The Repository Abyss Report: Into the Dark

The 52 Hz Whale Paper Never Used the Word Lonely

September 8, 2026 · Runtime 16:40 · Watch on YouTube

Since 1989 a single source in the North Pacific has been making a tone near 52 hertz. It was tracked for twelve consecutive seasons by the US Navy's SOSUS arrays and published in Deep-Sea Research I in 2004. That paper is thirteen pages long, and the word lonely does not appear in it once.

Chapters

Transcript

A tone, and a word that is not in the paper 0:00

Somewhere in the North Pacific, since at least nineteen eighty-nine, something has been making a tone near fifty-two hertz. It has been recorded for fifteen years and tracked for twelve. In two thousand and four, four scientists published everything they had on it, in a peer-reviewed journal. That paper is thirteen pages long. The word lonely does not appear in it once.

Three claims, one survives 0:24

Almost everything you know about this animal was added afterward. There are three claims that get repeated about it. That no other whale can hear it. That it is alone. That we know what it is. One of those is false by measurement. One was contradicted by the documentary named after it. The third the paper never made. Start with the file.

The file, running ten times too fast 0:49

NOAA's Pacific Marine Environmental Laboratory publishes the call. One file, two hundred and forty kilobytes, free to download since the early two thousands. Almost every version of this story you have ever heard is that file. And almost none of them mention the line printed next to it on the page. The recording is running ten times too fast. This is it, exactly as they publish it.

Speed undone: 151 seconds near 50 hertz 1:26

I ran it through the same script this channel wrote for the Bloop. It undoes the publication speed on both axes at once. Time multiplied by ten. Frequency divided by ten. What comes out is not a fifteen-second chirp at five hundred hertz. It is a two and a half minute event, sitting between forty-nine and fifty-three hertz. So here is the same file, slowed back down to the speed it was made at. This is the pitch of the actual animal, and almost nobody has heard it. If you are on a phone speaker you will hear very little, because a phone speaker cannot reproduce fifty hertz. Hold that thought. It is going to matter.

SOSUS was built to find submarines 2:18

The system that heard it was not built for whales. SOSUS was a Cold War installation: arrays of hydrophones on the seafloor, listening for Soviet submarines. For forty years the data was classified and biologists could not touch it. In nineteen ninety-two it was partially declassified. The tracking of this animal begins in the same year, and that is not a coincidence.

The second network listened for volcanoes 2:43

There was a second network, and it belonged to NOAA. The Vents Program had been running autonomous hydrophones in the eastern Pacific since August nineteen ninety-one. Its job was to listen for volcanoes. Earthquakes and eruptions along the mid-ocean ridge. It picked up the fifty-two hertz calls in the Gulf of Alaska. Neither network was designed to hear this. Both did.

1989, and the waters that barely moved 3:09

The timeline in the paper is precise. Distinctive tones near fifty-two hertz were first noticed in nineteen eighty-nine. For the next three years they came from waters that barely moved, around forty-six north, one twenty-six west. Systematic tracking began in nineteen ninety-two. It ran for twelve consecutive seasons, and every one of them was different.

Twelve seasons, August to February 3:34

The calling is seasonal and the season has a shape. Calls were recorded from August to February, with most of them in December and January. Twelve tracks. Three started in August, four in September, three in October, two in December. But all twelve ended within a few weeks of each other. Nine in January. Three in early February. The beginning wanders. The ending does not.

Table 3, plotted for the first time 4:03

Table three of that paper is the part almost nobody has looked at. It gives you the start date, the end date, the latitude and the longitude of every one of those twelve seasons. Plotted, it is a map of one animal's twelve winters across an ocean basin. As far as I can tell, this is the first time it has been drawn for a general audience.

Meander, west-east, north-south 4:26

The authors sorted the twelve tracks into three shapes, and said plainly that the sorting was arbitrary. Some meandered over short ranges. Some ran west to east. Some ran north to south. The sequence, season by season, was A, B, C, B, A, C, C, B, A, C, A slash C, C. There is no pattern in that, and the paper does not pretend there is.

