OK, a Voice.
But Why a Glass Speaker?

A companion piece to Behind the Scenes of the Galvanic Voice. The short version, for anyone arriving here first: the Galvanic Voice is a marine alert device that watches the boat’s sensors continuously and announces what matters in calm spoken sentences — instead of identical beeps on a chartplotter you may or may not be looking at. Why voice rather than yet another set of warning lights is the subject of the other piece. This one answers the question that comes up immediately after that one: OK, a voice — but why a glass speaker?

Speakers on a sailing boat are a small disaster. They corrode after a season. They look ugly drilled into a bulkhead. They leak water in five years no matter what the brochure said. And they point in one direction — at exactly one person — when half the crew is somewhere else entirely. I did not set out to design a glass speaker. I set out to ask a different question, and the glass speaker is what fell out of that question.

What a Marine Speaker Actually Has to Do

The job of a loudspeaker on a sailing boat is not to play music well in a living room. It is to put a calm, intelligible spoken sentence — typically something between “shallow water, three metres ahead” and “man overboard, port quarter, bearing two-eight-zero” — into the ear of whichever crew member needs to hear it, at the moment it matters, in conditions that range from a quiet anchorage to thirty knots of breeze with rain on the dodger.

The traditional answer to that job is to install a marine speaker — a cone-and-magnet enclosure mounted to a bulkhead, wired to an amplifier, originally designed to play music. It works, more or less, until:

• It corrodes after a season at sea, because the cone is rarely properly sealed against salt.
• It leaks water through the gasket five years in, which almost all of them eventually do.
• It looks ugly drilled into a bulkhead, which is its own argument against installing it.
• It points in one direction — at the boat’s owner, usually — when the off-watch is asleep below and the cook is in the galley.
• You eventually install three or four of them around the boat and still miss whoever happens to be standing somewhere none of them are pointing at.

Every cruising sailor reading this has owned, installed, or replaced one of these. Nobody is happy about it.

I Did Not Want to Build a Better Speaker. I Wanted to Ask a Different Question.

The right engineering question, when you find yourself dissatisfied with every available answer in a category, is usually not “how do I make a better X?” but “why am I committed to X in the first place?” I sat with the speaker question for some time before realising I had been asking it the wrong way round.

The right question is not “how do I install a loudspeaker on a boat?” It is “where does the sound need to arrive?”

The honest answer is: wherever the crew happens to be. At the helm. At the chart table. In the galley. On the foredeck. In the bunk. Distributing a fixed installed speaker to each of those places is the wrong solution — because each installation is its own corrosion / leak / drilling / aesthetic / wiring problem, and because the crew moves anyway. The right solution is to put the sound somewhere most of the crew is in the line of, most of the time, in a way that does not require drilling holes in five places.

The Glass Panel on the Front of the Device Is, As It Turns Out, a Surprisingly Good Loudspeaker

Once you ask the question that way, glass becomes interesting — and there are two physics reasons why it works specifically well in this role, both worth saying explicitly.

The first is stiffness. Glass has a very high Young’s modulus — the engineering term for how resistant a material is to bending — on the order of twenty to fifty times higher than the paper or polypropylene used in conventional speaker cones. A stiff panel moves as a coherent surface rather than flexing and distorting locally, which means a compact transducer can deliver high sound-pressure-level output from a panel that fits on a helm console. With a conventional cone of the same overall dimensions, you simply cannot get the same loudness without breakup and distortion. The stiffness of the material is what makes a limited-size, high-output marine loudspeaker physically possible at all — without it, the device would either have to be much larger, or much louder than it actually is.

The second is the geometry of radiation. A cone speaker projects a beam — and the beam narrows with frequency, which is why a marine cone mounted in the cockpit does not reliably reach the chart table, and why every cruising boat that has ever installed one ends up installing a second. A flat panel of comparable dimensions broadcasts across the full hemisphere it faces: every direction the glass is pointing, all at once. For an alert that has to reach the helm, the companionway, and whoever happens to be standing in the saloon simultaneously, hemispherical dispersion is not a nice-to-have — it is exactly what the job calls for, and exactly what a cone cannot give you.

