Eleven Contacts, One Firmware Build. Line 14 Of The Changelog Explains Them.
Nine charter skippers logged eleven large mid-water targets on one lake in a single season. We asked each of them for the unit model and the build number the log header records. Eight of the nine matched.
BURLINGTON, Vermont — Line 14.
It sits in the release notes for firmware build 4.2.12, between a chart-overlay redraw fix and a note about Bluetooth pairing.
14. Bottom tracking. Corrected a condition in which, where the temperature gradient exceeds 1.4 °C over less than two metres, a second-return seabed echo could be resolved as the true bottom and the true seabed return presented as a discrete mid-water target. Affects HS-660, HS-660c. Introduced in 4.2.11.
Between 9 June and 28 August 2025, nine charter skippers on the broad lake logged eleven large mid-water contacts. Eleven contacts, one lake, eleven weeks. Sighting groups called it the best run of instrument evidence in forty years, and they were not being foolish. Instruments do not get excited.
We asked the nine skippers for one thing, and it was not the screenshot.
What is at the top of the file
Every one of these units writes a header before anything else: model, serial, transducer, and the firmware build running when the file was opened. Four lines of plain text. Nobody reads it.
All nine gave us the raw files, unedited, within three weeks. That should be said early: it did not have to happen.
Eight of the nine headers say the same thing: 4.2.11.
What build 4.2.11 did
Build 4.2.11 shipped on 22 May 2025. The eight units took it between 24 May and 6 June, over marina wifi, unnoticed.
Dr. Ingrid Solheim-Baptiste, a fisheries acoustician paid by this newspaper to read all eleven files, puts the failure in one sentence. The machine got confused about which echo was the floor.
Sound goes down, hits the bottom, comes back up, bounces off the underside of the surface, goes down and returns a second time. That second arrival is the lake bed drawn twice as deep as it is. Bottom-tracking software exists partly to throw it away.
Under a sharp thermocline, the routine in 4.2.11 kept it. It locked onto the doubled echo as the true bottom, and then had to put the real seabed return somewhere. It put it at half the depth it now believed the lake to be — open water, a discrete target with hard edges.
"The contact is real in the sense that the return is real," Solheim-Baptiste said. "It is the lake bed. It is the lake bed at the wrong depth, drawn at a moment when there is nothing at that depth to argue with it."
There is a check, and it is why this desk went to print. In ten of the eleven files, the contact sits within four per cent of half the charted depth.
What Marielle Duclos-Kinnear saw
Duclos-Kinnear skippers out of Point Bay and logged two of the eleven. She was shown the arithmetic in her own wheelhouse.
"Then that is what it was," she said. "I am not going to be sentimental about a machine. I wrote down what the screen said, I gave you the file, you did the work. I would report it again tomorrow."
September
Norvane's reply says a dealer reproduced the fault on 2 August, twenty-six days before the last of the eleven contacts. The fix shipped on 9 September 2025. Seven of the eight units took it within eleven days; the eighth in October, when the boat came out.
No contact of this kind has been logged on that lake since. We asked all nine skippers, the two survey contractors on the channel, and a dive club with its own sounder log.
Eleven months. Nothing.
That is not proof. Absence over eleven months on one lake is thin, and turns into something only if it holds five years. We will ask again every September.
The one that does not fit
Aldous Perreault-Ngo logged a contact on 4 July, on a seven-year-old unit of a different make that has no wireless and has never been updated.
His contact sits at 31 metres over 74 metres of water. That is not half of anything. Solheim-Baptiste has read the file three times and declines to name a candidate.
Perreault-Ngo said the useful thing first. "You have got ten of them," he said. "You have not got mine. Do not write it as though you have got mine."
What this desk takes from it
Nine people saw something on a screen, wrote it down accurately, reported it, and then handed a stranger the underlying data with no conditions. Most institutions this newspaper deals with in a year do not manage the last part once.
The screens were wrong for fifteen weeks because of a regression in a bottom-tracking routine. The correction was published on 9 September 2025, in public, free, at line 14, and nobody who cares about this lake read it — including us, until June.
Norvane's reply runs at the foot of this page and its central complaint is fair. It is a consumer echo sounder with a twenty-degree cone, sold to find fish in a hurry, never offered as an instrument of record. We asked it to be one. So did everybody else.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
We stopped asking what the eleven contacts were and asked what recorded them. Every one of these units writes a four-line header at the top of each log file giving the model, the serial and the firmware build, so we asked all nine skippers for the raw file rather than the screenshot, tabulated the headers, and then read the manufacturer's release notes for every build either side. A fisheries acoustician with no interest in the lake was paid to read all eleven files and told before she started that her report would be published whatever it said.
Who we spoke to
- Marielle Duclos-Kinnear, Charter skipper; logged two of the eleven contacts. Interviewed aboard at Point Bay; supplied both raw log files and the unit serial 12 June 2026 Was the first of the nine to say yes, and telephoned four of the others on our behalf.
- Aldous Perreault-Ngo, Charter skipper; the ninth unit. Interviewed by telephone twice; supplied the 4 July file June and July 2026 His contact is the one this story does not explain, and he said so before we did.
- Dr. Ingrid Solheim-Baptiste, Fisheries acoustician, Bergen. Read all eleven raw log files at this newspaper's request and expense; report published in full Report dated 3 July 2026 Read the same desk's Loch Ness logs last year and was not told what this cluster was reputed to be.
