Glitch on BBC1 around time of regional opt out
Martin Evans
but probably on other sources) during BBC Yorkshire and Lincolnshire
(Leeds) regional opt out and in? During the 6pm news this appears to
be done sometime ahead of time rather than being done during a black
screen / quiet sound period.
For a few weeks before Christmas 2004 this gave around a one and a
half second repeat of the London feed when switching from a direct to
indirect London feed and a similar sound and vision gap going the
other way. More recently It is far shorter and gives the effect of a
double stutter at around 6:25pm and a brief sound gap at 6:55pm - it
usually happens during the weather forecast not when the regional
trailers or fade outs are run.
--
Mark S
"Martin Evans" <mce...@dial.pipex.com> wrote in message
news:t03i015bocij9p60a...@4ax.com...
> Has anyone else noticed the glitch (noticed on analogue terrestrial
> but probably on other sources) during BBC Yorkshire and Lincolnshire
> (Leeds) regional opt out and in? During the 6pm news this appears to
> be done sometime ahead of time rather than being done during a black
> screen / quiet sound period.
Not many black screen periods during the news. Not intentionally, anyway.
Roger Wilmut
<mce...@dial.pipex.com> wrote:
Yes, I can confirm this as described. Also at just before 6.00 p.m. the
picture on my Philips TV twitches momentarily. I have also had the
glitches during 'Breakfast'.
Mark Carver
>>For a few weeks before Christmas 2004 this gave around a one and a
>>half second repeat of the London feed when switching from a direct to
>>indirect London feed and a similar sound and vision gap going the
>>other way. More recently It is far shorter and gives the effect of a
>>double stutter at around 6:25pm and a brief sound gap at 6:55pm - it
>>usually happens during the weather forecast not when the regional
>>trailers or fade outs are run.
>
>
> Yes, I can confirm this as described. Also at just before 6.00 p.m. the
> picture on my Philips TV twitches momentarily. I have also had the
> glitches during 'Breakfast'.
Same on BBC South analogue at the 'soft opt' point. There's a 5 or 6
frame 'stutter'. The luminance level drops a bit too.
--
Mark
Please replace invalid and invalid with gmx and net to reply
Martin Underwood
news:36u3ieF...@individual.net...
On BBC South (from Oxford transmitter) there's a stutter just as the
national news ends and the opening titles/music for BBC South news begins. I
think there's also a stutter as they switch back to national news just
before 7 PM. I've never understood why this should be: surely each region
takes its feed permanently from either BBC TV centre in London (for
networked programmes) or from their own studios - and surely those two
sources are in sync so there should be no stutter. What's different about
BBC Suuth (for example) switching between national and one of their VCRs or
cameras and them switching between one of their cameras and another of their
cameras? And why is this only a recent phenomenon?
Also I've noticed that during the closing credits of many BBC dramas
(Casualty, Holby City, The Rotters' Club) there's a stutter as the
cUntinuity announcer compresses/crops the screen to start the "coming next"
trailer that belongs *after* the credits, not during them.
Googolplex
Listen lad I remember the days when TV-AM used to switch over to Anglia
at 9:25 am in 't mornings and prior to the spinning night and the
oh-so-missed music there'd be a pop on the sound and the frame would
actually roll, like the vertical hold on the set was briefly messed up.
You kids these days have got it easy, with your minor 6 frame stutter!
Paul Ratcliffe
wrote:
> Has anyone else noticed the glitch (noticed on analogue terrestrial
> but probably on other sources) during BBC Yorkshire and Lincolnshire
> (Leeds) regional opt out and in?
Most regions have it. It is symptomatic of lack of thought and design of
the opt-opt systems.
> During the 6pm news this appears to be done sometime ahead of time
> rather than being done during a black screen / quiet sound period.
When is there ever black & silence in the 6pm news, or at any other time
convenient for the needs of the regions? Never.
People switch off or over during black and silence, or so we are led to
believe by the marketing types, which is why it is a heinous crime to
transmit (it also puts up the TX 'leccy bill).
> For a few weeks before Christmas 2004 this gave around a one and a
> half second repeat of the London feed when switching from a direct to
> indirect London feed and a similar sound and vision gap going the
> other way.
I guess they were routing digital network to the analogue transmitters.
Not very clever.
> More recently It is far shorter and gives the effect of a
> double stutter at around 6:25pm and a brief sound gap at 6:55pm - it
> usually happens during the weather forecast not when the regional
> trailers or fade outs are run.
Since the regions put out there own trails, they can't opt out half way
through them as they are putting them out. Duh?
Paul Ratcliffe
> On BBC South (from Oxford transmitter) there's a stutter just as the
> national news ends and the opening titles/music for BBC South news begins. I
> think there's also a stutter as they switch back to national news just
> before 7 PM. I've never understood why this should be: surely each region
> takes its feed permanently from either BBC TV centre in London (for
> networked programmes) or from their own studios
They do.
> - and surely those two sources are in sync so there should be no stutter.
They are not. Stations generally free-run relative to network and drift
during the day means there is a timing mismatch at the switching point.
There is also invariably a synchroniser in the feed to the studio, but not
normally to the transmitter, which introduces a frame of delay (it doesn't
need to be like this, but most regions seem to be badly engineered).
Then there are the Energis path switches to contend with. These can happen
without notice and the paths from London to the region are not the same
length. This usually gives a large frame bounce type disturbance.
> What's different about
> BBC Suuth (for example) switching between national and one of their VCRs or
> cameras and them switching between one of their cameras and another of their
> cameras?
VTRs (actually, the way things are going, you might well be more correct).
That switch happens on the vision mixer where everything has to be
synchronous.
The above switch is usually on a separate router where everything does
not have to be synchronous and often isn't.
> And why is this only a recent phenomenon?
It isn't. Some things have made it worse at certain times than it used to
be. Nothing has made it better than it used to be.
> Also I've noticed that during the closing credits of many BBC dramas
> (Casualty, Holby City, The Rotters' Club) there's a stutter as the
> cUntinuity announcer compresses/crops the screen to start the "coming next"
> trailer that belongs *after* the credits, not during them.
That's bad design in the whizzzzzzy new broadcast centre. They are switching
in and out a DVE with a frame of delay (or is it 2?) which looks particularly
vile on fast scrolling credits i.e. most programmes.
Add to that the random aspect ratio changes and the cuts (not fades) in
sound levels, and you have a pretty poor result all round.
Is it any wonder people turn off or over during the credits (which no-one
apart from actors' mothers reads anyway according to the channel
controller)?
Martin Underwood
news:slrnd0kfta...@news.pr.network...
> On Wed, 9 Feb 2005 12:38:00 -0000, Martin Underwood <m...@privacy.net>
> wrote:
>
>> On BBC South (from Oxford transmitter) there's a stutter just as the
>> national news ends and the opening titles/music for BBC South news
>> begins. I
>> think there's also a stutter as they switch back to national news just
>> before 7 PM. I've never understood why this should be: surely each region
>> takes its feed permanently from either BBC TV centre in London (for
>> networked programmes) or from their own studios
>
> They do.
>
>> - and surely those two sources are in sync so there should be no stutter.
>
> They are not. Stations generally free-run relative to network and drift
> during the day means there is a timing mismatch at the switching point.
Really? Why is this. Is it not possible to sync the studio to a standard
reference (eg provided by BBC TV Centre in London) so that each studio can
switch between remote and local to its heart's content without any
interference. There may be different delays between the same signal provided
by two different regions due to delays in the camera-to-transmitter path,
but that's not important unless you are in a fringe area and switch between
two different regions' outputs of the same channel. (When I lived in
Bracknell and could get ITV from both Crystal Palace and Hannington I
sometimes compared the fractional-second delay in the sound between CP
[sound through VCR to hifi] and Hannington [sound through TV]. Nowadays
there could be much longer delays due to frame stores etc - this was in
all-analogue days.)
>> Also I've noticed that during the closing credits of many BBC dramas
>> (Casualty, Holby City, The Rotters' Club) there's a stutter as the
>> cUntinuity announcer compresses/crops the screen to start the "coming
>> next"
>> trailer that belongs *after* the credits, not during them.
