There is no documented evidence that the Nazis, Joseph Goebbels, the Rockefeller Foundation, or any hidden group moved music from 432 to 440 Hz. A=440 was proposed in 1834, broadcast by the U.S. government from 1936, recommended at a London conference in 1939, and standardized by ISO in 1975. The Goebbels story traces to a 1988 political publication. The health studies so far are small pilots. Preferring 432 is fine on its own terms.

I want to be upfront about the tone here, because this topic usually gets one of two treatments. Either someone posts a dramatic video with a cathedral and a drone pad, or someone else dunks on everyone who ever liked a 432 version. Neither helps you decide anything about your record. What helps is doing it the way you'd troubleshoot a weird noise in a mix: name the claim, find where it came from, check it against what's documented, and work out what evidence would actually settle it.

So that's what this is. A rumor file, labeled as rumors, with receipts.

The Goebbels story, and where it actually starts

The most repeated version says Joseph Goebbels, or the Nazi propaganda apparatus generally, pushed the world to 440 to make people anxious or easier to control. It's sticky because part of it is true in general: propaganda ministries really did manipulate media. But a general truth isn't evidence for a specific acoustic claim.

It turns out we can trace this one. Ruth E. Rosenberg's article "Perfect Pitch: 432 Hz Music and the Promise of Frequency" in the Journal of Popular Music Studies (2021) follows the claim back to a 1988 article by Laurent Rosenfeld in a publication of Lyndon LaRouche's organization, titled "How the Nazis Ruined Musical Tuning." It argued that Goebbels, through Radio Berlin, was instrumental in the 1939 conference where 440 was agreed. The same network's Schiller Institute says on its own campaign page that it launched a drive in Milan in April 1988 to restore a pitch based on middle C at 256 Hz, which it tied to A=432 and to Verdi. So the Nazi story and the 432 campaign arrived together, from the same political source, in the same year.

Now line it up against the record covered in our history of concert pitch. Johann Scheibler proposed 440 in 1834, and the Whipple Museum at Cambridge records it being endorsed at Stuttgart that year. ISO's own magazine dates the American Federation of Musicians' adoption of 440 to 1917. NIST notes that WWV has broadcast A440 since 1936. Lynn Cavanagh's history of the A=440 standard, citing a 1949 account by Llewelyn Lloyd of the British Standards Institution, describes the 1939 London conference with the BSI handling the arrangements. And the final standard came from ISO in 1955 and 1975, long after the war, from an international body.

For the Goebbels version to hold up, you'd want primary policy documents showing intent, a physiological mechanism, controlled evidence that 440 does what's claimed, and a chain connecting a Nazi decision to an ISO standard published three decades later. None of those pieces has surfaced in the sources that keep repeating it.

Rockefeller, Rothschild, and the swap-a-villain pattern

The second family of stories keeps the plot and changes the villain. A research note from Middlebury's Center on Terrorism, Extremism, and Counterterrorism, "Fashwave (432 hz): How Audio Frequencies Take on Meaning in Neofascist Communities" (2024), quotes one widely viewed video claiming a 1953 worldwide agreement, based on a Rockefeller Foundation idea, moved music to a "consciousness suppressing" 440. It also quotes a top comment blaming the Rothschilds, and points out the familiar antisemitic shape of that accusation.

The dates don't even line up with the story they're telling. Nothing in the standards record happened in 1953; the milestones are 1834, 1859, 1939, 1955, and 1975. And the structure is worth noticing on its own: a messy, decades-long process involving instrument makers, broadcasters, musicians' unions, and standards bodies gets collapsed into one switch and one motive. That's emotionally satisfying. It just isn't how standards get made.

It's also worth knowing, gently, what company some of these stories keep. The same Middlebury note documents fashwave channels labeling tracks "(432 hz)" alongside extremist imagery. Liking a 432 version of a song doesn't make anyone part of that, obviously. But it's a good reason to be careful about which explanations you repeat when you share one.

