
Recipe — octave stacks with fading edges
Start with a complex tone containing partials at f, 2f, 4f, 8f… each weighted by a bell-shaped spectral envelope centred mid-audible range. Raise all partial frequencies together; when highs exit audible window, new lows enter — pitch height increases perceptually while absolute frequency range stays bounded. Shepard's original paper presented the auditory equivalent of Penrose stairs.[1]
Risset's computer music implementations made the illusion mainstream in composition and sound design.[3] Deutsch catalogued it among pitch illusions proving central constructive processing.[4]
Why the brain is fooled
Pitch perception integrates spectral height (which partials are loudest) and chroma (pitch class modulo octave). Shepard tones decouple them — chroma cycles while height trends. Terhardt's virtual pitch theory predicts ambiguity when multiple octave-spaced templates compete.[5]
Griffiths and Warren linked planum temporale to hierarchical pitch processing — endless glide produces sustained prediction error without resolution, engaging attention.[6] Huron's expectation framework: infinite rise violates learned melodic boundaries — awe or unease.[7]

Ascending versus descending variants
Direction depends on envelope motion and whether frequencies glide up or down. Films use ascending Shepard tones for tension (dolly zoom sonic cousin); descending versions feel sinking or relief depending on context. Risset rhythms can make tempo seem to accelerate forever — separate but related paradox family.[3]
Relation to missing fundamental and harmonicity
Shepard tones rely on harmonic octave spacing — Norman-Haignere et al.'s cortical harmonicity detectors respond strongly.[8] Missing fundamental article cousin: both show pitch is inference, not raw frequency readout.[9] Inharmonic mistuning destroys Shepard effect — integer ratios essential.

Physiological responses — not everyone enjoys it
Endless rise can induce mild vertigo or anxiety — vestibular and auditory self-motion pathways crosstalk in some listeners (Todd et al.).[10] Sokolov orienting never receives closure — sustained novelty without habituation per Rankin if motion is smooth enough to avoid seams yet never resolves.[11]
Sound Bubbles sleep presets exclude Shepard motion; focus presets may use imperceptibly slow pitch drift on non-harmonic beds instead.
Music, film, and game design
Batman Nolan-era soundtracks popularised rising Shepard tones before drops. Video games use infinite fall for hypnosis aesthetics. McDermott's scene analysis warns such stimuli monopolize attention — poor background for work or rest.[12]
Banbury et al. documented cognitive costs when audio demands processing — illusions are not neutral wallpaper.[13]
Predictive coding interpretation
Friston's free-energy principle: brain minimizes prediction error — Shepard tone refuses stable minimum, keeping precision high.[14] Clark's predictive processing: endless updraft means no successful model — listener stays alert.[15] Companion sound for sleep needs the opposite: satisfiable predictions.
Sound Bubbles design stance
- No Shepard ramps in Sleep or Calm gardens — orienting and unresolved pitch.[11]
- TimeLine Drift on broadband noise adjusts level and filter, not octave-chasing partials.[11]
- Experimental Orbit modes may use subtle harmonic motion for curious listeners — labelled clearly.[1]
- Respect nausea/vertigo reports — disable motion.[10]
Build your own — cautionary steps
- Generate octave harmonic stacks with smooth spectral envelope — avoid clipping partials.[1]
- Glide frequencies slowly; fast glide feels siren, not staircase.[3]
- Test duration under two minutes first — monitor discomfort.[10]
- Never hide Shepard tone in "relaxation" products without informed consent.[20]
Historical context
Shepard's work sat alongside visual object permanence experiments — cross-modal theme that perception constructs stability from ambiguous data. Deutsch's illusions volume places Shepard tone in tradition from Tritone paradox to scale ambiguity.[4] Cognitive science tourism, but also serious window into pitch neural code.
Summary
Shepard tone illusion is perpetual pitch glide via fading octave partial stacks — acoustic Penrose stairs documented by Shepard and Risset, explained by virtual pitch and harmonic cortical tuning.[1][3][8] Fascinating for attention; hostile to sleep and background focus.[11][13]
Sound Bubbles reserves such paradoxes for optional curiosity, not default wellbeing beds.[11]
Limits
Individual pitch resolution and hearing loss alter illusion strength.[21] Not vertigo therapy — can worsen discomfort.[10] Illusions educate perception; they do not heal or harm by frequency magic.[20]
How this article was researched
We combine first-hand experience placing and tuning Sound Bubbles gardens with citations from peer-reviewed journals, reviews, and institutional pages (including NIH/NCBI, sleep and hearing literature, acoustics, and attention research). Where evidence is mixed or early, we say so. On wellbeing topics we stay cautious: these are companion soundscapes, not cures.
References
Sources cited in this article. Prefer primary literature and institutional guidance; Sound Bubbles is not a medical device and these citations do not imply clinical endorsement.
- (1964). Circularity in relative pitch. Journal of the Acoustical Society of America. doi:10.1121/1.1917109
- (1964). Circularity in relative pitch (extended). Journal of the Acoustical Society of America.
- (1969). Pitch and rhythm paradoxes in music. Journal of the Acoustical Society of America.
- (2013). The Psychology of Music. Elsevier.
- (1984). Pitch, consonance, and harmony. Journal of the Acoustical Society of America. doi:10.1121/1.390978
- (2004). The planum temporale as a computational hub. Trends in Neurosciences. doi:10.1016/j.tins.2004.10.011
- (2006). Sweet Anticipation: Music and the Psychology of Expectation. MIT Press.
- (2022). Neural population tuning reveals harmonic structure in music and speech. Nature Neuroscience. doi:10.1038/s41593-022-01114-5
- (1973). An optimum processor theory for the central formation of the pitch of complex tones. Journal of the Acoustical Society of America. doi:10.1121/1.1912371
- (2003). Vibration sensitivity and the lower limit of pitch. Journal of the Acoustical Society of America. doi:10.1121/1.1568949
- (2009). Habituation revisited. Neurobiology of Learning and Memory. doi:10.1016/j.nlm.2008.09.015
- (2013). Auditory scene analysis. Neuron. doi:10.1016/j.neuron.2013.07.018
- (2001). Auditory distraction and short-term memory. Human Factors. doi:10.1518/001872001775992390
- (2010). The free-energy principle. Nature Reviews Neuroscience. doi:10.1038/nrn2787
- (2013). Whatever next? Predictive brains, situated agents. Behavioral and Brain Sciences. doi:10.1017/S0140525X12000477
- (2024). Sound and music based interventions — evidence overview. NIH.
- (2024). Hearing and communication. NIH.
- (2021). World report on hearing. WHO.
