Good look at video compression with some nice bits of history

[ppppenguin]

Well worth watching. I understand this stuff fairly well but though his explanation was really good.

https://www.youtube.com/watch?v=r6Rp-uo6HmI

[Raditechman]

I read your post while I was listening/watching this lecture about compression
https://www.crowdcast.io/e/gresham-compression/1

John

[Radio Fixer]

Well the first one was quick and easy and as said very good. Started on the second and thats going to take more time Ugh!.: never enough. But I will continue later.

thanks Gary

[ppppenguin]

I haven't registered to see the one suggested by John but most of us don't need to understand much beyond the one I recommended.

I'm far from an expert but know a little more than the outline. One point that's often not well made is that there are two sets of complicated maths going on. The first is data compression. Getting many gallons into a pint pot. This leaves you with a number of bits per second that you need to record or transmit. The other is channel coding. This takes the bits and organises them in a way that suits the channel. So it will be different for Freeview, satellite and cable. Channel coding has to allow for imperfect tranmission so you end up sending more bits than just the wanted signal.

Even within Freeview (for example) there's a whole family of channel coding parameters. Such as number of bits per symbol, guard band, number of carriers per multiplex. Don't be put off by the maths in this paper. You can skate lightly over the complicated stuff and still get a good idea of how and why: https://tech.ebu.ch/docs/techreview/trev_278-stott.pdf This paper was written in 1998 and there now be detailed improvements but the theory remains the same.