C2.

Tempering a Head

In a second example reported on this page, a different iPhone app (insTuner) was used on the same 72 cm Fiedler timpano. When the head was tempered at a low C2, insTuner displayed C1. This is another octave-low indication, not independent confirmation of a missing acoustic component, a perceived lower octave, or successful tempering.
VP-C2-C1

insTunerC1

Interpreting the Spectrum Display

The display below was generated for the original C3 example on a 72 cm Fiedler timpano using the SpectrumView Frequency Analysis iOS App on an iPad2. The original interpretation associated displayed features with modes (1,1) through (6,1) and with a possible C2 component near 65.41 Hz. Virtual pitch cannot create a line in the recorded acoustic spectrum. If a feature near 65.41 Hz is present, it needs a physical or measurement explanation, such as recorded low-frequency sound, distortion in the instrument or recording chain, background noise, or an analysis artifact. Its origin cannot be established from this screenshot alone. Nor can the absence of membrane motion at that frequency be inferred from a tuner indication. Identifying individual membrane modes requires evidence beyond the position of a peak in a single sound spectrum.

SpectrumView display from the original timpani example; interpretation discussed in the text

Using Tuners When Tempering Timpani

For a detailed tempering procedure, see Chapter 5: Tempering Timpani. For fine adjustments, see Step 6 Fine-Tuning. Use a tuner as a supplementary comparison tool, not as an automatic certificate that a head is clear. Keep the intended note and octave explicit, use comparable strike positions, mallets, dynamics, and microphone placement, and compare like with like. If the display jumps an octave, check it against a reference tone and the sound itself rather than adjusting the head simply to obtain the lower reading. Claims that an octave-low indication reliably diagnoses successful tempering require a controlled comparison with independent acoustic measurements and listening judgments.


This page distinguishes three questions: what frequencies are present in the recorded sound, what note an electronic tuner estimates, and what pitch a listener hears. These may be related, but none is an interchangeable measurement of the others. The historical examples document tuner behavior; they do not by themselves establish an octave-lower percept or validate a clearing method.

The next question is: which features of a timpani sound support a stable pitch judgment? Modern pitch research distinguishes information available from spectrally resolved harmonics from the temporal-envelope cues produced when several harmonics interact within an auditory filter. These are not simply two fixed listener types. Low-numbered resolved harmonics often support more precise pitch judgments than unresolved harmonics, but the underlying neural coding remains a research question. See Oxenham’s review. Applying this framework to timpani requires attention to the actual frequencies, levels, and decay of its partials, not an assumption that an octave-lower fundamental must be heard.

The following page considers pitch perception in relation to the preferred diametric modes and their approximately harmonic relationships. Keep three distinct issues separate: frequency relationships between modal families, splitting within a modal family, and the pitch perceived by a listener. A clearer or steadier tone does not, on its own, establish which of these has changed.

<<Previous – Next>>

Scroll to top