On top of those two physics arguments, glass also happens to be a remarkably good match to the frequencies marine alerts need to carry. The audio band the human brain reads as urgent — voice intelligibility, alert chimes, the warning tones in IMO MSC.302(87) — lives in a relatively narrow range, roughly a couple of hundred hertz to a couple of kilohertz. A glass panel of the right thickness has natural modal behaviour in exactly that band.

The piece of hardware that drives all of this is small. It lives behind the glass. The glass itself does the announcing — out into the cockpit, out through the companionway, out across the saloon — in every direction the panel is facing.

Why This Particular Glass, in This Particular Place

Gorilla-grade glass was not chosen because it sounds good in a brochure. It was chosen for four converging reasons:

• The marine-display industry already uses it. Chartplotter and MFD cover panels have been on this material for years. The supply chain, the toughness story, the optical clarity, the UV resistance — all of that is already solved by an industry that ships millions of units a year.
• It survives the marine environment. Salt spray, UV, the winch handle dropped on it from a metre up. Marine speakers do not survive any of those things gracefully; this glass does.
• It is the right thickness to be excited efficiently. Too thick and the transducer has to work too hard; too thin and the panel is fragile. The thickness used by display covers turns out to be very close to the optimum for an alert-band radiator.
• It is also the screen. The same glass that is announcing the alert is also the surface the LEDs shine through, the surface the sailor reads, and the surface the sailor will gesture at to acknowledge the alert. One piece of glass, four jobs — none of them duplicated, none of them requiring an extra hole in the boat.

What a Galvanic-Voice-Fitted Boat Does Not Need

Worth itemising explicitly, because the negative space is the point:

• A bulkhead speaker drilled into the saloon (one less hole; one less gasket to fail).
• A bulkhead speaker drilled into the cockpit.
• A weatherproof cone speaker bracket on the dodger.
• An external amplifier rack tucked under the chart table.
• Four runs of speaker cable through the headliner.
• The annual ritual of cleaning corroded speaker terminals.
• The five-year ritual of replacing the leaking ones.

The boat keeps its bulkheads intact. The installer makes one hole, in one place, for one device. The device announces from its own glass, in every direction that piece of glass is facing.

And, Because People Ask: Is This Actually New?

The honest answer is yes, in the senses that matter for a sailing boat. Surface-exciter audio — driving a panel rather than a cone — is not a brand-new physics discovery in 2026; it has been used in consumer-electronics and museum installations for years. What is new is the application: treating the structural glass of a marine helm-station device as the loudspeaker diaphragm, with all the engineering refinements that the marine environment imposes — the glass chemistry, the bonding choice, the exciter coupling against salt and vibration, the DSP correction for a thick rigid panel rather than a soft cone, the multi-zone deployment pattern that puts sound in different parts of the boat without putting holes in them.

Because it matters when somebody asks whether this approach is something we just stumbled into last month: the design is the subject of multiple patent applications — covering the core technique, the marine-environment refinements, and several adjacent embodiments. Patents pending across the lot. The detailed claim language is on file with the patent office and is not the point of this piece; the existence of the work is. We did not pick glass because it sounded clever in a brochure. We picked it because we did the work, and the work pointed there.

It Was Never a Clever Trick

I am sometimes asked, with admiration that I appreciate but slightly misdirect, about the cleverness of the glass speaker. I want to be honest: the glass speaker is not a clever trick I wanted to show off. It is the only honest answer to the question “where does the sound need to arrive on a sailing boat, without ruining the boat to put it there?”

Everything else — the transducer choice, the glass thickness, the placement on the helm console — fell out of that single decision. Like a lot of the engineering on the Galvanic Voice, it looks like a feature decision from the outside and was, from the inside, simply the only answer left when the worse ones had been ruled out.