The sound takes another step upward. Then another. It returns to where it began, yet the next step still seems to rise. A Shepard sequence can give the impression of climbing without reaching a higher final floor. The impossibility belongs to the interpretation of a changing mixture, not to a physical frequency increasing beyond every limit.
The metaphor of a circular staircase is useful if it does not conceal the mechanism. There are several tones at once, spaced by octaves. Their strengths change across the spectrum so that the high and low edges become inconspicuous. The sequence gives local evidence of upward movement while making absolute height harder to pin down.
An original 24-step Shepard-style sequence
Listen at a comfortable low level. Before looking at the construction below, note whether the second traversal feels higher, similar, or less convincing than the first. There is no required reaction. A listener can follow one component, notice a reset, or experience a strong overall ascent. Those reports ask different questions about the same signal.
One name, several heights
A musical C can occur in many octaves. The note name identifies pitch class; the octave helps specify height. Tamara Smyth's UC San Diego explanation describes Shepard tones as octave-separated sinusoids beneath a fixed spectral envelope. As their pitch classes move, low components enter quietly and high components leave quietly. After twelve semitone steps the pattern can return to its beginning.
The quiet edges do important work. Imagine a line of people walking up an escalator while the lowest person joins behind a curtain and the highest disappears behind another. If you follow the middle group, you see continual upward movement without clearly seeing the membership change. The sound uses an analogous redistribution of components, though the auditory experience is not literally a visual tracking process.
In the sequence here, the twelve pitch-class states are repeated. The physical ingredients do not march to infinitely high frequencies. What repeats is a set of mixtures arranged so adjacent states support an ascending judgment. Repetition and perceived direction can coexist because judging a small step is not the same task as assigning an absolute position to the whole mixture.
What an experiment can ask that a demo cannot
The Praat reference entry identifies Roger Shepard's 1964 paper, Circularity in Judgments of Relative Pitch. The crucial phrase is “judgments of relative pitch.” A laboratory can present tone pairs in controlled orders and measure the pattern of higher/lower responses rather than rely on one person's impression of an endless scale.
For a simple paper model, draw twelve positions around a clock. Mark each neighboring clockwise step “up.” Following those local labels eventually returns you to the same position. The diagram does not prove the auditory illusion; it clarifies the logical shape of the result. A consistent relationship between nearby pairs need not create one globally ordered line.
Pitch circularity is not confined to one sound recipe
Diana Deutsch, Kevin Dooley and Trevor Henthorn's 2008 study produced circularity with tones containing full harmonic series, using changing relative harmonic amplitudes. Sixteen participants judged ordered tone pairs. This was a different construction from the classic octave-spaced mixture, and the study examined individual differences as well as the overall pattern.
That extension matters because it separates a perceptual question from one famous demonstration's engineering. If several stimulus families can produce related judgments, an account must explain more than one clever audio file. It also warns against treating the staircase image as a literal recipe: different physical constructions can support a similar perceptual relationship.
The original demonstration in this article uses the simpler octave-spaced approach. It does not replicate the 2008 harmonic-amplitude experiment. Naming that boundary lets you compare the methods without confusing shared vocabulary with identical stimuli.
Why the illusion is useful to a composer
A composer can exploit a directional impression without literally exhausting the instrument's register. The effect may suggest building tension or movement, but a description of the synthesis cannot dictate a listener's emotion. Loudness, harmony, rhythm, context and personal associations still participate.
The same caution makes the illusion more rewarding. It is possible to understand exactly how the file is constructed and still hear a compelling rise. Explanation does not have to erase experience. It can give you two parallel ways to follow it: the frequencies and amplitudes cycling through a bounded arrangement, and a percept that keeps interpreting the next local step as upward.
Sources and a closer look: UC San Diego: Shepard-tone construction; Praat: Shepard 1964 bibliographic reference; Deutsch, Dooley and Henthorn: harmonic-series pitch circularity.
Explore art, experiments and other unexpected connections at Incoprea.
