Quality post-switchover

[Another John]

Haven't posted in this group before so excuse any ignorance which
emerges below:

We're in the North East, Tyne Valley, and so have only had our Switch
Over in September.

Pre switchover, we (in our village) had excellent digital reception. We
have clear "sight" of Pontop Pike (can't actually see it, but there are
no obstacles between us and it; moreover, there are no trees etc
around). Although our own aerial is in the loft, as I said, reception
was very good indeed pre-switchover.

Post-switchover, we are finding that our PVR, which is a five-year-old
Wharfedale, is having trouble making perfect recordings: there are
frequent tiny skips in recorded programmes. Tiny skips, but enough to
ruin the recording.

I've noticed that most of the trouble is on the commercial channels.
However today I found that last night's MOTD (BBC1) was also affected.

Any suggestions, folks? Ought I to replace our PVR with a newer model?
Or are there more technical, but more certain, ways of dealing with the
problem?

Thanks a lot
John

[Scott]

On Sun, 18 Nov 2012 21:51:41 +0000, Another John <lal...@hotmail.com>
wrote:

My guess is that you are locked on to a transmitter other than your
nearest one. I had this problem also. The automatic scan did not
automatically pick the strongest signal.

Find out the actual channel numbers for your intended transmitter:
http://www.ukfree.tv/transmitters.php should help. Then carry out
manual scans for each of the multiplexes.

[Victor Delta]

"Scott" <newsg...@gefion.myzen.co.uk> wrote in message
news:lmmia89vdhpt3q24s...@4ax.com...

Also do a factory settings reset if not done before.

V

[R. Mark Clayton]

"Another John" <lal...@hotmail.com> wrote in message
news:lalaw44-856FE0...@surfnet-nl.ipv4.ptr.145.109.x.invalid...

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)

[Andy Burns]

Scott wrote:

> Another John <lal...@hotmail.com> wrote:
>
>> Post-switchover, we are finding that our PVR, which is a five-year-old
>> Wharfedale, is having trouble making perfect recordings: there are
>> frequent tiny skips in recorded programmes. Tiny skips, but enough to
>> ruin the recording.
>

> My guess is that you are locked on to a transmitter other than your
> nearest one.

Could be, or 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

[Brian Gaff]

I had to use one after they put up the power to stop locking up on Crystal
Palace, so overload is a possibility if the transmitter is close.

Brian

--
Brian Gaff....Note, this account does not accept Bcc: email.
graphics are great, but the blind can't hear them
Email: bri...@blueyonder.co.uk
______________________________________________________________________________________________________________


"Andy Burns" <usenet....@adslpipe.co.uk> wrote in message
news:f4-dnZt-ksIL5zTN...@brightview.co.uk...

[Another John]

In article <f4-dnZt-ksIL5zTN...@brightview.co.uk>,
Andy Burns <usenet....@adslpipe.co.uk> 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.

John

[NY]

"Another John" <lal...@hotmail.com> wrote in message

news:lalaw44-31BF7C...@surfnet-nl.ipv4.ptr.145.109.x.invalid...

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.

[Stewart]

>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

[Jim Lesurf]

In article <8qKdncC0EuEwkzfN...@brightview.co.uk>, NY

<m...@privacy.net> wrote:

> 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.

I suspect this is driven by the marketing dept, just as it was in the days
of serious FM tuners. Back then many tuners had signal strength meters that
were half-way to the end-stop before you could hear music in mono without
background noise, and at the end-stop long before the signal was big enough
to maximise the output SNR. Bit like the old cars with speedo dials showing
up to 140 mph...

"That tuner is more sensitive. Look, you can see its meter reads a much
better signal than the other one!"

Slainte,

Jim

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

[Ian Jackson]

In message <8qKdncC0EuEwkzfN...@brightview.co.uk>, NY
<m...@privacy.net> writes

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

[Ian Jackson]

In message <5a8ka85dtjtiars9d...@4ax.com>, Stewart
<Ste...@gmail.com> writes

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

[Ian Jackson]

In message <IVdKhNEu$jqQ...@g3ohx.demon.co.uk>, Ian Jackson
<ianREMOVET...@g3ohx.demon.co.uk> writes

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

[Bill Wright]

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

[Bill Wright]

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

[Stewart]

On Mon, 19 Nov 2012 15:04:07 +0000, Bill Wright <bi...@invalid.com>
wrote:

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.