- Norvane Marine Electronics, Manufacturer of eight of the nine units. Written questions on 26 June 2026 with the eleven header extracts attached; reply received 14 July June–July 2026 Printed in full under Right of Reply. Disputes the framing and makes a point this desk accepts.
- Release notes, builds 4.2.9 through 4.3.1, Manufacturer's public support portal. Downloaded, and each checked against third-party archive captures taken within 48 hours of the original posting June 2026 The wording of line 14 has not been altered since it was posted.
What we could not confirm
- The contact of 4 July, logged on a seven-year-old unit of a different make that has never been connected to a network. Nothing in this story touches it. Dr. Solheim-Baptiste cannot place it at half the charted depth and will not guess.
- Whether the regression produced contacts on other water. The build shipped worldwide. Either this happened on a hundred lakes and nobody assembled it, or something about this thermocline was necessary, and we cannot tell you which because we looked at one lake.
- Whether anything was also physically there. A doubled seabed return is drawn at a plausible depth; a fish holding at that depth in the same second is indistinguishable on the file. The acoustician says three of the eleven cannot be separated on the data alone.
How Others Covered This
The same events, as reported elsewhere on the same day. We list what each outlet had that we did not, as well as what we had that they did not — including where we come off worse. Why we print this.
-
The Meridian TelegraphELEVEN TIMES THIS SUMMER, SOMETHING BIG MOVED UNDER THE BROAD LAKE
Ran the cluster as the strongest instrument record in forty years and interviewed six of the nine skippers.
Had that we did not
Two screenshots and a very good passage on what a charter season actually costs.
Left out
The make and the build. Its own reporter had the files and did not open the header, which is not a criticism we are entitled to make loudly, because neither did we until June.
-
Signal & LedgerNorvane Ships 4.2.12 With Bottom-Tracking Fix
A 180-word trade note on the day the release went out, listing all nineteen items.
Had that we did not
Line 14, verbatim, nine months before anybody connected it to anything.
Left out
Nothing. This is the only outlet that got it right and it got it right by accident, in a column nobody on this beat reads.
-
They Buriedthis newspaperEleven Contacts, One Firmware Build. Line 14 Of The Changelog Explains Them.
Asked nine skippers for the log header rather than the screenshot, and had all eleven raw files read by a paid acoustician.
Had that we did not
The eleven headers, the acoustician's report entire, the release notes, and the one file that does not fit.
Left out
Our headline says the changelog explains them. It explains ten of them. The eleventh does not arrive until the fifth crosshead of six, and it should be in the headline, because that is the honest count. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Katarzyna Bramwell-Oyediran, vice-president for product, Norvane Marine Electronics on 26 June 2026 by our reply editor, in writing, with all eleven header extracts and the acoustician's draft report attached, and eighteen days to respond. Replied on 14 July 2026. Printed in full and unedited, at her request and ours.
We published that note. It has been on our support site since 9 September 2025 and it is the fourteenth item in a list of nineteen, which is exactly where a bottom-tracking fix belongs in a list of nineteen items.
I would rather answer your framing than your facts, because your facts are broadly right and there is no dignity in pretending otherwise. Build 4.2.11 contained a regression in the bottom-tracking filter. It affected two models. A dealer in Kingston reproduced it for us on 2 August, our own people reproduced it in the tank in Gdynia eleven days later, and it was corrected in the next scheduled release. Fifteen weeks is our ordinary cycle for a defect that is not a safety defect, and a display artefact is not a safety defect.
Here is my objection, and I would ask you to print it in the same size type as the rest.
You are asking a fish-finder to be a scientific instrument. It is not one, and it has never been sold as one. The unit costs one thousand three hundred and forty dollars. It is a consumer echo sounder for finding fish quickly in moving water, and page 3 of the manual says that returns must be interpreted by an experienced operator and that display artefacts are normal in strongly stratified water. That sentence is not legal cover. It is a true statement about the physics of a single-beam transducer with a twenty-degree cone, and it was true of every echo sounder ever built, including the ones your acoustician uses, which cost forty times as much and also lie to her.
Second. Nobody asked us. Not once, in the whole of that summer, did any skipper, any club, any sighting society or any newspaper telephone this company to say: we are seeing something odd, what firmware are we on. Your reporter is the first person in this entire affair to ring the manufacturer. I find that more remarkable than the defect.
Third, and this is the one that will annoy you. If the standard is that any artefact in any consumer product becomes a matter of public interest the moment somebody builds a story on top of it, then the standard cannot be met by anyone. We ship release notes because we think they should exist. We do not ship them because we expect a newspaper to read them, and I notice that you did not read them either until nine months after we posted them.
What I would have preferred is the story you actually reported. Nine working people handed a newspaper their raw data with no conditions attached and it turned out to be our software. That is a good story and it is creditable to them and discreditable to us, and I have no complaint about it being told.
My complaint is the headline. It reads as though we put something at line 14 in order to bury it. Line 14 is not a hiding place. It is line 14.
Published unedited under our right-of-reply guarantee.
How was this story?
We publish the result, whatever it is. Reader verdicts appear on the front page and in our newsroom metrics.
Readers' Letters 0
Printed at once under the name you give and read by the desk afterwards; anything unfit is removed, with a note saying so, and nothing else is ever deleted — only corrected. Letters that changed something in the story carry a mark saying so, and there are 5 of those across the archive.