In article ,
Jaques d'Alltrades wrote:

It's not so simple. Digital TV needs line of site for good reception.

Line of what?

Line of site. If you don't line up your aerial correctly with the site
of the transmitter, you will not get good reception. Really!

It's line of *SIGHT* - nothing to do with the 'site' of the transmitter.
One aerial should be visible from the other. (FSVO visible)

Sorry. I should have said "THIS IS A JOKE".

Er, no. If the digital signal isn't good enough for its error
correction to work, you will get effectively nothing. You will
continue to see and hear SOMETHING for even lower quality analogue
signals, though few people will want to.

Not IME, and I've been dabbling in radio from around 1955 and (slow
scan) TV from the 'eighties, and paid a little attention to talks on
digital stuff at the Amateur Radio Club meetings.

Where you get problems is when the signal is either intermittant or
fluctuating in amplitude. While the second shouldn't affect the output,
it does. A weak digital signal can be amplified and cleaned through
filters.

So can analogue, actually. The fact that most receivers are crap
doesn't mean they have to be.

A weak signal can be cleaned up, yes, but you get problems when you
lose enough of the bits to go over the capacity of the error correction.
ECCs that degrade gracefully are rare, and there are theoretical
reasons to doubt that good ones exist.

I haven't looked into the digital signal technology in detail, but
it is very similar to tape recording technology as far as encoding
goes. And digital recording needs EITHER a significantly higher
bandwidth OR a significantly better signal to deliver the same
quality of analogue signal back.

The whole idea of digital is that it operates on a precise frequency -
as far as the generating and transmitting equipment will permit. It's
all about money - you can cram many more channels into the same area of
the radio spectrum. Being a series of pulses, as long as you can
separate them from background 'hash' a ver weak signal will suffice.

Grrk. You can transmit only a constant signal (i.e. containing no
information) on a precise frequency. Digital transmission actually
uses a MUCH wider frequency range than analogue! Seriously.

HOWEVER, your other point is correct. You can separate transmissions
on different frequencies rather better and, provided that you can
recognise enough bits, strength is irrelevant. But that isn't quite
the same thing as we were talking about ....

This is because you are starting and ending with analogue; when you
are starting and ending with digital (as in computer data), then
transmitting in digital is better. You can only lose information
by converting between the two, not gain it.

That's the same with anything - take a photograph of a photograph of a
photograph and the result isn't to bad, but take a photocopy of a
photocopy of a photocopy and the results can be quite poor.

Yes.

However, with modern equipment the actual losses in conversion of audio
from analogue to digital and vice-versa are miniscule. The limiting
factors are the clipping and compression used in the digital conversion,
and the retranslation of the received data.

No, that misses the point here. If you need to force the converted
signal through a narrow 'channel' (digital or analogue), you lose
information. Inevitably. You can do the conversion near-perfectly,
but a bounded analogue bandwidth needs an infinite digital one and
conversely. The conversion isn't the problem.

Don't get me going about digital radio either...

Don't get me going about analogue tape storage of computer data ....


Regards,
Nick Maclaren.