>
> That's bad design in the whizzzzzzy new broadcast centre. They are
> switching
> in and out a DVE with a frame of delay (or is it 2?) which looks
> particularly
> vile on fast scrolling credits i.e. most programmes.
> Add to that the random aspect ratio changes and the cuts (not fades) in
> sound levels, and you have a pretty poor result all round.
Why can't the DVE be left switched in all the time but set normally to pass
the signal through unhindered and only altered to split-screen mode when the
continuity announcer feels the urge to burble?
You mention random aspect ratio changes. What's the recommended best
practice on digital TV when both 4:3 and 16:9 material is shown? Should both
be in full horizontal resolution, with the anamorphic flag switched
according to source material's aspect ratio, or should 4:3 be transmitted as
if it was widescreen (possibly with vertical coloured borders, as for some
football matches on ITV!), thus using only part of the horizontal
resolution? I've seen both styles used and always wondered why there isn't a
single standard. I appreciate that switching within a programme (eg between
modern material and archive material) might get a bit tedious for the viewer
but switching at programme-programme or programme-advert junctions is OK.
ITV3 seem to have got that switching down to a fine art.
charles
Underwood <m...@privacy.net> wrote:
> Is it not possible to sync the studio to a standard reference (eg
> provided by BBC TV Centre in London) so that each studio can switch
> between remote and local to its heart's content without any interference.
In the analogue days this was always a problem since the studio had to be in
sync to London if contributing and in sync from London if "opting out".
Nowadays, with frame stores, there shouldn't be a problem, except that some
bean counter probably decided not to provide frame stores.
Dave Liquorice
>> They are not. Stations generally free-run relative to network and
>> drift during the day means there is a timing mismatch at the
>> switching point.
>
> Really? Why is this. Is it not possible to sync the studio to a
> standard reference (eg provided by BBC TV Centre in London) so that
> each studio can switch between remote and local to its heart's
> content without any interference.
Well if there is a disturbance anywhere in the link from TVC to you,
your station syncs take a hit. Might not be very good in the middle of
a one off, never to be repeated, recording.
You also need to take into account the round trip distance between the
regional SWC and the studio. Network is (was in my day...) normaly
routed straight through SWC but when the regional opt out studio asks
for "the switch" network is routed through the studio, this longer
route takes a finite amount of extra time.
--
Cheers new...@howhill.com
Dave. pam is missing e-mail
Martin Underwood
news:nyyfbegfubjuvyypb...@news.howhill.com...
> On Wed, 9 Feb 2005 18:43:26 -0000, Martin Underwood wrote:
>
>>> They are not. Stations generally free-run relative to network and
>>> drift during the day means there is a timing mismatch at the
>>> switching point.
>>
>> Really? Why is this. Is it not possible to sync the studio to a
>> standard reference (eg provided by BBC TV Centre in London) so that
>> each studio can switch between remote and local to its heart's
>> content without any interference.
>
> Well if there is a disturbance anywhere in the link from TVC to you,
> your station syncs take a hit. Might not be very good in the middle of
> a one off, never to be repeated, recording.
Surely it's not beyond the wit of man to have a local oscillator that's kept
in sync with TVC but which is capable of free-running if the TVC sync feed
fails...
> You also need to take into account the round trip distance between the
> regional SWC and the studio. Network is (was in my day...) normaly
> routed straight through SWC but when the regional opt out studio asks
> for "the switch" network is routed through the studio, this longer
> route takes a finite amount of extra time.
Why is Network fed directly to the switching centre rather than permanently
going via each region's studio to avoid having to switch a variable delay in
or out? Or to put it another way, why aren't the studio and the SWC one and
the same thing?
Matthew Sylvester
> Surely it's not beyond the wit of man to have a local oscillator that's kept
> in sync with TVC but which is capable of free-running if the TVC sync feed
> fails...
The SPGs in the regions (I'm basing this on Bristol, but the others are
going to be more-or-less the same) are locked during the day to a local
atomic frequency standard, so if all things work properly there is
little discernable drift through the day relative to TVC. We genlock to
London first thing in the morning (I'll be doing that in about seven
hours - why aren't I in bed?) and all being well will still be in sync
at the end of of the 2200 news sequence. If there is a problem we can
re-genlock during the day, but that means ensuring that nobody is
editing or recording at the point where we press the switch.
> Why is Network fed directly to the switching centre rather than permanently
> going via each region's studio to avoid having to switch a variable delay in
> or out? Or to put it another way, why aren't the studio and the SWC one and
> the same thing?
'Cos the studio is busy doing other things (rehearsing, pre-recording
etc) so can't sit there with network on the desk all day. In the Nations
they have a Continuity studio doing the switch so this doesn't arise.
The problem in the regions is that the network feed has to be
synchronous in two places at once, the input to the mixer, and the opt
switch. This is tricky, as even in analogue world there is a delay
through the mixer, proc amp and SIS coder. The fudge is to have the sync
point at the opt switch, and feed the mixer through a synchroniser so
its feed is exactly one frame late reaching the switch. This means that
when the switch is thrown, you get a clean cut with no roll or colour
flash, but the programme jumps back one frame. If you are seeing a jump
bigger than this something is going wrong - I'll keep an eye on
Hannington tomorrow and see what is happening.
Gareth Rowlands
Martin Underwood wrote:
> > Well if there is a disturbance anywhere in the link from TVC to you,
> > your station syncs take a hit. Might not be very good in the middle
> > of a one off, never to be repeated, recording.
> Surely it's not beyond the wit of man to have a local oscillator
> that's kept in sync with TVC but which is capable of free-running if
> the TVC sync feed fails...
Is there an SPG available on the commercial market that can do this ?
The old BBC Natlock PG's usually ran in 'protect' mode that limited
locking corrections to a rate a 2" Quad machine could keep up with.
Cheers,
Gareth.
--
http://www.rat.org.uk gareth at lightfox dot plus dot com
DB
>> that's kept in sync with TVC but which is capable of free-running if
>> the TVC sync feed fails...
>
> Is there an SPG available on the commercial market that can do this ?
Yes there is. I'm fairly sure the Trilogy SPGs will do this.
Mark Carver
Not without a glitch IME
--
Mark
Please replace invalid and invalid with gmx and net to reply.
Mark Carver
> Mark Carver wrote:
>> Same on BBC South analogue at the 'soft opt' point. There's a 5 or 6
>> frame 'stutter'. The luminance level drops a bit too.
>>
>
> Listen lad I remember the days when TV-AM used to switch over to Anglia
> at 9:25 am in 't mornings and prior to the spinning night and the
> oh-so-missed music there'd be a pop on the sound and the frame would
> actually roll, like the vertical hold on the set was briefly messed up.
Well that was because BT would switch the input to the transmitters
using a 2x1 relay :-) Much the same on Friday evenings in London at the
Thames/LWT switch over.
(In some regions during the early days of TV-am I think BT had to
re-patch the feeds, which took minutes ?)
Mark Carver
[snip]
> The problem in the regions is that the network feed has to be
> synchronous in two places at once, the input to the mixer, and the opt
> switch. This is tricky, as even in analogue world there is a delay
> through the mixer, proc amp and SIS coder. The fudge is to have the sync
> point at the opt switch, and feed the mixer through a synchroniser so
> its feed is exactly one frame late reaching the switch. This means that
> when the switch is thrown, you get a clean cut with no roll or colour
> flash, but the programme jumps back one frame. If you are seeing a jump
> bigger than this something is going wrong - I'll keep an eye on
> Hannington tomorrow and see what is happening.
Thanks for that, I paid special attention this morning, hard to tell as
the soft opt point was carefully selected. On previous occasions when
it's happened during speech, there is a noticeable 'stutter', might
(should !) only be a frame or two after all, just seems worse.
I'd be interested in your observations.
Mark Carver
> I do remember regional continuity on all junctions on BBC1. That was a
> mess. A few frames of the "national" logo was often seen.