708 km in 47 days 4:52

The shortest track is the first one, and it barely qualifies as a track. Ninety-two to ninety-three: seven hundred and eight kilometres in forty-seven days. An average of zero point seven kilometres per hour, which is slower than walking pace, by a lot. Significant movement was recorded on six days out of forty-seven. And in that season it called on fifty-five of sixty-three days. Eighty-seven per cent.

11,062 km in 176 days 5:22

Ten years later the same animal did something else entirely. The two thousand two to two thousand three track ran one hundred and seventy-six days. Eleven thousand and sixty-two kilometres. That is fifteen times the distance of the first track, in under four times the days. Half of it meandering, half of it a straight run south.

2.8 km/h, and what the shapes tell you 5:46

Across all twelve, the average speed was two point eight kilometres per hour. The range runs from zero point seven to three point eight. And the shapes separate cleanly on that number. The meandering tracks averaged one point nine. The west-east tracks, two point five. The north-south tracks were the fastest, at three point five. When it went south, it went with purpose.

The gaps: 1 day, 16 days, 78 days 6:12

Now the part that should bother you. This animal goes quiet. Gaps of one to sixteen days happen irregularly through every season. And there are longer ones. Thirty-three days. Forty-two. Seventy-eight. In those windows there is no signal at all, from anywhere, for weeks at a time. So how do you track something that has stopped making the only sound you can find it by?

How you track a thing that stops calling 6:39

You do not. You assume. The paper says so directly: tracking was continued across the gaps. The justification is that speed, direction and distance stayed consistent on both sides of the silence. And that the calls afterwards sounded like the calls before. It is a reasonable assumption and it is clearly labelled as one. Hold on to that, because a lot rides on it later.

39 calls, 42 minutes, one afternoon 7:06

Table one publishes something rarer than the map. Thirty-nine calls from a single day, the third of February nineteen ninety-three. Each one timed to the second. Individual calls last five to seven seconds. The gaps between them run six to twenty-two seconds. They cluster into nine groups, separated by two to six and a half minutes. It is forty-two minutes of one animal's voice, written out line by line.

Claim one: nothing can hear it 7:35

So. The first claim. You have read it a hundred times, in roughly these words. Other whales sing between fifteen and twenty-five hertz, so this one calls at a frequency none of them can hear. That claim is the entire emotional engine of the story. It is also the one that can be checked with a download and ten minutes.

Blue whale, measured: 16 hertz 7:58

The same NOAA page that publishes the fifty-two hertz file publishes the others. Blue whale, northeast Pacific. Same ten times speed-up. Measured with the same script, corrected the same way, it sits between sixteen and seventeen and a half hertz. That is the population this animal has spent twelve winters swimming through.

Fin whale, measured: 21 hertz 8:30

Fin whale, Atlantic, same treatment. Twenty-one to twenty-three and a half hertz. So the numbers in the popular version are roughly right. Blue whales are low. Fin whales are low. Fifty-two hertz is about three times a blue whale's call. The gap is real. What everyone gets wrong is which side of it is the strange one.

Which of these is actually inaudible 9:02

The floor of human hearing is about twenty hertz. Below that you may feel a sound, but you do not hear it as a pitch. The blue whale's call, at sixteen hertz, is under that floor. The fin whale's, at twenty-one, is sitting right on it. Fifty-two hertz is not. It is an ordinary low bass note, around a low G sharp. You could hear it in a room.

The comparison the story picks 9:27

Sit with that for a second, because it inverts the whole story. The popular version always names the whales that sing between fifteen and twenty-five hertz. Every single one of those calls sits at or below the floor of human hearing. So in the comparison the story itself picks, the voice adopted as the symbol of a sound nobody can hear is the one you could actually hear. The inaudible ones are the whales nobody worries about.

7 hertz to 35 kilohertz. They are not deaf 9:55

And a whale is not a person. The best current summary of marine mammal hearing covers low-frequency cetaceans. It puts their functional hearing at roughly seven hertz to thirty-five kilohertz. Fifty-two hertz is not near an edge of that. It is in the middle of it. Christopher Clark, at Cornell, put it about as plainly as it can be put. Blue whales, fin whales and humpbacks can all hear this animal. They are not deaf.

