[rocqua]

Also, 20-30 Db of attenuation for light is already quite a lot. Whilst for a radio signal it is still very conceivable that 30Db of attenuation still allows for a signal to be received.

Our eyes simply aren't very sensitive instruments. And the visible part of the spectrum is uncharacteristically full of 'noise', so it makes some sense that our eyes don't need to detect any signals that are too far below the noise-floor.

That makes me wonder. How much 'darker' is any given bit of radio spectrum as compared to the visual spectrum earth at night.

[elbasti]

> our eyes aren't very sensitive instruments

I take umbrage with that statement! Our eyes are exquisitely sensitive, and most importantly, have staggering dynamic range.

Our eyes are capable of perceiving a single photon [1], albeit noisily (I've been lucky enough to have performed this experiment myself!).

But the greatest thing about our eyes is the dynamic range: the difference in brightness between a moonless, starry night (which we are perfectly capable of navigating by eyesight) and a bright sunny day is nine orders of magnitude. A bright day is a billion times brighter!

Show me an RF receiver or light camera with that dynamic range!

The one place our eyes are limited is in frequency range.

[1] https://www.nature.com/articles/ncomms12172

[rocqua]

I knew the dynamic range was large. I did not know about the sensitivity! That is quite impressive.

The magnitude of the dynamic range is even more impressive if converted to 'stops' from photography, yielding about 30 stops (1 stop halves the light). Whereas a really good camera will do about 15 stops. Though I suppose that the camera gets 15 stops in a single 'scene'. Whilst the 30 stop figure for the human eye does not hold up if half your vision is taken up by daylight and the other half by a night sky. For a single 'scene' I think it becomes hard to define the dynamic range of a human eye though.

[joshspankit]

As well: because of reed solomon a modern digital signal can be reconstructed from flickers and fragments of the radio waves

[effie]

What does this mean? A fragment of a radio wave is still a radio wave and can carry information irrespective of whether reed solomon or other encoding is used, no?

[joshspankit]

Think of it like trying to receive morse code:

If the signal gets choppy and you miss some of it, how do you know what you missed?

`..--- -.... ----- -----` (2600) could come through as `..- -. —` (Uno). Uno is a valid word, so passes validation, but it’s the wrong message.

What Reed-Solomon allows us to do is pad the message with n% of error correcting ‘bits’. That way, if >100-n% of the message gets through, the whole message can be reconstructed from whatever bits that made it. And if not enough of the message made it, it immediately fails the validation check and so you know you must resend the message.

It’s so handy and so solid that it’s used almost everywhere. From optical discs to ECC ram to radio communications and a bunch more.