It was a mess down here in the Saarf too. Poor old Paul Harris night
after night. It seemed to be a totally self-op operation ?
On a visit to Belfast in 1990 I noticed nasty non sync crashes in and
out of every local BBC 1 NI junction, I assume their synchroniser was
away being repaired ?
Matthew Sylvester
> Thanks for that, I paid special attention this morning, hard to tell as
> the soft opt point was carefully selected. On previous occasions when
> it's happened during speech, there is a noticeable 'stutter', might
> (should !) only be a frame or two after all, just seems worse.
>
> I'd be interested in your observations.
All seemed fine to me this morning. As you say the opt point was chosen
well, which can make a great difference - doing it during music makes it
sound terrible, as the brain seems to notice the change in rhythm very
easily.
Roderick Stewart
> >> Is it not possible to sync the studio to a standard reference (eg
> >> provided by BBC TV Centre in London) so that each studio can switch
> >> between remote and local to its heart's content without any interference.
>
> > In the analogue days this was always a problem since the studio had to be in
> > sync to London if contributing and in sync from London if "opting out".
>
It was. I remember sometimes being involved with this from the Lime Grove end,
and finding that the biggest trouble was getting the regions to send us colour
bars a few minutes before the start of the programme, when for some reason they'd
rather be rehearsing. I think the worst case lockup time for Natlock was about
4.5 minutes, so ideally opt-ins and opt-outs should have been planned to be kept
separate by that amount, but I doubt very much that many people knew this or
understood the reasons for it.
Rod.
DB
> of every local BBC 1 NI junction, I assume their synchroniser was away
> being repaired ?
That wasn't limited to Belfast. When I moved to Wales in the late 80s, one
of the first things I noticed was the glitches in and out of every programme
when Cardiff did their own continuity.
Mark Carver
> Really? Why is this. Is it not possible to sync the studio to a standard
> reference (eg provided by BBC TV Centre in London) so that each studio can
> switch between remote and local to its heart's content without any
> interference.
As an aside to this discussion, I remember in the 70s some ITV companies
would add or subtract one line per field to their local pulses before
opting to network. They could therefore gently phase themselves in sync
before cutting. They also lengthened or shortened the line period to get
the H timing in step too. (Not sure about ScH phase :-) )
My old bedroom telly had an attenuator because the three channels would
splash over each other due to overloading. If I removed the attenuator I
could often see BBC 1 floating in the background on ITV.
You could clearly see the 'pulse cross' suddenly and firmly moving in
the seconds before an opt to network. Southern TV going into News at Ten
from their station clock offered the clearest view. Also LWT would be
very busy during the 'Final Score' segment of World Of Sport.
Somewhere buried in the IBA COP are the allowed parameters for doing
this. Did the Beeb ever do the same trick ?
Alan S.
Since we are remembering, in my teenage years I had a Pye 405-line set
that would show a wave running up (usually) the screen as Associated
Rediffusion prepared to cut to ATV for Crossroads (yes, I admit it!). As
a kid I was not certain why this happened but working in the business I
learned about Genlock etc.
There were times during the 1960's Winter mains voltage reductions when
the audible 10,125Khz line whistle changed note significantly when A-R
cut to another region. I always assumed this was because 405 was mains
locked and the volatge/frequency was different?
Alan S.
charles
Alan S. <use...@nojunkaspal.co.uk> wrote:
[Snip]
> There were times during the 1960's Winter mains voltage reductions when
> the audible 10,125Khz line whistle changed note significantly when A-R
> cut to another region. I always assumed this was because 405 was mains
> locked and the volatge/frequency was different?
If it was mains locked, & I would have expected that in the early 60s, then
that would have applied throughout the country, with the exception of NI
where there was no national grid connection. If the BBC were having a
contribution from Belfast in mains lock days, a reference feed of 50Hz was
sent on a control line. The SPGs had a facility for external 50c/s.
If there really was a change in the 10KHz whistle, then that implies
non-mains lock with fairly crude oscillators. I think the stability was
supposed to be 1 part in 10,000 for the 405 system.
The BBC went off mains lock a bit before colour started, probably 1966.
Alan S.
news:4d3b25a8...@charleshope.demon.co.uk:
I cetainly remember the sound of the line whistle changing on ITV when
they cut to a different region. Maybe it was just the LOPT being given a
kicking by the non-sync cut? I don't remember it happening on the BBC!
Mark Carver
> I cetainly remember the sound of the line whistle changing on ITV when
> they cut to a different region. Maybe it was just the LOPT being given a
> kicking by the non-sync cut?
I think that's exactly what it was, I heard the same in my younger days
on 625 lines during non sync jumps (BBC and ITV).
The sudden flash of chroma noise when mixing from a source with a colour
burst to one without is another memory. Notably ITV dissolving from a
black and white film, to a colour break-caption. That would have been
the TV's chroma gain going through the roof as the burst reduced in
amplitude.
Happy memories :-)
--
Mark
Please replace invalid and invalid with gmx and net to reply.
Dave Liquorice
> Surely it's not beyond the wit of man to have a local oscillator
> that's kept in sync with TVC but which is capable of free-running if
> the TVC sync feed fails...
And what do you do when the fed comes back but at a different relative
timing?
> Why is Network fed directly to the switching centre rather than
> permanently going via each region's studio to avoid having to switch
> a variable delay in or out?
See Pauls response, the studio does other things during the day, is
generally open access and isn't manned 24/7 (not that SWCs are either
but they are not quite so easy to gain access to. Think network
security.
Martin Underwood
>
>> Surely it's not beyond the wit of man to have a local oscillator
>> that's kept in sync with TVC but which is capable of free-running if
>> the TVC sync feed fails...
>
> And what do you do when the fed comes back but at a different relative
> timing?
If a local SPG has drifted off-sync in the half-hour or so of a regional
programme, then you need a better SPG.
I'm very surprised that all TV premises throughout the country aren't
normally (excluding the rare cases when the feed fails) driven from a single
stable source somewhere - well, maybe more than one source for resilience.
Certainly I'd expect all BBC studios to have a common source and all ITV
studios to have another common source.
It's interesting to hear people talk about Nationwide, because as a viewer I
can remember all sorts of cockups cueing the presenters but I can't remember
any problems with rolling pictures or stuttering due to lack of
synchronisaton. I wonder if it's because they used to make use of a brief
fade to black and no sound at the switchover points, whereas now the
switchover tends to occur during the title music at the end of BBC 6 o'clock
news when the stutter, especially on sound, is very noticeable.
How do Telethon events manage things - Children in Need etc. In that case,
regions are switching frequently between local and network, and are even
sometimes sending their feed to network. I don't remember any stuttering
glitches on those programmes.
Roderick Stewart
> It's interesting to hear people talk about Nationwide, because as a viewer I
> can remember all sorts of cockups cueing the presenters but I can't remember
> any problems with rolling pictures or stuttering due to lack of
> synchronisaton.
Good! We must have got it right most of the time then.
> I wonder if it's because they used to make use of a brief
> fade to black and no sound at the switchover points, whereas now the
> switchover tends to occur during the title music at the end of BBC 6 o'clock
> news when the stutter, especially on sound, is very noticeable.
The intention of using Natlock was that incoming sources from the regions were
synchronous and could be selected through normal mixer switching without any
sync disturbances. Like any other human invention it wasn't perfect, but
probably the most logical way with the technology of the time, and mostly it
seemed to work.
Rod.
Martin Underwood
news:VA.0000090...@escapetime.nospam.plus.com...
> In article <420c94e1$0$45620$ed2e...@ptn-nntp-reader04.plus.net>, Martin
> Underwood wrote:
>> It's interesting to hear people talk about Nationwide, because as a
>> viewer I
>> can remember all sorts of cockups cueing the presenters but I can't
>> remember
>> any problems with rolling pictures or stuttering due to lack of
>> synchronisaton.
>
> Good! We must have got it right most of the time then.
Given the technical and cueing difficulties, I'm amazed that Nationwide ran
as smoothly as it did - most of the time.