Cosmic numbers, checked with a calculator

The third family is mathematical: 432 shows up in astronomy, ancient measures, and geometry, so it must be the universe's frequency. Some of the arithmetic is real. The jump from "this number appears in patterns" to "music tuned to it does something to your body" is where it falls apart, and you can see why with a short script. It does five things: shows that hertz depends on the human-defined second, scores every integer from 400 to 480 against two equally arbitrary sets of "pattern" tests, checks the C=256 claim, repeats the BBC's crystal-divider arithmetic, and simulates a small study where the true effect is zero.

#!/usr/bin/env python3
"""rumor_math.py: test the arithmetic that 432 Hz stories lean on.
Usage: python3 rumor_math.py
Needs: numpy, scipy.

Part 1 shows a hertz value depends on the human-defined second.
Part 2 scores every integer 400..480 against two equally arbitrary
'pattern' test sets, to show that whoever picks the tests picks the winner.
Part 3 checks the C=256 / A=432 claim under equal temperament.
Part 4 repeats the BBC crystal-divider arithmetic for 432, 439, 440.
Part 5 simulates a small cross-over study with NO real effect to see
how often at least one outcome still comes out 'significant'."""
import math
from fractions import Fraction
import numpy as np
from scipy import stats

print("PART 1  A hertz is cycles per *second*, and the second is a convention")
for name, sec in [("SI second", 1.0), ("minute", 60.0), ("French decimal second (0.864 s)", 0.864)]:
    print(f"  A=432 expressed per {name:32s}: {432*sec:9.2f}   A=440: {440*sec:9.2f}")

def factors(n):
    f, p = [], 2
    while p * p <= n:
        while n % p == 0: f.append(p); n //= p
        p += 1
    return f + ([n] if n > 1 else [])
FIB = {5, 8, 13, 21, 34, 55, 89, 144, 233, 377}
dsum = lambda n: sum(map(int, str(n)))
SET_A = {  # the tests 432 enthusiasts tend to reach for
    "digit sum 9":          lambda n: dsum(n) == 9,
    "divisible by 12":      lambda n: n % 12 == 0,
    "divisible by 8":       lambda n: n % 8 == 0,
    "only 2s and 3s":       lambda n: set(factors(n)) <= {2, 3},
    "Fibonacci factor":     lambda n: any(n % f == 0 for f in FIB),
}
SET_B = {  # equally arbitrary tests, chosen by someone else
    "divisible by 11":      lambda n: n % 11 == 0,
    "multiple of 10":       lambda n: n % 10 == 0,
    "divisible by 55 (Fib)":lambda n: n % 55 == 0,
    "digit sum 8":          lambda n: dsum(n) == 8,
    "divisible by 8":       lambda n: n % 8 == 0,
}
print("\nPART 2  Pattern scores for every integer 400..480")
for label, T in (("test set A", SET_A), ("test set B", SET_B)):
    sc = {n: sum(t(n) for t in T.values()) for n in range(400, 481)}
    best = max(sc.values())
    winners = [n for n, s in sc.items() if s == best]
    print(f"  {label}: top score {best}/5 -> {winners};  432 scores {sc[432]}, 440 scores {sc[440]}")

print("\nPART 3  Does C=256 give A=432?")
print(f"  equal temperament, A = 256 x 2^(9/12)  = {256 * 2 ** (9/12):.2f} Hz")
print(f"  Pythagorean sixth, A = 256 x 27/16     = {256 * 27/16:.2f} Hz")
print(f"  an equal-tempered master retuned to A=432 puts C4 at {432 / 2 ** (9/12):.2f} Hz")
print(f"  a 440 master puts C4 at {440 / 2 ** (9/12):.2f} Hz")

print("\nPART 4  Crystal-divider arithmetic from a 1,000 Hz base")
for f in (432, 439, 440):
    r = Fraction(f, 1000)
    print(f"  {f} Hz = 1000 x {r.numerator}/{r.denominator}   prime: {factors(f) == [f]}")

