What does Harmonic Surprise, Shepard Tones and Musical Chills conclude, in numbers, about “Shepard Tones and the Endless Accelerando: Octave Partials, a Bell-Shaped Envelope and the Tritone Paradox”?

Answer: Chapter 3 builds Shepard tones and Risset's accelerando and shows that the spectral overlaps of the two directions of a tritone are exactly equal, both 2.5858, so direction can only come from the listener.

Scope of the numbers: Its parameters are illustrative; all calculations are generated by its programs; it is not medical advice, claims no health effect, and is not marketing-compliance or legal advice. Chapter 1 writes harmony as a Markov chain: under an illustrative matrix a deceptive cadence carries 3.06 bits of surprisal and an authentic cadence only 0.51.

中文版

Data table

The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. standard-library Python onlyprinted values that match the answer: 2.5858

dup-overlapdown-overlapnearer reading
12.65662.4217up
32.64032.4976up
52.60792.5600up
62.58582.5858tie
72.56002.6079down
92.49762.6403down
112.42172.6566down
first tone C C# D D# E F F# G G# A A# B
peak at C D D D = U U U U U = D D
peak at F# U U U = D D D D D = U U
pairs the two listeners hear the same way: 0 of 12
Show the full program
# Chapter 3. p32: why the pitch circle cannot decide up or down at the
# tritone. Part A: tone 0 and tone d (Shepard tones of p31) are compared
# by spectral overlap: each partial of tone 0 is moved up d semitones or
# down 12 - d semitones and multiplied by the weight tone d really has
# there. The larger overlap is the nearer reading. At d = 6 both moves
# have the same length and, by symmetry of the sum, the same overlap.
# Part B: the tritone paradox (Deutsch 1986) as a listener template:
# pitch class x is heard as "high" in proportion to cos(2 pi (x-h)/12);
# h is the listener's peak class. Templates are ILLUSTRATIVE.
import math

F0, MU, SIG = 27.5, math.log2(27.5) + 4.5, 1.5
PC = "C C# D D# E F F# G G# A A# B".split()


def w(x):
    return math.exp(-(x - MU) ** 2 / (2 * SIG ** 2))


def overlap(shift):
    xs = [math.log2(F0) + k for k in range(-8, 20)]
    return math.fsum(w(x) * w(x + shift / 12) for x in xs)


print(" d   up-overlap  down-overlap  nearer reading")
for d in (1, 3, 5, 6, 7, 9, 11):
    u, v = overlap(d), overlap(-(12 - d))
    verdict = "up" if u > v + 1e-9 else "down" if v > u + 1e-9 else "tie"
    print("{:>2} {:>12.4f} {:>13.4f}  {}".format(d, u, v, verdict))


def heard(x, h):
    a = math.cos(2 * math.pi * (x - h) / 12)
    b = math.cos(2 * math.pi * (x + 6 - h) / 12)
    return "U" if b > a + 1e-9 else "D" if a > b + 1e-9 else "="


print("first tone   " + " ".join("{:>2}".format(p) for p in PC))
for h in (0, 6):
    row = [heard(x, h) for x in range(12)]
    print("peak at {:<3}  ".format(PC[h])
          + " ".join("{:>2}".format(r) for r in row))
agree = len([x for x in range(12) if heard(x, 0) == heard(x, 6) != "="])
print("pairs the two listeners hear the same way: {} of 12".format(agree))

Single source and citation

Fan, Y. (2026). Harmonic-Expectation Violation Entropy in Orchestral Scoring, the Shepard-Tone Illusion and the Dynamics of Musical Chills (Version 1.0.0), Chapter 3: Shepard Tones and the Endless Accelerando: Octave Partials, a Bell-Shaped Envelope and the Tritone Paradox. 5D Library. DOI pending

DOI pending · monograph page goes live after publication · 5DL-B-163 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026163,
  author    = {Fan, Yafu},
  title     = {Harmonic-Expectation Violation Entropy in Orchestral Scoring, the Shepard-Tone Illusion and the Dynamics of Musical Chills},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

Other chapter answers in this book