>> I wonder if it's because they used to make use of a brief
>> fade to black and no sound at the switchover points, whereas now the
>> switchover tends to occur during the title music at the end of BBC 6
>> o'clock
>> news when the stutter, especially on sound, is very noticeable.
>
> The intention of using Natlock was that incoming sources from the regions
> were
> synchronous and could be selected through normal mixer switching without
> any
> sync disturbances. Like any other human invention it wasn't perfect, but
> probably the most logical way with the technology of the time, and mostly
> it
> seemed to work.
So if Natlock allowed the regions' signals to be synchronous back in the
1970s, why is it that they are (apparently) not synchronous nowadays when it
comes to a region swithching between the network and its own output? Doesn't
one tend to imply the other?
Dave Liquorice
>>> Surely it's not beyond the wit of man to have a local oscillator
>>> that's kept in sync with TVC but which is capable of free-running
>>> if the TVC sync feed fails...
>>
>> And what do you do when the fed comes back but at a different
>> relative timing?
>
> If a local SPG has drifted off-sync in the half-hour or so of a
> regional programme, then you need a better SPG.
You not thinking this through properly. Local SPG is synced to remote
"master" SPG master. The remote SPG fails (both 11kV lines fail,
generator doesn't start etc). When the remote SPG comes back the
chances of it still having the same relative timing as the local SPG
(that hasn't drifted in the mean time) are pretty small. What does the
local SPG now do? It's "master" source now has a different timing...
> I'm very surprised that all TV premises throughout the country
> aren't normally (excluding the rare cases when the feed fails)
> driven from a single stable source somewhere - well, maybe more than
> one source for resilience.
More than one source is exactly what is happening for just the reason
you say, resilience. The glitch is more down to the extra time it
takes the network feed to do the down and up to the regional opt
studio. Not differences between network and local station timings.
> How do Telethon events manage things - Children in Need etc. In that
> case, regions are switching frequently between local and network,
> and are even sometimes sending their feed to network.
These days, synchronisers.
Peter Hayes
> In message <420a918d$0$93946$ed2e...@ptn-nntp-reader04.plus.net>
> Martin Underwood wrote:
>
> > > Well if there is a disturbance anywhere in the link from TVC to you,
> > > your station syncs take a hit. Might not be very good in the middle
> > > of a one off, never to be repeated, recording.
>
> > Surely it's not beyond the wit of man to have a local oscillator
> > that's kept in sync with TVC but which is capable of free-running if
> > the TVC sync feed fails...
>
> Is there an SPG available on the commercial market that can do this ?
>
> The old BBC Natlock PG's usually ran in 'protect' mode that limited
> locking corrections to a rate a 2" Quad machine could keep up with.
There were similar problems for mechanical devices long before Quad VTs.
From the "Marconi-EMI System of Television" Handbook, 1939,
http://www.seahaze.demon.co.uk/tv.png
--
Peter
charles
Martin Underwood <m...@privacy.net> wrote:
> "Dave Liquorice" <new...@howhill.com> wrote in message
> news:nyyfbegfubjuvyypb...@news.howhill.com...
> > On Wed, 9 Feb 2005 22:41:13 -0000, Martin Underwood wrote:
> >
> >> Surely it's not beyond the wit of man to have a local oscillator
> >> that's kept in sync with TVC but which is capable of free-running if
> >> the TVC sync feed fails...
> >
> > And what do you do when the fed comes back but at a different relative
> > timing?
> If a local SPG has drifted off-sync in the half-hour or so of a regional
> programme, then you need a better SPG.
Don't think so. The official PAL stability is 1 in 10e6. If my arithmetic
is correct there could have been a drift of 1.8mS in half an hour. That's
about 28 lines.
> I'm very surprised that all TV premises throughout the country aren't
> normally (excluding the rare cases when the feed fails) driven from a
> single stable source somewhere - well, maybe more than one source for
> resilience. Certainly I'd expect all BBC studios to have a common source
> and all ITV studios to have another common source.
The BBC certainly didn't used to rent a line just for the distribution of a
timing reference and I doublt if they do now in these "cost concious days".
The network feed coming up the line obviously could be used, but during
regional news times (that can get used for other things - such as direct
items from Westminster). But it's not so much a stable frequency reference,
but a relative timing reference. If a regional studio has to be synchronous
when opting out it cannot also be synchronous when contibuting to Network.
ISTR that the London > Birmingham circuit used to equate to 12 lines.
Nowadays, with digital codecs it will be much more.
> It's interesting to hear people talk about Nationwide, because as a
> viewer I can remember all sorts of cockups cueing the presenters but I
> can't remember any problems with rolling pictures or stuttering due to
> lack of synchronisaton. I wonder if it's because they used to make use
> of a brief fade to black and no sound at the switchover points, whereas
> now the switchover tends to occur during the title music at the end of
> BBC 6 o'clock news when the stutter, especially on sound, is very
> noticeable.
I suspect there were lots of non-sync cuts. Presentation was less "arty".
> How do Telethon events manage things - Children in Need etc. In that case,
> regions are switching frequently between local and network, and are even
> sometimes sending their feed to network. I don't remember any stuttering
> glitches on those programmes.
Probably CiN is using a synchoniser for its contributions and the regions
are remaining synched to Network.
charles
Martin Underwood <m...@privacy.net> wrote:
> So if Natlock allowed the regions' signals to be synchronous back in the
> 1970s, why is it that they are (apparently) not synchronous nowadays when
> it comes to a region swithching between the network and its own output?
> Doesn't one tend to imply the other?
You haven't been reading what I've already posted. If a region is
synchronous into London it will NOT be synchronous with the network signals
out of London. As I said, for Birmingham it used to be about 25 lines
difference, it would be more the further away from London and even more now
with digital processing of the signals.
In the days of BT circuits, the path delay was a constant. I suspect that with digital circuits, provided by someone else, the routes could well vary.
Dave Liquorice
> Timing considerations do appear to have become complacent since then
> (for example the uneccessarily long delays between the live and
> displayed images by the stage of the Concerts in Buck House -
> unsynchronised synchronisers being used by the looks of it.
Or simply the time it takes the display to process and show the image.
Even ordinary plasma screens take several frames and that delay has
tripped up more than one installation...
Of course with a digital truck there will be a time delay through it
as well. Gareth, how long is that, in general terms, from light
hitting camera sensor to signal come out of the main output?
> Have you compared the delay of a GSM mobile phone to live sound?
That amount of delay is border line to what I can tolerate in a
conversation.
Alan S.
news:420d1202$0$44975$ed2e...@ptn-nntp-reader04.plus.net:
Were decent Synchronisers available in 1981?
I've been watching some VHS recordings of a certain Royal Wedding which
took place in July of that year and, as far as I can see, they were
merrily mixing between sources all over the UK with no apparent sync
problems. It was also a joy to see something with, also, no apparent
sound/vision sync problems. All lip sync looked good. Happy Days!
Alan S.
Phil
Phil: And there was ICElock as well - used by OBs in particular - by
receiving an off aire signal anywhere in the uk, they could be told by a
digitaly inserted signal in vertical blanking how far out of sync they
were, and the equipment then applied the correction.
NAtionwide used to demonstrate the problem - or perhaps hide it:
That monitor wall in the early days that showed more regions than there
were lines available nback to London/LimeGrove. In those days a region
'sat on London to opt ouit to do there bit, but if they were then due to
contribute to Nationwide in the second half, they had to do a rapid retime
of their SPG - or you got a bump. It was alsmost instant with ICElock as
I recall in a demo once when I visited the Mill (when down at Evesham in
79).
They were also early users of the '3-line synchroniser' which meant they
didn't have to get it precisely right. And then, Nationwide went on a
Train tour - and that was the first obvious uyse I recall of a FRAME
synchroniser (as the picture kept freezing when the signal was lost).
Its too easy to think, at the present time, that delaying pictures (and
keeping sound in sync!!) is dead easy.