[Ian Jackson]

In message <ifkka8lru0nvc7n3j...@4ax.com>, Stewart
<Ste...@gmail.com> writes

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

[Jim Lesurf]

In article <k8dhp7$a2f$1...@speranza.aioe.org>, Bill Wright

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. :-)

[Woody]

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

[Ian Jackson]

In message <RVwqs.654089$EF4.5...@fx23.am4>, Woody
<harro...@ntlworld.spam.com> 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

[Bill Wright]

Woody wrote:

> 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.

Greater sensitivity when dealing with weak signals does not
automatically imply earlier overload when dealing with strong ones.

>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.

I've experimented many a time with this sort of thing. Varying the
signal input to an amplifier whilst observing the BER. But the measuring
instrument will respond differently to signals at different levels. One
partial solution is to vary the amplifier input with an attenuator, but
to apply compensating attenuation between the amp output and the
measuring instrument, so the latter has a constant input level.

>
> 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.

At DSO we had to go round reducing the gain at system head-ends. And
since DSO I've done several domestic installs using a loft aerial and
four way splitter!

> [Ducks down under flameproof cover.]

Why?
Bill

[Jim Lesurf]

In article <Oe57jgM+...@g3ohx.demon.co.uk>, Ian Jackson

<ianREMOVET...@g3ohx.demon.co.uk> wrote:
> In message <RVwqs.654089$EF4.5...@fx23.am4>, Woody
> <harro...@ntlworld.spam.com> writes

> > 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.

It isn't "absolutely constant". But the coding does make it fairly
'efficient' in IT terms. Which means the power averaged over a few symbol
periods does tend to be fairly steady. (All vague terms, of course.)

In theory an occasional combination of symbols might cause a peak spike
that is far bigger than 3dB above average. Caused by all the symbols having
max amplitude and the right relative phases. But I think the spec *expects*
such signals to be clipped at/before TX and for this not to lose data given
the error corrections. Can't recall if the assumed clip is +3dB or
something else, though.

Slainte,

Jim

[Andy Wade]

On 19/11/2012 15:28, Jim Lesurf wrote:
> In article <k8dhp7$a2f$1...@speranza.aioe.org>, Bill Wright
> <bi...@invalid.com> wrote:
>> 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)?

Yes, at least for the PSB muxes where the change was from 16-QAM rate
3/4 back to 64-QAM rate 2/3. "Back" since that was the mode originally
used by On/ITV digital (other than the change to 8K carriers of course).
That change is worth about -4dB, so Bill's pretty-well spot on. The
COM muxes are now on 64-QAM rate 3/4, sacrificing about 1.5 dB for extra
capacity.

--
Andy

[Andy Wade]

On 20/11/2012 09:39, Jim Lesurf wrote:

> It isn't "absolutely constant". But the coding does make it fairly
> 'efficient' in IT terms. Which means the power averaged over a few symbol
> periods does tend to be fairly steady. (All vague terms, of course.

Yes the amplitudes and phases of the individual carriers are effectively
randomised by all the coding and interleaving, so the baseband signal is
like Gaussian noise, and the RF envelope has a Rayleigh APD.

> In theory an occasional combination of symbols might cause a peak spike
> that is far bigger than 3dB above average. Caused by all the symbols having
> max amplitude and the right relative phases. But I think the spec *expects*
> such signals to be clipped at/before TX and for this not to lose data given
> the error corrections. Can't recall if the assumed clip is +3dB or
> something else, though.

It's certainly not clipped that hard - a peak to mean power ratio of
around 10 dB is nearer the mark. For DSO the feeder cables at many TX
sites had to be uprated because the peak voltages to be handled were
higher than for analogue, despite the mean power for digital being less
than the analogue peak sync power. BBC white paper 156 is an
interesting read on this subject:
http://www.bbc.co.uk/rd/publications/whitepaper156.shtml

DVB-T2 has built-in measures for the control of what it calls PAPR (peak
to average power ratio).

--
Andy