Hacker News Reader

Comment by mchahn | Hacker News Reader

> using lossless analog compression

What does this mean? Analog is inherently lossy to begin with.

It's an awkward term, I think by "compression" they actually mean "encoding". The images were probably sent using something like SSTV:

https://en.wikipedia.org/wiki/Slow-scan_television

or just FM - FM on tape was used for data aquisition in the 60s in cases where dynamic range and linearity were more important than bandwidth.

I think you have got that backwards.

A digital signal is inherently lossy as it takes only finite number of values because digital is discrete (0 and 1), whereas analog is continuous and in theory it can encode an infinite number of values, subject to your codec apparatuses limits.

I was referring to the fact that an "exact value" doesn't even exist. There is noise on every analog signal. If there is no exact value then lossless becomes a meaningless term.

Strictly, that doesn't matter. Depending on the signal being sent, digital is lossless, as well. Specifically, for a bandlimited signal, digital can be lossless.

Basically, the sampling theorem is a hella fun topic. There was a very good video that served as a fun primer on this from a few years ago, if you want me to look it up.

Analog compression was a very important topic before the days of digitization; it was the subject of my dad's first job out of university in 1961 (television bandwidth compression in that case).

And in fact analog modulation can precisely encode the source signal; consider how music or voice was sent uncompressed over radio/through the phone.

Wasn't phone audio compressed in that frequencies outside a narrow range for human voice were cut?

Reply on news.ycombinator.com