THe early VTRs has 1/ no timebase correction, and then about 1 microsecond
- using a box as large as a desktop TV (AMTEC) with varicap diodes for
variable delays (analogue). Later came digital correction and memory of 1
line or 3 or 4 or 16 then 32 and finally a frame in the late eighties/
early nineties. Timing considerations do appear to have become complacent
since then (for example the uneccessarily long delays between the live and
displayed images by the stage of the Concerts in Buck House -
unsynchronised synchronisers being used by the looks of it.
As Charles has also repeatedly mentioned. digital encoding brings with it
delays of multiples of frames - as compression is based on whole-frame
images, and this presents problems with cueing. Have you compared the
delay of a GSM mobile phone to live sound?
--
Phil Spiegelhalter: Ph...@fillin.co.uk
==== Technical Training for Broadcasters =====
*RE CUE Mobile DV Multi-Camera Production and Non-Linear Editing*
Martin Underwood
news:4d3bb7cf...@charleshope.demon.co.uk...
> In article <420d1202$0$44975$ed2e...@ptn-nntp-reader04.plus.net>,
> Martin Underwood <m...@privacy.net> wrote:
>
>
>> So if Natlock allowed the regions' signals to be synchronous back in the
>> 1970s, why is it that they are (apparently) not synchronous nowadays when
>> Doesn't one tend to imply the other?
>
> You haven't been reading what I've already posted. If a region is
> synchronous into London it will NOT be synchronous with the network
> signals
> out of London.
I've read this, but I'm not sure I understand "if a region is synchronous
into London it will NOT be synchronous with the network signals
out of London". That's not a criticism of you but something crucial I've
obviously failed to grasp :-( I would have expected that if you are in sync
for contribution to London, then you'd be in sync for the signal coming back
from London.
Dave Liquorice
> ... but something crucial I've obviously failed to grasp :-(
Quite.
> I would have expected that if you are in sync for contribution to
> London, then you'd be in sync for the signal coming back from
> London.
It takes a finite amount of time for your signal to get to London and
a finite amount of time to get back again.
charles
Alan S. <use...@nojunkaspal.co.uk> wrote:
> Were decent Synchronisers available in 1981?
No
> I've been watching some VHS recordings of a certain Royal Wedding which
> took place in July of that year and, as far as I can see, they were
> merrily mixing between sources all over the UK with no apparent sync
> problems. It was also a joy to see something with, also, no apparent
> sound/vision sync problems. All lip sync looked good. Happy Days!
Natlock ruled.
charles
Underwood <m...@privacy.net> wrote:
> I've read this, but I'm not sure I understand "if a region is synchronous
> into London it will NOT be synchronous with the network signals out of
> London". That's not a criticism of you but something crucial I've
> obviously failed to grasp :-( I would have expected that if you are in
> sync for contribution to London, then you'd be in sync for the signal
> coming back from London.
To be synchronous, for mixing purposes means that the frame syncs, PAL
ident, line syncs and finally colour phase must be matched to the mixing
desks own feed of syncs. And, yes, the frequency stability of the two
signals must be such that there is no drifting out of phase during the
period of time when mixing takes place. An accuracy of 35nS is at the back
of my mind for sync timing, but its 30+ years since I did any of that work.
All done with cut cables inside TVC.
So, consider a region contibuting to London. It's sync pulse genrator needs
to be locked to London, for stability, but the timing of the signals from it
needs to be such that these signals are synchoronous in London after
transiting a few hundred miles of cable (in the simplest situation). This
means that at a given moment in time, the signals in the region are a number
of lines early wrt London.
Now consider the mixed distributed signal. It will arrive in the region
after transiting the same few hundred miles of cable and will therefore be
dealyed on the direct output of that region by twice as much delay as the
region's signal suffered reaching London. In the old days, as I think I
said, this total delay was about 24 lines in the case of Birmingham. So, it
won't be synchronous.
Does that make sense.
Martin Underwood
news:420dfde7$0$42579$ed26...@ptn-nntp-reader03.plus.net...
Ah!!!!!!
The penny has FINALLY dropped... Now I see what I'd forgotten: the time for
the signal to travel to and from London and to go through any equipment in
London.
Even assuming straight line paths and signals travelling at the speed of
light (rather than some unknown fraction of it), it would take significant
time for the signals to make the round trip:
For example:
London to Bristol is about 100 miles or 160 km (1.6 x 10^5 m)
Light travels at 3 x 10^8 m/sec
So round trip time is (2 x 1.6 x 10^5) / (3 x 10^8) or approx 1 msec.
Or, to put it another way, about 16 lines.
I feel a right plonker now :-(
Peter
"Martin Underwood" <m...@privacy.net> wrote in message
news:420dfde7$0$42579$ed26...@ptn-nntp-reader03.plus.net...
The point you are missing is the time it takes for a signal to cross the
country is very fast but not infinitely fast. By the time the signal gets to
the region it will be delayed. The region syncs up to the signal arriving
there, which will be a little later than when it left London. Then if the
signal has to go back to London it will have two lots of delay. This could
all be solved by moving the regions to White City.
Peter
Phil
> To MIX in LOndon irt has to be within about 1us for MONOCHROME
> or 3ns for colour
Phil: Should have read 25nS for Monochrome or SECAM
And if you can do it without introducing delays..
a/ You can predict the future **
b/ you can travel in time like Dr Who
c/ Can I have tonights wining lottery numbers please
** eg my prediction that VHS would never catch on
Mark Carver
> In article <Xns95FB738F0D4D6...@195.129.110.67>,
> Alan S. <use...@nojunkaspal.co.uk> wrote:
>>Were decent Synchronisers available in 1981?
> No
Well they were, they were just very expensive. ISTR the first
synchroniser to be used in anger was at the 1976 Olympics. There was a
sequence during the opening ceremony that required a mix from a camera
on an airship, to the main stadium cameras. Therefore the signal from
the remote camera had to be rendered locally synchronous. Quantel built
the kit, and have never looked back !
My local ITV station TVS started using syncronisers in 1983.
They were made by Wokingham firm Questech, I knew the cap that designed
them.
>>I've been watching some VHS recordings of a certain Royal Wedding which
>>took place in July of that year and, as far as I can see, they were
>>merrily mixing between sources all over the UK with no apparent sync
>>problems. It was also a joy to see something with, also, no apparent
>>sound/vision sync problems. All lip sync looked good. Happy Days!
>
>
> Natlock ruled.
How did ITV do it then ?
Phil
Mark Carver <mark....@invalid.invalid> wrote:
> Well they were, they were just very expensive. ISTR the first
> synchroniser to be used in anger was at the 1976 Olympics. There was a
> sequence during the opening ceremony that required a mix from a camera
> on an airship, to the main stadium cameras. Therefore the signal from
> the remote camera had to be rendered locally synchronous. Quantel built
> the kit, and have never looked back !
> My local ITV station TVS started using syncronisers in 1983.
> They were made by Wokingham firm Questech, I knew the cap that designed
> them.
(charles)...
> > Natlock ruled.
> How did ITV do it then ?
Phil: My first job in television, was at Thames in R+D over a vacation (75)
Tom Sloley was working on an ITV (faster) version of Natlock - and I did
some of the board layouts. When I joined the Beeb a year later, Natlock
seemed a bit strange (especially since it was based on 625NTSC frequencies
- because the BBC had aready bought the crystals for that - and they were
expensive in those days!!!).
Other projects going on were inbuilt region idents, resurrecting (but not
for long) Thames' automation system which involved PDP8i's, and VDUs
(handbuilt prototypes) which relied on a FIELD GLASS BLOCK DELAY LINE!
which occupied the whole of the base of the VDU - simply to display the
next 10 and 20 scheduled items. There was also an early Telextext decoder.
Circuits really were 'breadboarded' using scrap bits of timber fromthe
adjacent carpenters workshop.... and of course there was Jenny Hanley on
site for Magpie, ... oh nostalgia isn't what it used to be.
back to the question - I answered it in my previous answer - pres locked
to the networked feed of ads - which did non sync vuts when the PO changed
regions. Then came framestores.. and much later, LNN playout.
Westward TV had a very advanced form of sync distribution in 76 when I
visited - but I can't recall all the details, wheras tradiional
distribuition at that time still often had separate, as opposed to
composite, syncs.
