Quality post-switchover

In message , NY
writes
"Another John" wrote in message

...
In article ,
Andy Burns wrote:

Could be... that the signal strength is now too high, if you have an
amplifier, try removing it, if you don't have one then try an attenuator
(Tool Satan have one that's cheap enough to suck it and see)

http://toolstation.com/shop/p86843


Ooohh! Now that's an interesting thought! When we first had freeview
installed (god knows when that was ... years and years and years -- was
the service called Top-up TV or summat?) ... we had a special aerial
engineer come out (there was a special fixed rate, such was the desire
of the providers to get people hooked up).

The engineer fitted an attenuator because he said "the signal's too
strong". Some years later, during one of the many equipment shuffles
that go on in this corner of the household, I removed it (for some
reason). We never noticed any difference at all.

So *perhaps*, after Switch-over, the signal is again "too strong". I
will seek out the attenuator (I never threw it away of course!) and see
if it makes a difference.

Meanwhile, all other suggestions welcome, and thanks very much for the
help so far, chaps.

It would be nice if set top boxes and other DVB tuners would indicate
if the signal strength is *too* strong. Most seem to just full-scale
strength but it doesn't run red (for example) if you might need an
attenuator.

The analogue signals were usually 10dB* higher in level than the
digitals (and often also on adjacent channels to the digitals). If the
signals were all rather too strong, it's likely that the digitals were
getting hammered mainly from the four/five high-level 'alien' analogues.
Now that analogues have gone, it's likely that receivers are happier
receiving six signals of the type they are really intended to receive -
even though their levels have been increased to be the same as what the
analogues were previously. If reception was OK before the final change
to digital-only, it's probable that is also be will be OK now.
*A few were 16dB, and a few were 6dB.
--
Ian

In message , Stewart
writes

They moved from QAM16 to QAM64 on switch over and if your reception is
marginal that might have been enough to dip below successful reception (it's
digital so it either works or it doesn't)

My money is on this as the problem !
At my girlfiend's house, she had perfect reception before DSO, but
intermittent drop -outs post DSO on BBC channels.
So even though power increased a lot at DSO, reception got worse
because the BBC change from 16QAM to 64QAM (to squeeze more programs
into the multiplex) is a less rugged signal.
The problem was fixed with a slight reorientation of the aerial (which
is in the loft). Loft erials are far more likely to have multipath
effects which affects 64QAM much more than 16QAM.

HTH

Indeed, the BER threshold for 64QAM is 6dB higher than for 16QAM. The
change to 64 could tip what was marginal (but adequate) reception over
the edge.
--
Ian

In message , Ian Jackson
writes
In message , Stewart
writes

They moved from QAM16 to QAM64 on switch over and if your reception is
marginal that might have been enough to dip below successful reception (it's
digital so it either works or it doesn't)

My money is on this as the problem !
At my girlfiend's house, she had perfect reception before DSO, but
intermittent drop -outs post DSO on BBC channels.
So even though power increased a lot at DSO, reception got worse
because the BBC change from 16QAM to 64QAM (to squeeze more programs
into the multiplex) is a less rugged signal.
The problem was fixed with a slight reorientation of the aerial (which
is in the loft). Loft erials are far more likely to have multipath
effects which affects 64QAM much more than 16QAM.

HTH

Indeed, the BER threshold for 64QAM is 6dB higher than for 16QAM. The
change to 64 could tip what was marginal (but adequate) reception over
the edge.

But, on second thoughts, would the change not have been accompanied with
a level increase of 10dB?
--
Ian

Ian Jackson wrote:

The analogue signals were usually 10dB* higher in level than the
digitals (and often also on adjacent channels to the digitals). If the
signals were all rather too strong, it's likely that the digitals were
getting hammered mainly from the four/five high-level 'alien' analogues.
Now that analogues have gone, it's likely that receivers are happier
receiving six signals of the type they are really intended to receive -
even though their levels have been increased to be the same as what the
analogues were previously. If reception was OK before the final change
to digital-only, it's probable that is also be will be OK now.
*A few were 16dB, and a few were 6dB.

A 20dB difference in the case of most transmitters round here.