print("\nPART 5  Small study, no true effect: how often does *something* hit p<0.05?")
rng = np.random.default_rng(432440)
N, SIMS = 33, 20000
for k in (1, 4, 8):
    a = rng.normal(size=(SIMS, N, k)); b = rng.normal(size=(SIMS, N, k))
    p = stats.ttest_rel(a, b, axis=1).pvalue
    rate = np.mean(np.any(p < 0.05, axis=1))
    print(f"  {k} outcome(s), n={N}: at least one p<0.05 in {100*rate:5.1f}% of null studies"
          f"  (theory {100*(1-0.95**k):.1f}%)")
df = N - 1; tcrit = stats.t.ppf(0.975, df)
for d in (0.2, 0.35, 0.5):
    nc = d * math.sqrt(N)
    power = 1 - stats.nct.cdf(tcrit, df, nc) + stats.nct.cdf(-tcrit, df, nc)
    print(f"  paired n={N}, true effect size d={d}: power {100*power:4.1f}%")

Real output from running it on this machine. No audio is involved; part 5 uses simulated random data.

PART 1  A hertz is cycles per *second*, and the second is a convention
  A=432 expressed per SI second                       :    432.00   A=440:    440.00
  A=432 expressed per minute                          :  25920.00   A=440:  26400.00
  A=432 expressed per French decimal second (0.864 s) :    373.25   A=440:    380.16

PART 2  Pattern scores for every integer 400..480
  test set A: top score 5/5 -> [432];  432 scores 5, 440 scores 2
  test set B: top score 5/5 -> [440];  432 scores 1, 440 scores 5

PART 3  Does C=256 give A=432?
  equal temperament, A = 256 x 2^(9/12)  = 430.54 Hz
  Pythagorean sixth, A = 256 x 27/16     = 432.00 Hz
  an equal-tempered master retuned to A=432 puts C4 at 256.87 Hz
  a 440 master puts C4 at 261.63 Hz

PART 4  Crystal-divider arithmetic from a 1,000 Hz base
  432 Hz = 1000 x 54/125   prime: False
  439 Hz = 1000 x 439/1000   prime: True
  440 Hz = 1000 x 11/25   prime: False

PART 5  Small study, no true effect: how often does *something* hit p<0.05?
  1 outcome(s), n=33: at least one p<0.05 in   4.8% of null studies  (theory 5.0%)
  4 outcome(s), n=33: at least one p<0.05 in  18.1% of null studies  (theory 18.5%)
  8 outcome(s), n=33: at least one p<0.05 in  33.4% of null studies  (theory 33.7%)
  paired n=33, true effect size d=0.2: power 20.0%
  paired n=33, true effect size d=0.35: power 49.6%
  paired n=33, true effect size d=0.5: power 79.5%

Part 2 is my favourite result in this whole series. With test set A, the kind of patterns 432 fans reach for, 432 is the unique winner out of 81 numbers. With test set B, just as arbitrary and picked by someone else, 440 is the unique winner. Whoever picks the tests picks the winner. That's not a trick against 432; it's how pattern-hunting works for any number.

Part 3 matters for the "scientific pitch" argument. Middle C at 256 Hz gives A=432 only if you use a Pythagorean major sixth (27/16). In the equal temperament that almost every modern record uses, C=256 gives A of about 430.54. Retune an equal-tempered master to A=432 and its middle C lands near 256.87, not 256. And part 4 kills a tidy story I've seen in both directions: 432 is just as easy to derive from a 1,000 Hz crystal base as 440 is (54/125 versus 11/25). The awkward number was 439, which is prime. That's why Britain's 439 lost, not because 432 was hidden.

One more point that people miss: a hertz is cycles per second, and the second is a human convention. Express the same notes per minute, or per old French decimal second, and every "sacred" number changes while the music doesn't.

The health claims, read carefully

This is where I'd ask for the most care, because people make real decisions on it. The study cited most often is Calamassi and Pomponi's "Music Tuned to 440 Hz Versus 432 Hz and the Health Effects: A Double-blind Cross-over Pilot Study" in Explore (2019). Thirty-three healthy volunteers listened to the same film soundtracks for 20 minutes on two days, once at each tuning. The 432 sessions showed a mean heart-rate drop of 4.79 bpm at p=0.05, a non-significant dip in blood pressure, and a respiratory change at p=0.06. Listeners also reported feeling more focused and satisfied. The authors themselves conclude the experiment should be repeated with a larger sample and randomized controlled trials.