My Evesham notes on SPGs are about glass block delay lines, and gates.
charles
Phil <Ph...@fillin.co.uk> wrote:
> Westward TV had a very advanced form of sync distribution in 76 when I
> visited - but I can't recall all the details, wheras tradiional
> distribuition at that time still often had separate, as opposed to
> composite, syncs.
There was a system called Unipulse being talked about (in the BBC) at the
time. Certainly, if starting from scratch, 1 cable instead of 7 would might
have made economic sense.
Roderick Stewart
> > It takes a finite amount of time for your signal to get to London and
> > a finite amount of time to get back again.
>
> signal is being passed.
>
> 299792458 metres per second in a vacuum; less in any other medium.
Almost exactly 200 metres per microsecond in a 75 Ohm cable. You can look
at the echo pulses on a scope and use this figure to measure the length of
a drum of the stuff.
Or you can buy something called a "time-domain reflectometer", which is a
poncy name for the same thing.
Rod.
Roderick Stewart
> The official PAL stability is 1 in 10e6. If my arithmetic
> is correct there could have been a drift of 1.8mS in half an hour. That's
> about 28 lines.
That's quite a close tolerance. I recall being told at Evesham that if a
radiolink camera was used on a vehicle alongside a racetrack, it was
necessary to ensure that the receiving aerial was off to one side of the
track, rather than one end of it, because the phase shift introduced by the
rate of change of path length as the vehicle moved would take the subcarrier
out of tolerance at around 30 mph.
I suppose nobody bothers nowadays - just bung it through a synchroniser and
don't worry about how it works.
Rod.
Roderick Stewart
> When I joined the Beeb a year later, Natlock
> seemed a bit strange (especially since it was based on 625NTSC frequencies
> - because the BBC had aready bought the crystals for that - and they were
> expensive in those days!!!).
4.4296875Mhz = "Natlock Frequency".
How sad is it that I remember that?
Rod.
Phil
Phil: Yes that was their system - unipulse
Don't know who made it though.
charles
Roderick Stewart <rj...@escapetime.nospam.plus.com> wrote:
> In article <slrnd0s4...@thinkpad.nowster.org.uk>, Paul Martin wrote:
> > > It takes a finite amount of time for your signal to get to London and
> > > a finite amount of time to get back again.
> >
> > It's all to do with the speed of light in the medium through which the
> > signal is being passed.
> >
> > 299792458 metres per second in a vacuum; less in any other medium.
> Almost exactly 200 metres per microsecond in a 75 Ohm cable. You can look
> at the echo pulses on a scope and use this figure to measure the length of > a drum of the stuff.
It depends on the cable. PSF1/2 had a velocity factor of 0.79, it was
different for PSF 1/3.
charles
Phil <Ph...@fillin.co.uk> wrote:
> In article <4d3c3aaf...@charleshope.demon.co.uk>, charles
> <cha...@charleshope.demon.co.uk> wrote:
> > In article <4d3c39c...@fillin.co.uk>, Phil <Ph...@fillin.co.uk> wrote:
> > > Westward TV had a very advanced form of sync distribution in 76 when I
> > > visited - but I can't recall all the details, wheras tradiional
> > > distribuition at that time still often had separate, as opposed to
> > > composite, syncs.
> > There was a system called Unipulse being talked about (in the BBC) at
> > the time. Certainly, if starting from scratch, 1 cable instead of 7
> > would might have made economic sense.
> Phil: Yes that was their system - unipulse
> Don't know who made it though.
vague memory says "Philips", but after 30 years I could well be wrong.
Gareth Rowlands
Alan S. wrote:
> Were decent Synchronisers available in 1981?
There were Marconi composite PAL 8 bit digital Field Synchronisers in use
in OB's that year.
They were 5u high, and you could turn the individual bits off to get TOTP
style level quantisation effects.
Cheers,
G.
--
http://www.rat.org.uk gareth at lightfox dot plus dot com
Gareth Rowlands
Phil wrote:
> Phil: And there was ICElock as well - used by OBs in particular -
> by receiving an off aire signal anywhere in the uk, they could be
> told by a digitaly inserted signal in vertical blanking how far
> out of sync they were, and the equipment then applied the
> correction.
I'm told by veterans of the Natlock era that ICElock was never used
by OB's. In the Type V mobile control rooms the wiring infrastructure
and control hardware were in place (and the buttons) - but the ICE
decoders were never fitted.
The famous use of ICE was to bring Pebble Mill in to lock, "just
before 1 o'clock".
Cheers !
Gareth.
Gareth Rowlands
Dave Liquorice wrote:
> Of course with a digital truck there will be a time delay through it
> as well. Gareth, how long is that, in general terms, from light
> hitting camera sensor to signal come out of the main output?
Most of the delay is in the CCD. The light falling on the sensor has
to be integrated over an exposure period, then transferred to the back
of the chip in vertical interval (In a FIT CCD). The charge then has
to be clocked out of storage in the CCD once the exposure period is
over.
I think the average time is 1/2 frame to integrate, 1/2 frame average
for the charge to be clocked out and start appearing as video in the
camera head.
Delay through the Scanner is made up of 1 line (approx 64uS) through
the mixer, and at most another 1-3 lines through the most devious of
line routing switchers and line synchronisers where you have mixers
feeding into other mixers (e.g. in VT) which may then need to hot
cut on a matrix to Line.
Going back to the root of your question which was raised by another
contributor. Very noticable delays seen on 'big screens' at 'big
events' have very little to do with careless and gratuitous use of
multiple frame synchronisers, and everything to do with the computer
processing power and time required by PC based software and hardware to
break an image up and reconstitute it as 9+ displays on individual VGA
"monitors".
Matthew Sylvester
> They were 5u high, and you could turn the individual bits off to get TOTP
> style level quantisation effects.
My university student TV station got hold of a secondhand synchroniser
around '87 that could do that. IIRC it had separate Y and C processing
internally, with Y sampled at 6 bits and C at 5 bits. It also had a key
input which you could use to selectively freeze parts of the picture.
Endless hours of fun... :o)
Peter F
"Gareth Rowlands" <gareth.see...@127.0.0.1> wrote in message
news:182e363...@lightfox.plus.com...
>
> Going back to the root of your question which was raised by another
> contributor. Very noticable delays seen on 'big screens' at 'big
> events' have very little to do with careless and gratuitous use of
> multiple frame synchronisers, and everything to do with the computer
> processing power and time required by PC based software and hardware to
> break an image up and reconstitute it as 9+ displays on individual VGA
> "monitors".
>
It might not apply to all large screens but the ones I've worked with don't
introduce any significant display through the screen processor itself if the
incoming is patched direct to the processor.
The processors don't need a particularly powerful pc to run. One of them I'm
pretty sure is a DOS program with a bit of a windows interface. From the
level of pc it will happily run on I reckon it's safe to assume that the
majority of the processing is done in the dedicated hardware cards. From
memory there are 2 types, one handling input and another seperate type for
output.
The actual processing is further distributed thus:
Screen consists of x number of individual panels. Each individual panel
contains its own processor board.
All the data is distributed via fibre optic. Generally, 1 fibre per row of
panels. The fibres just daisy chain from panel to panel.
So, all the main screen processor is really doing is splitting one video
signal into say 5 rows of data on fibre. It's not dealing with individual
led's.
The individual processor boards in the panel are taking on a great deal of
the actual workload; and even then all each one really does is take the data
appropriate for its panel and sending it to the individual led clusters
(which also have a small role in the whole decoding scheme of things) and
then chuck the rest out on a fibre to the next panel.
I can however throw some light on where the delay is coming from...
Couple of frames maybe from a vision mixer, or some new toy that was found
in the rack, stuck in between the incoming feed and the feed to the
processor.
Much more noticeable delay when, instead of using a proper feed on a cable
direct from the scanner truck, it's decided that using an off air signal is
easier.
Note to self: Freeview reception, major broadcast productions and the
security services will probably not work together in harmony!
There you go, probably a little more information than anyone really needed.
