[mikepurvis]

Acoustic modems are fascinating devices— you're transmitting at most tens of bits per second before error correction. To make such a paltry amount of data do anything useful is a pretty interesting challenge. You're limited to just bare telemetry and a few status flags / commands.

[Loocid]

>you're transmitting at most tens of bits per second before error correction.

You can get in the kb/s range over a couple of kilometres. You don't have to drop to tens of bits per second until you're trying to transmit over many nautical miles.

Examples: https://geo-matching.com/uploads/default/b/e/benthos-modem-p...

[vorpalhex]

This stuff is for automated buoys or sensors, not for pushing video frames.

Being to push a handful of ints every few seconds is more than enough for a lot of cases, especially if the "modem" (radio?) is cheap enough.

[dylan604]

but why do buoys or sensors need to transmit that data underwater? They could easily do that via radio transmissions above the surface.

to me, anytime underwater communication is concerned, it is solely to be in the realm of military submersed vessel communication. There's no compelling reason for any other communication to undergo the expense of underwater transmission.

[azalemeth]

I guess the other use case that comes to my mind is the oil and gas industry who have a lot of underwater equipment, which could conceivably use an acoustic modem as a backup device.

[acters]

also substantially slower than the speed of light, while not the biggest factor is also slower than other forms of transmission.

As you said, there is a lot of noise that needs powerful error correction to be used, such as how the Reed-Solomon code was used in deep space communication. Prepackaging information before relying on wireless communication is usually the most necessary part of any reliably complex system.