That's an honest pilot, and the right reading is "interesting, unconfirmed." Part 5 of the script shows why. With 33 people and no real effect at all, measuring one outcome gives a false "significant" hit about 5% of the time. Measure four outcomes and at least one hits in about 18% of null studies. Measure eight and it's about a third. And if a real effect were small (d=0.2), a study this size would only detect it about 20% of the time. Small samples cut both ways: they miss real effects and they flag fake ones.

There's an engineering confound too. A 432 version made by classic resampling isn't only lower, it's about 1.85% slower, as What 432 Hz actually changes in a finished master explains. A study that wants to say anything about pitch needs to separate pitch from tempo, which means a time-preserving shift on one arm, and the trade-offs of that are covered in Resampling, pitch shifting, and tempo change are three different jobs. Nothing here supports claims that 432 heals, repairs DNA, or that 440 harms anyone.

The rumor file at a glance

Why these stories feel true

The persuasive frequency rumors all share a recipe. Start with a true fragment: standards did change, institutions were involved, propaganda has used music, and people really do feel different when a song is lower and slower. Then add a dramatic missing cause, a villain with a motive, without the documents or mechanism that would connect the dots. Then repeat it until it feels remembered rather than demonstrated. Numbers add a coat of authority on top.

A quick source check that works for this topic and most others:

• Is there a primary document, a dated record, or a named study you can open yourself?
• Does the timeline fit the events it claims to explain?
• Is a mechanism proposed, and has anyone tested it with controls?
• Does the claim depend on a unit, a rounding choice, or a cherry-picked pattern?
• Who first said it, and what were they campaigning for at the time?

The reasonable middle

Here's what's solid. 440 is a documented standard with a long, boring, practical history. Pitch varied enormously before it. An exact 432 retune is audible and measurable: 31.77 cents lower, and 1.85% longer if you resample. And personal preference is a perfectly good reason to make one. If a lower version of your record sounds better to you, that's enough. It doesn't need a conspiracy behind it, and it's stronger without one.

The best antidote to a scary story is a fair test. How to run a 432 vs 440 listening test you can trust walks through level matching, blind switching, and scoring, the same discipline Reference matching without guesswork applies to mix references.

Sources worth reading

• Ruth E. Rosenberg, "Perfect Pitch: 432 Hz Music and the Promise of Frequency", Journal of Popular Music Studies 33(1), 2021.
• Calamassi and Pomponi, "Music Tuned to 440 Hz Versus 432 Hz and the Health Effects", Explore, 2019 (PubMed abstract).
• Middlebury Institute CTEC, "Fashwave (432 hz): How Audio Frequencies Take on Meaning in Neofascist Communities", 2024.
• Lynn Cavanagh, "A brief history of the establishment of international standard pitch a=440 hertz" (PDF).
• ISO, ISO 16:1975 standard tuning frequency, and NIST's A440 glossary entry.

Who changed the tuning standard from 432 Hz to 440 Hz?

Nobody, because there was no 432 standard to change. 440 was proposed by Johann Scheibler in 1834, broadcast by NIST's WWV from 1936, recommended at a London conference in 1939, and published by ISO in 1975. Earlier pitch varied widely.

Did Joseph Goebbels or the Nazis choose 440 Hz?

No primary evidence supports it. The claim traces to a 1988 article in a publication tied to Lyndon LaRouche. The 1939 London conference was arranged through the British Standards Institution, and 440 was in use well before the Nazi era.

Is 432 Hz the frequency of the universe?

A hertz is cycles per second, and the second is a human-defined unit. Change the unit and the number changes. Number patterns around 432 depend on which patterns you choose to look for.

Is there scientific evidence that 432 Hz music is healthier?

There are small pilot studies, including a 33-person cross-over study that found a borderline heart-rate difference. Its own authors called for larger trials. That is a hypothesis worth testing, not proof of a health benefit.

Is it wrong to prefer 432 Hz?

Not at all. Preferring a lower, slightly slower version of a song is a legitimate taste. It just doesn't need a conspiracy or a medical claim to be valid. Judge it with a fair, level-matched A/B test.