Treasure it!
All the best,
Peter
Phil
Peter F <pet...@cheerful.com> wrote:
> "Gareth Rowlands" <gareth.see...@127.0.0.1> wrote in message
> news:182e363...@lightfox.plus.com...
> >
> > Going back to the root of your question which was raised by another
> > contributor.
(PHIL: I E ME)
> > Very noticable delays seen on 'big screens' at 'big
> > events' have very little to do with careless and gratuitous use of
> > multiple frame synchronisers, and everything to do with the computer
> > processing power and time required by PC based software and hardware to
> > break an image up and reconstitute it as 9+ displays on individual VGA
> > "monitors".
> >
> It might not apply to all large screens but the ones I've worked with don't
> introduce any significant display through the screen processor itself if the
> incoming is patched direct to the processor.
> The processors don't need a particularly powerful pc to run.
(SNIPPED)
> All the best,
> Peter
PHIL: Quite - a simple 2d picture 'manipulation (display) requires MINIMAL
processing. An A.D.O. used Z80 processors for calculations, and was
available over 20 yeats ago!
There remains no excuse for the job not being done properly. (Keep I.T.
people away from broadcast!) **
Given the full loop of 1 camera accumulation (FIELD) 1 line through mixer,
and then a MAXIMUM wait of 1 field to reach that displayed part of the
image - which will then be displayed fro a further field - an join the
live action being reaccunmulated in the camera for the howl round (BUT
without the 'lag' of the original Dr. Who titles) does not excuse the
ecessive delay that was apparent.
Not genlocking different 'facilities' companies boxes at each stage could
well add an average 1/2 field per 'box'... how many of these wer in the
chain? ... for no added benefit (just lack of understanding about basic
fundementals like 'genlocking'.
** a '9 image' videowall display is no more complex than 9 parallel ADO
boxes - in fact a lot simpler - so net delay 1 field.
You could argue that a frame is required where the desired pixel did not
align itself on a source pixel from the latest field --- in which case
either A/ it would assess motion as present, and interpolate from the
pixels above and below (and around), or B/ use a source pixel from the
previous field.. hence a 'frame delay'... but only because there was no
difference in the image at that point - so no perceived delay anyway.
(Note that the delay involved in assessing motion includes using 'earlier'
field's pixels - but these are not redisplayed - and so do not contribute
to any perceived delay.. they were simply a calculation/comparison
resource.)
Paul Ratcliffe
> London to Bristol is about 100 miles or 160 km (1.6 x 10^5 m)
> Light travels at 3 x 10^8 m/sec
> So round trip time is (2 x 1.6 x 10^5) / (3 x 10^8) or approx 1 msec.
> Or, to put it another way, about 16 lines.
M'learned colleague Batch always used to say 13 lines of advance for a
Nationwide contribution from BS.
Paul Ratcliffe
<cha...@charleshope.demon.co.uk> wrote:
> The network feed coming up the line obviously could be used, but during
> regional news times (that can get used for other things - such as direct
> items from Westminster). But it's not so much a stable frequency reference,
> but a relative timing reference.
Heck, I don't remember that happening for a long time, but we have more
inter-regional cicuits these days.
Using NET1 distribution as a reverse contribution circuit was always seen as
something to be avoided for fairly obvious reasons.
ISTR it did happen quite regularly on a Friday evening during the 6.30 opt
as we were taking material for the then new Sunday parliamentary programmes.
> If a regional studio has to be synchronous
> when opting out it cannot also be synchronous when contibuting to Network.
> ISTR that the London > Birmingham circuit used to equate to 12 lines.
> Nowadays, with digital codecs it will be much more.
That's all irrelevant now and has been for years.
Paul Ratcliffe
<rj...@escapetime.nospam.plus.com> wrote:
> It was. I remember sometimes being involved with this from the Lime Grove end,
> and finding that the biggest trouble was getting the regions to send us colour
> bars a few minutes before the start of the programme, when for some reason they'd
> rather be rehearsing.
That's what Bars-to-Line was for wasn't it? The studio had Mixer Out
distributed everywhere from a point prior to the Bars-to-Line relay so they
could continue rehearsing whilst sending line-up to keep the chumps in NC1
or wherever happy.
Of course, when Studio B in BS had a bit of a vision refurb. a few years ago,
some muppet from P&ID (we dubbed him The Mad Scotsman, can't remember his
real name) decided he didn't want to reimplement the automatic killing of
Bars-to-Line when the TX red light went on as he couldn't be bothered
thinking about how to do it. At the time, the gallery TX monitor was wired,
as it always had been, from Mixer Out rather than Studio Out and the
inevitable happened - someone opted out without realising and put 100%
colour bars to the transmitter.
The next morning there was a big fuss and Studio Out suddenly appeared
on the gallery TX monitor, much to the annoyance of production.
We are still living with the consequences of that, despite having moved
studios, some 14 years on, as the same system was reimplemented when we
moved back into A (and I think for the same reason).
Paul Ratcliffe
>> The intention of using Natlock was that incoming sources from the regions
>> were
>> synchronous and could be selected through normal mixer switching without
>> any
>> sync disturbances. Like any other human invention it wasn't perfect, but
>> probably the most logical way with the technology of the time, and mostly
>> it
>> seemed to work.
>
> So if Natlock allowed the regions' signals to be synchronous back in the
> 1970s, why is it that they are (apparently) not synchronous nowadays when it
> comes to a region swithching between the network and its own output? Doesn't
> one tend to imply the other?
No, you are talking about switching at two different places, and therefore
at two different times. Natlock was used to advance a region's timing to
make it correct at the switching point in London when contributing to London.
It was obsoleted by synchronisers.
There is no way all the regions could drive London's timing to make the
switching point in the region correctly timed as they would all be fighting
each other.
It is not necessary as it could be done properly with a synchroniser, but
it isn't.
Paul Ratcliffe
<Spa...@pembers.freeserve.co.ukulele.invalid> wrote:
> In the moribund days of University Challenge on ITV when it went out on
> a Sunday morning there was once a major power cut at Granada in
> Manchester, resulting in black level on YTV and a freeze-frame on
> Central which lasted an awfully long time. I expect they had to get the
> continuity announcer out of bed.
Or out of the bar. Redvers used to like a drop, er, allegedly.
Paul Ratcliffe
>>> Surely it's not beyond the wit of man to have a local oscillator
>>> that's kept in sync with TVC but which is capable of free-running if
>>> the TVC sync feed fails...
>>
>> And what do you do when the fed comes back but at a different relative
>> timing?
>
> If a local SPG has drifted off-sync in the half-hour or so of a regional
> programme, then you need a better SPG.
Maybe, but what happens when (as mentioned before) there is a path change
on the Energis network and your feed changes its timing completely?
If you're locked to network, you get a station wide bump as the SPG relocks.
If you're not, then you get a momentary bump on output when you opt out or
opt back in.
> I'm very surprised that all TV premises throughout the country aren't
> normally (excluding the rare cases when the feed fails) driven from a single
> stable source somewhere - well, maybe more than one source for resilience.
They are not for the above reasons. And how would you be able to lock to
more than one stable source anyway, unless their relative timings were
exactly the same, which would never happen?
> Certainly I'd expect all BBC studios to have a common source
Well they don't.
> and all ITV studios to have another common source.
No idea.
> It's interesting to hear people talk about Nationwide, because as a viewer I
> can remember all sorts of cockups cueing the presenters but I can't remember
> any problems with rolling pictures or stuttering due to lack of
> synchronisaton. I wonder if it's because they used to make use of a brief
> fade to black and no sound at the switchover points
Sometimes they did and sometimes they didn't.
> switchover tends to occur during the title music at the end of BBC 6 o'clock
> news when the stutter, especially on sound, is very noticeable.
That's the extra frame going in or coming out. It really upsets the NICAM
some times and others it doesn't. I agree it's horrid. That could all be
solved with a small delay in the network feed in the region to mop up the
delay round the studio loop so the synchroniser didn't have to 'advance'
the pictures by putting in almost but not quite a frame of delay instead
of just delaying the pictures a small amount.
