How does Television work?

[Mike Watterson]

I'm sure you are more expert at this than me. But some things in retrospect seem to have started too early, like WS CRTs before there was Digital TV. Digital Video on CDs. Digital standards for HD when SD was still hard. DAB radio. DAB has been turned off here last May 2021. It seems more political than any big consumer advantage over FM.

Now lack of standards on minimum technical quality and content. I still remember the obsession with both of those when I was in BBC Communications in the first half of 1970s.

We now have 42" to 55" 4K HDR TVs at about the cost of a 1950s radio (in real terms), but so thin the speakers are worse than a 1960s pocket radio. Most of the channels with junk or repeats and over compressed. No wonder some people watch streaming channels on a tablet and use the big TV for the games console.

[Panrock]

All fascinating!

But what about the basic principles, e.g:

If one were to slow time right down and watch an old CRT display, one would see a scanning spot with trail, painting the picture in orderly fashion, pixel by pixel, line by line. Regardless of picture content.

But if one were to do the same with a modern digital display panel - or say a mobile phone - what would one see?

[Refugee]

The simplest thing to do is to look at hoe they sent photos back from mars in the 1970s.
It took an hour or more to send a photo back.
On a modern screen the lines are all stored in RAM and the pixels on the screen are direct driven.
Updates are sent as required using one of many thousands of software compression methods to get moving images.
The jumpy and blocky screens are down to penny pinching on the bandwidth.

[Panrock]

Thanks Ref.

So there are still 'lines'?

Would one perhaps see the picture growing out of blocks of different sizes sprouting in random places on the screen?

[ppppenguin]

There are still lines and pixels at the camera and screen. In the transmission path the data is processed in many ways. The main data compression methods used happen to be based on blocks, typically of 8 pixels by 8 lines. So when things go wrong you start to see the blocks.

[ppppenguin]

A quick look at possible routes from camera to screen.

The image is originated as an array of pixels and lines by a camera or graphics generator. The data rate at this point is very high, far too high for transmission.

The images are stored either at full data rate, or very lightly compressed.

The data rate must then be reduced to something suitable for transmission. Within each frame this is usually done by block based systems, similar to those used for JPEG. There is a tradeoff between quality and data rate. For moving pictures you need only transmit changes between frames. That's what MPEG does. There are clever and complex methods that work very well indeed.

You then need to package up this data for transmission. The coding method is chosen to suit the nature of the channel. For terrestrial TV (Freeview) this is done by a coding system called COFDM. The details are not important here. Suffice to say that it's resistant to ghosting and other propagation problems. One weakness is that it doesn't work well with fast moving receivers, but that isn't usually a problem. Satellite doesn't have the same propagation problems and so uses a simpler coding system. Broadband internet uses COFDM to send data reliably over the relatively poor quality channel of a copper pair. That data can be anything from emails to full HD video.

The receiver reverses all of these steps.

One inherent problem is that the processing time rises as compression gets better. This isn't just a matter of faster chips. If you're looking over a group of frames to establish what's moving and what's not, the delay is inherently that long. For normal broadcast this means that the viewer sees things a little later, usually of no importance. Where it matters is on contribution circuits. We've all seen interviews done over a connection with excessive delay. There are odd sounding pauses.

[BusyBee]

I must admit I don't know anything much about the modern digital TV but , as far as digital display is concerned, I think the point is about the pixels. They could really be processed in many different ways although it is sensible to think of a sequence of lines as previous. As Jeffrey says, this is done in blocks. I would imagine these are displayed each block being updated, each instantaneously so in practice essentially line and frame update in each block is instant hence combined update so blockiness if it fails. I would also think that the normal practice of video update is used wherin digital memory is written to then this is written to the screen by dedicted hardware. This can have multi parallel interface so, at least in principle, many blocks can be written instantaneously and (theoretically) the whole screen could be updated at the same time with a massively parallel interface. There can be various methods of comparing and checking the signal block by block so error (hence noise / corruption) correction is done digitally. In this I suspect there may be lots of sporadic corruption but with a normal signal minor fluctuations would, I suspect, not be unusual but constantly corrected.

As I said in the first place though, I have no working experience. I just know what is possible but apologies if I wander too much

Tracy

[Panrock]

It seems to me here that the term 'lines' may be just a convention to describe horizontal rows of pixels. Equally, you could say there are vertical 'lines' when talking about the vertical rows!  Whereas, in analogue television displays, 'lines' do really have a special status.  Or... "have I missed something here"?

If there are 8 x 8 pixel 'blocks', presumably this sets the minimum size when we see whole chunks of them in error. I wonder, are the 'blocks' transmitted in traditional order... ie. left-to-right, top-to-bottom?

[Murphyv310]

Interesting thread, I'm also learning things here Steve as I closed down my business in 2006 and lost all interest in modern practices then.
Keep the info coming please.

[`FRANK.C]

720 line systems, always progressive scan, have a total of 750 lines, 30 of which are vertical blanking.
1080 line systems, inerlaced or progressive scan, have a total of 1125 lines, 45 of which are vertical blanking.
The horizontal blanking interval varies with standard to give the same 74.25MHz or 148.5MHz luminance sampling rate.
The Cb and Cr colour difference sampling rates are each half the luminance rate.

Hi Jeffrey
The vertical blanking lines referred to above I would assume are to accommodate CRT TV's.
In a digital system it would appear to be wasteful to transmit them. So are they just inserted at the receiving end or they actually transmitted.
I have to agree this is a interesting thread. I have learned a lot.

Frank