Bill

Ian Jackson wrote:

Indeed, the BER threshold for 64QAM is 6dB higher than for 16QAM. The
change to 64 could tip what was marginal (but adequate) reception over
the edge.

But, on second thoughts, would the change not have been accompanied with
a level increase of 10dB?
I know the difference between 16 and 64QAM is supposed to be 6dB, but it
doesn't actually feel like 6dB. It feels like about 3dB.

Bill

In article , Bill Wright
wrote:
Ian Jackson wrote:

Indeed, the BER threshold for 64QAM is 6dB higher than for 16QAM. The
change to 64 could tip what was marginal (but adequate) reception
over the edge.

But, on second thoughts, would the change not have been accompanied
with a level increase of 10dB?
I know the difference between 16 and 64QAM is supposed to be 6dB, but it
doesn't actually feel like 6dB. It feels like about 3dB.

Don't know the actual reason for that in practice. Have they changed
something else as well like the coding ratio(s)? Note also that a higher
bitrate makes a bigger target for occasional errors to hit. :-)

That said, it is worth bearing in mind that the bulk of the information
theory analysis that people do tends to blythly assume linearity.[1] So
behaviour can be somewhat worse if there is some amount of, say,
intermodulation between muxes or carriers in a mux. In such cases winding
up the power tends to also wind up the level of distortion.

Slainte,

Jim

[1] One of the reasons it is hard to assess the more 'enthusiastic' claims
some make about analog audio systems like Vinyl LP is that attempts to
estimate its data rate ignore effects of its the level of nonlinearity.
Nightmare to try and take that into account! Given this, I can't say I
blame people for ignoring it. :-)

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

On Mon, 19 Nov 2012 15:04:07 +0000, Bill Wright
wrote:

Ian Jackson wrote:

Indeed, the BER threshold for 64QAM is 6dB higher than for 16QAM. The
change to 64 could tip what was marginal (but adequate) reception over
the edge.

But, on second thoughts, would the change not have been accompanied with
a level increase of 10dB?
I know the difference between 16 and 64QAM is supposed to be 6dB, but it
doesn't actually feel like 6dB. It feels like about 3dB.

But if you had very adequate signal before, the +7dB power increase
counts for very little.
Wheras the 6dB loss (or 3dB per Bill) for 64QAM vs 16QAM hurts.
Anyway, my findings with attic aerials (actually only one) which are
more subject to multipath seem to confirm this.

In message , Stewart
writes
On Mon, 19 Nov 2012 15:04:07 +0000, Bill Wright
wrote:

Ian Jackson wrote:

Indeed, the BER threshold for 64QAM is 6dB higher than for 16QAM. The
change to 64 could tip what was marginal (but adequate) reception over
the edge.

But, on second thoughts, would the change not have been accompanied with
a level increase of 10dB?
I know the difference between 16 and 64QAM is supposed to be 6dB, but it
doesn't actually feel like 6dB. It feels like about 3dB.

But if you had very adequate signal before, the +7dB power increase
counts for very little.
Wheras the 6dB loss (or 3dB per Bill) for 64QAM vs 16QAM hurts.
Anyway, my findings with attic aerials (actually only one) which are
more subject to multipath seem to confirm this.

My guess is that a signal showing a poor BER because of multipath
(rather than lack-of-signal noise) would not show any improvement simply
by increasing the TX power.
--
Ian

Some quick comments.

The Tx power is measured differently between analogue and
digital. Analogue was measured on peak black (which is higher
than sync) but the average of the whole signal was very much
lower than that. Digital on the other hand is essentially a
continuous block of signal (the COFDM) which reads a steady power
irrespective of content. Hence the 'real' power of digital
transmissions is higher than the analogue effectively was.

Someone spoke about removing an attenuator and it made no
difference. It won't. Provided the receiver can decode and where
necessary correct the incoming signal you will see no degradation
on the picture for a significant range of signal strength. If the
OP was working off a main station - I think he said Pontop - then
his signal quality wll be high, and it is the signal quality that
matters with digital.