Claim two: it is alone 10:25

Which leaves the second claim, and it is the load-bearing one. That this is one animal, and that it is alone. Every version of this story treats that as the finding. So it is worth asking what, in the actual data, it rests on.

One observation holds the whole thing up 10:41

One observation. Only one series of these calls was ever recorded at a time, with no call overlap. That is the whole argument, and the authors state it as an inference, not a result. One source at a time suggests one animal. It is a good inference. It is also the kind that a single overlapping recording would end.

2010: widely separated sensors 11:04

In twenty ten, sensors off California picked up the pattern on widely separated instruments. John Hildebrand, at Scripps, said what that could mean, and did not overstate it. It could mean multiple animals are singing. Widely separated is the operative phrase. One animal cannot be in two places.

Somebody went looking, in 2018 11:26

Then somebody went looking. In twenty eighteen a seven-day expedition sailed with three of the people best placed to settle it. John Calambokidis of Cascadia Research. Ana Sirovic. John Hildebrand again. The film that came out of it in twenty twenty-one is called The Loneliest Whale. And the thing it establishes, in its own final act, is that there is more than one.

The film disproves its own title 11:51

Read that again, because it is the strangest fact in this episode. A documentary funded and filmed to find the loneliest whale in the world ends by telling you the whale is not the only one. It has been five years. The name has not moved an inch. The correction was published inside the thing doing the publicising.

Claim three: what the paper actually says 12:13

Third claim. That we know what this animal is. You will see it called a blue whale. You will see it called a blue-fin hybrid, confidently, in a lot of places. Here is the paper's own sentence on the subject. We do not know the species of this whale, whether it was a hybrid or an anomalous whale that we have been tracking. That is the finding. Not a gap in the finding.

3.5% fin whale DNA, in a different ocean 12:38

The hybrid idea is not nonsense, and there is real genetics behind the general case. A twenty twenty-four whole-genome study looked at North Atlantic blue whales. It found fin whale DNA making up about three and a half per cent of their genome. The gene flow runs one way, from fin into blue. It is a good paper. It is about a population in a different ocean. It is not about this animal, and nobody has claimed it is.

No sighting, no photograph, no sample 13:08

Because nobody has a sample. This animal has never been seen. Never photographed. Never biopsied. There is no length, no sex, no age, no photograph of a fluke. Everything anybody has ever known about it arrived as pressure changes on a microphone on the seafloor. For twelve years, that was enough to draw a map. It was never going to be enough to name a species.

December 2004, and the letters 13:36

So where did the word come from? The paper went online in October two thousand and four. In December, a newspaper wrote it up. And the letters started arriving at Woods Hole. From people who were deaf. From people who were single. From people who felt unheard. None of that is in the paper. All of it is in the story now.

The pitch has been dropping 13:59

There is one detail the romantic version always leaves out, and it is the most humanising thing in the record. The pitch has been dropping. Over fifteen years of monitoring, the dominant frequency slid down about two hertz, to near fifty. The authors read that the ordinary way. It is what you would expect from an animal getting older and bigger. The calls also became more regular, with fewer long silences.

What would close this case 14:27

This channel does not end an open case on a sad chord, so here is what would actually close it. Three things, in order of how hard they are. One: two fifty-two hertz sources recorded overlapping, from separate positions, on the same arrays. That single observation ends the one-animal inference outright.

One dart, one week of sequencing 14:48

Two: a ship put on a calling animal while it is still calling, close enough to photograph. Three: a skin sample. With the genome work that now exists, one biopsy dart answers the species question in a week. Hybrid, anomalous blue, or something nobody has described. The order matters, because each step makes the next one possible.

Accurate to 1-10 km, in the 1990s 15:13

And none of that is fantasy. Look at what the tracking already did. Source positions were accurate to between one and ten kilometres. Seventy per cent of them came from calls matched on three to five arrays or more. One to ten kilometres is close enough to put a ship on top of something. What is missing is not the technology. It is a vessel, a season, and someone willing to fund a search that might find nothing.

