Public corrections
What was wrong here
Every statement on these pages that turned out to be wrong — with the date, the old wording, and what stands there now. We do not change things quietly.
In a field where almost everyone overstates, the only currency that counts is checkability. A site that was always right has either never claimed anything concrete or tidied its mistakes away. We would rather put ours on display. If you find a correction here that we have missed, send it to us — it will end up on this page.
A cortisol figure the data do not support
We had taken the numbers from the abstract of the study, not from the data. Reading the full text showed: the 1.35 is the starting value of the 440 group, not its final one — we had mixed up beginning and end. And the changes within the groups were none of them statistically sound; the 432 group started out at half the value already. With fifteen people per group, the difference in final values is in all likelihood an accident of how the groups were drawn.
Aravena et al., J Appl Oral Sci 28 (2020), e20190601 · independent review, University of Padua, 2023
Milan was 1881, and 432 was not Verdi’s number
Two errors in one line. The Congresso dei Musicisti Italiani met in June 1881, not 1880. And Verdi was indeed the most prominent advocate of a lower pitch — but his favourite was the French 435; he accepted 432 only, on the grounds that the difference was “so small, almost imperceptible to the ear”. The battle cry “Verdi tuning” for 432 was not coined until 1988, a good eighty years after his death.
Gazzetta musicale di Milano, 13 Nov 1881 · Carteggio Verdi–Ricordi 1882–1885 · Rosenberg, J Popular Music Studies 33 (2021)
A caveat missing that the authors state themselves
On the study that found a lower pulse at 432 Hz, one sentence was missing: the authors write themselves that they cannot say whether this is down to the number 432 or simply to its being lower. The finding stands — the qualification belongs with it, and it was not there.
Hohneck et al., BMC Complement Med Ther (2025), PMID 39844155
A version number twenty releases old
It was never visible: the entry sits in the structured data and is read only by machines. That is precisely why it went stale unnoticed — for search engines the Tuner was a 0.1.0 for months while the page offered a 0.21.1. A number nobody sees is a number nobody checks. It now comes from the same line as the button; if the link is missing, the build stops.
“Measured”, where nothing was measured
The radio assumes for every station that it sits at 440 and takes off a flat 1.8 per cent. Nothing is measured there — and three other places on the same page say so explicitly. A word in the status line that contradicts that costs more than truth; it devalues the very argument by which we set the Tuner apart from everything else.
How we handle it
What ends up here
Every factual statement that was wrong — figures, years, attributions, technical details. Not typos, rewordings, or changes of mind about design.
The old wording stays
A correction that does not show the error is not one. So both stand here side by side — and the old sentence is not prettied up.
If you find something
Write to us. We check it against the primary source, not against an abstract — that was exactly the error in the first entry above. If it holds, it comes here, with a date.