Digital TV's are, from my experience, rather more sensitive than
the older analogues were - they are more modern after all. As a
result it is quite easy to overload the front end. The best
option is to buy a 20dB variable attenuator and find where the
signal indication comes off 100%, then watch it for a bit and see
what happens. If it pixelates occasionally remove a bit of
attenuation to see if it stops, however don't bother if it
occasionally has short freezes (a second or so) especially if the
sound and vision do not freeze at the same time. I have seen this
on several makes of set, branded and own brand, and have yet to
find a solution, although a ferrite around the mains lead and the
aerial cable near entry to the TV did help on one set.

Finally unless you are in a known fringe area consider removing
any masthead preamp, or at least replace it with one of lower
gain. The nature of digital TV - especially the use of adjacent
channels which analogue did not - could easily overload the amp
if it is near its limit and cause all sorts of problems.



[Ducks down under flameproof cover.]


--
Woody

harrogate three at ntlworld dot com

In message , Woody
writes
Some quick comments.

The Tx power is measured differently between analogue and
digital. Analogue was measured on peak black (which is higher
than sync)

No. With 625-line, negative modulation, the sync tips are the peak of
the RF envelope. The power (well, at least the voltage level of the
received signal) is usually measured as 'RMS during peak' - the RMS
signal level at the maximum level of the RF envelope, ie during each of
the 6.4us line syncs and during the broad syncs in the vertical blanking
interval. [Note that it is NOT the 'peak' power/voltage.]

With PAL-I, black is 24% voltage (-2.4dB) below sync, and white 80%
(-14dB). The tips of a full white with a fully saturated yellow colour
subcarrier are around 87% (-18dB).

but the average of the whole signal was very much
lower than that.

It certainly could be (depending on the video content). A completely
black picture is the highest average power (around 2dB less than sync
tips), and a completely white is the lowest. From memory, the range of
maximum-to-minimum average power is around 12dB.

Digital on the other hand is essentially a
continuous block of signal (the COFDM) which reads a steady power
irrespective of content. Hence the 'real' power of digital
transmissions is higher than the analogue effectively was.

I'm not sure that the power of a digital signal is absolutely constant.
I believe that, depending on signal content, a 16QAM MUX can have
occasional peaks around 3dB higher than the measured RMS value, and even
somewhat higher for 64QAM. However, I may have got this wrong.

Someone spoke about removing an attenuator and it made no
difference. It won't. Provided the receiver can decode and where
necessary correct the incoming signal you will see no degradation
on the picture for a significant range of signal strength. If the
OP was working off a main station - I think he said Pontop - then
his signal quality wll be high, and it is the signal quality that
matters with digital.

Digital TV's are, from my experience, rather more sensitive than
the older analogues were - they are more modern after all.

I doubt if digital receivers are really more sensitive. It's just that
they can produce perfect pictures down to an SNR/BER of around 20dB (and
then they pack up), whereas 20dB on an analogue set would give you very
snowy pictures.

As a
result it is quite easy to overload the front end. The best
option is to buy a 20dB variable attenuator and find where the
signal indication comes off 100%, then watch it for a bit and see
what happens.

As I've already said, I think that before the analogue switch-off, some
of the overload problems may have been caused by the analogue signals -
typically 10dB* higher than their long-suffering digital neighbours.
Occasionally 6, 16 or (as Bill says) 20dB).

If it pixelates occasionally remove a bit of
attenuation to see if it stops, however don't bother if it
occasionally has short freezes (a second or so) especially if the
sound and vision do not freeze at the same time. I have seen this
on several makes of set, branded and own brand, and have yet to
find a solution, although a ferrite around the mains lead and the
aerial cable near entry to the TV did help on one set.

Finally unless you are in a known fringe area consider removing
any masthead preamp, or at least replace it with one of lower
gain. The nature of digital TV - especially the use of adjacent
channels which analogue did not - could easily overload the amp
if it is near its limit and cause all sorts of problems.

The OP said that he had 'clear "sight" of Pontop Pike ' (although he
couldn't actually see it), so was unlikely to be using an amplifier.

[Ducks down under flameproof cover.]

Nearly time to duck down beneath the eiderdown (flameproof, or
otherwise).
--
Ian