The record, with nothing added 15:42

So here is the record, with nothing added. A tone near fifty-two hertz, from a source that has never been seen. Tracked across twelve winters and sixty-six thousand kilometres of open ocean. Species unknown. Number unknown. Almost certainly audible to every whale it has ever passed. That is a remarkable thing to have on file. It did not need to be sad to be worth twelve years of somebody's attention.

Next: how to read a dive 16:11

Next time, this channel stops telling you about a dive and teaches you to read one. The two laser dots that give you scale. Why the colour drains in the order it does. What a depth number implies about pressure. And how a sentence somebody types in a control room becomes the only way anyone will ever find that clip again. Every example comes from the seven dives already on this channel. No mystery. Just the skill.

Description and sources

Three things get repeated about this animal: that no other whale can hear it, that it is alone, and that we know what it is. This episode checks all three against the record. The first fails to a measurement anyone can repeat -- NOAA publishes the audio, and the blue whale call everyone says is normal sits below the floor of human hearing while the 52 hertz call sits above it. The second was contradicted by the documentary named after it. The third the paper never claimed. What survives is a table of twelve tracks, 66,218 kilometres of open ocean, and an exact statement of what would close the case.

PRINT-READY, FROM THIS CHANNEL

Read This Dive, Vol. 01 — the method, and five dives read with it

https://therepository.gumroad.com/l/read-this-dive

PRIMARY SOURCES

Watkins, W.A., Daher, M.A., George, J.E. & Rodriguez, D. (2004). Twelve years of tracking 52-Hz whale calls from a unique source in the North Pacific. Deep-Sea Research I 51(12): 1889-1901. doi:10.1016/j.dsr.2004.08.006 -- the primary source, read in full. Tables 1, 2 and 3 and the Appendix are this episode's backbone.
NOAA PMEL Acoustics Program, 52Hz sound: https://www.pmel.noaa.gov/acoustics/whales/sounds/sounds_52blue.html -- ak52_10x.wav, public domain, published at 10x real speed.
NOAA PMEL Acoustics Program, whale sound samples: https://www.pmel.noaa.gov/acoustics/sounds_whales.html -- the blue whale, fin whale and humpback files, all measured here with the same script.
Southall, B.L. et al. (2019). Marine mammal noise exposure criteria. Aquatic Mammals 45(2): 125-232 -- functional hearing for low-frequency cetaceans, about 7 Hz to 35 kHz.
Jossey, S., Haddrath, O., Loureiro, L. et al. (2024). Population structure and history of North Atlantic Blue whales inferred from whole genome sequence analysis. Conservation Genetics 25: 357-371. doi:10.1007/s10592-023-01584-5 -- unidirectional gene flow from fin into blue whale, about 3.5% of the genome.
Smithsonian Magazine (2015), reporting Chris Baranuik for BBC Nature -- the 2010 California detections, John Hildebrand (Scripps) on multiple animals, and Christopher Clark (Cornell) on other whales being able to hear this one.
Zeman, J. (2021). The Loneliest Whale: The Search for 52 -- the 2018 expedition with John Calambokidis, Ana Sirovic and John Hildebrand.
NOAA Ocean Exploration, EX1706 dive 12, 27 July 2017 -- hydrophone display video and frame grabs, public domain, via the NCEI geoportal.

Abyss Report reads the deep ocean through the record that resolved it -- instrument, date, position and the primary sources above.

About the material. This episode is acoustic and declared as such. Every spectrogram was measured in production from NOAA PMEL's published WAV files, with the tenfold publication speed-up undone on both axes -- none is traced from an image. The moving footage is NOAA Ocean Exploration video of a hydrophone display aboard NOAA Ship Okeanos Explorer, EX1706 dive 12, public domain, credited on screen; it shows what watching a hydrophone looks like, and is not the 52 hertz watch itself. Nothing here is generated.

#52HertzWhale #Bioacoustics #BlueWhale #NOAA #DeepSea