That would sort out the vertical and horizontal picture glitches as well
because the synchroniser would be operating on the same pair of fields in
the 8 field sequence instead of the next pair.
> How do Telethon events manage things - Children in Need etc. In that case,
> regions are switching frequently between local and network, and are even
> sometimes sending their feed to network. I don't remember any stuttering
> glitches on those programmes.
Feeds to network are handled seamlessly because they use synchronisers in
London for all incoming feeds. The regional opt outs operate as normal,
complete with glitches. If you haven't seen them, you haven't been looking
closely enough.
Paul Ratcliffe
wrote:
>> I do remember regional continuity on all junctions on BBC1. That was a
>> mess. A few frames of the "national" logo was often seen.
>
> It was a mess down here in the Saarf too. Poor old Paul Harris night
> after night. It seemed to be a totally self-op operation ?
I think it was in most if not all regions. There were lots of stories
usually involving some combination of the following elements: continuity
announcers, cockups, the bar, copious quantities of alcohol, swearing,
dodgy phone calls going out live because the Pres. mic was left up etc.
Chris Youlden
> Matthew Sylvester wrote:
> [snip]
>
>> The problem in the regions is that the network feed has to be
>> synchronous in two places at once, the input to the mixer, and the opt
>> switch. This is tricky, as even in analogue world there is a delay
>> through the mixer, proc amp and SIS coder. The fudge is to have the sync
>> point at the opt switch, and feed the mixer through a synchroniser so
>> its feed is exactly one frame late reaching the switch. This means that
>> when the switch is thrown, you get a clean cut with no roll or colour
>> flash, but the programme jumps back one frame. If you are seeing a jump
>> bigger than this something is going wrong - I'll keep an eye on
>> Hannington tomorrow and see what is happening.
>
>
> Thanks for that, I paid special attention this morning, hard to tell as
> the soft opt point was carefully selected. On previous occasions when
> it's happened during speech, there is a noticeable 'stutter', might
> (should !) only be a frame or two after all, just seems worse.
>
> I'd be interested in your observations.
>
There is another issue with Southampton opts. The vision mixer used for
Oxford opts, breakfast opts and some other news opts is digital - Grass
Valley 2100 Pres mixer (at least it was when I left over 3 years ago).
The vision delay through the mixer is much more than one line as I think
one poster has estimated, it is in fact 2 frames. Audio must therefore
be delayed to cope with this. When Southampton opt, Network to the
transmitters is replaced (on analogue) by network with 2+ frames delay.
This will be quite enough to cause stutter. The reverse process happens
at the end of the opt.
The procedure used to be that one should 'hot-opt' i.e. throw the switch
with as deadly an accuracy as possible at the point of opt. But this is
often a tall order when what follows (or precedes) is a very active
transition (remembering that the mixer is driven by one person and has
built in audio).
The DTT feeds have logic which refuses to take the mixer output if the
network source is cut up and will therefore switch when that flag
disappears i.e. when local pictures are cut up.
On another point made elsewhere in this thread, there is another reason
why a region does not stay locked to Network which is not so important
these days. If a network fault occurs, say on the Energis network, a
rerouting of that feed somewhere around the country can throw timings
completely out of the window. If you are genlocked when this happens
your programme's pulses can fly all over the place whilst on air. And if
you attempt to genlock just before going back to network the same
problem can befall you. Regions also used to opt on different Networks
during the day, and being in sync with one meant you were not
necessarily in sync with the other.
Southampton previously had a very slow genlock system (not sure if they
do now) which could take a while to sync up with Network.
Curtains if timings changed just before a switch.
Chris Y
Chris Youlden
Agreed Paul. We had a mark on the meter on Chain B pulses enabling us
to quickly advance them to the required position after Points West and
before our contribution to Nationwide, then switch to auto. To do all
of this automatically would have taken most of the programme.
Chris Y
Andy Hame
"Paul Ratcliffe" <ab...@orac12.clara34.co56.uk78> wrote in message
news:slrnd150jv...@news.pr.network...
> On Fri, 11 Feb 2005 12:19:20 +0000 (GMT), charles
> <cha...@charleshope.demon.co.uk> wrote:
>
>> The network feed coming up the line obviously could be used, but during
>> regional news times (that can get used for other things - such as direct
>> items from Westminster). But it's not so much a stable frequency
>> reference,
>> but a relative timing reference.
>
> Heck, I don't remember that happening for a long time, but we have more
> inter-regional cicuits these days.
> Using NET1 distribution as a reverse contribution circuit was always seen
> as
> something to be avoided for fairly obvious reasons.
> ISTR it did happen quite regularly on a Friday evening during the 6.30 opt
> as we were taking material for the then new Sunday parliamentary
> programmes.
I would suspect that the last time the Network distribution was used for
this would be while London Switching Centre was still staffed - about 1990.
We used to allow 2 minutes after the opt point before switching to the
Westminster Inject Point or Norman Shaw North (as it usually would be). I
cannot remember exactly what happened to the timing when we did this, it
must have jumped as I don't think that the Ceefax data bridges had any
synchronising function.
I do remember one of my supervisors of the time managing (after allowing me
to line up one opt, a WIP to Leeds via Birmingham & Manchester Network 1) to
move me out of the way with a flourish and to great dramatics count the 2
minutes down, and then switch this bored looking MP picking his nose onto
Network 2 into the middle of Star Trek.
I think I managed to keep a straight face!
We often used Network 2 distribution during the day to get programmes around
the country, as it would only be showing test card or Ceefax in vision, and
the regional Comms Centres had test card or Ceefax in Vision available
locally, together with a 1/4" tape (usually dusty) of non copyright muzac.
I cannot really believe we used to do that - it seems a lifetime ago!
Andy.
Dickie mint
> I do remember one of my supervisors of the time
Who, Who ?!!!! Initials will do!
>
> We often used Network 2 distribution during the day to get programmes around
> the country, as it would only be showing test card or Ceefax in vision, and
> the regional Comms Centres had test card or Ceefax in Vision available
> locally, together with a 1/4" tape (usually dusty) of non copyright muzac.
Yes, and I got so bored in Brum doing it I actually tried to do it
seamlessly. Cut to Local TC and cue tape at the end of a tune from
London! And back similarly. Invariably someone had forgotten to rewind
the tape, and I usually forgot to check it was ready to go. After
rewinding was the other shift's job.
Richard, in retirement :-)
--
Andy Hame
"Dickie mint" <richard_ta...@trapyahoo.co.uk> wrote in message
news:37j9j1F...@individual.net...
> Andy Hame wrote:
> > I do remember one of my supervisors of the time
>
> Who, Who ?!!!! Initials will do!
>
"Understood, take it as done!"
Dickie mint
> DJM.
> "Understood, take it as done!"
>
>
Thought so! It's him to a T.
Mark Carver
> There is no way all the regions could drive London's timing to make the
> switching point in the region correctly timed as they would all be fighting
> each other.
> It is not necessary as it could be done properly with a synchroniser, but
> it isn't.
Perhaps the BBC network topology could move in the direction of ITV's ?
For instance since Meridian moved to their new 'news only' shed, the
three regional news programmes are fed up a line to LNN South Bank.
All pre-recorded material is played out from LNN, including Meridian's
commercials. The transmitters AIUI are now fed directly from LNN, so the
regional news simply arrives in London, is rendered locally synchronous,
disappears into LNN's giant matrix, and gets fed on to the appropriate
Tx chain. I think all English and Welsh ITV regions will be handled this
way soon, from either LNN or YTV Leeds ?
As for D-Sat the BBC regions are all back hauled to London anyway (in
the MPEG final emission domain ?) after analogue switch off would it not
make sense for the Beeb (sorry BBC Broadcast :-)) to mirror ITV's
arrangement using London (and Manchester?) for all regional playout ?
This would then release the BBC English regions to do what they like
timing wise, and give London direct control of all transmitters (handy
for O-Bits etc) ?
--
Mark
Please replace invalid and invalid with gmx and net to reply
Mark Carver
[snip]
> By the way, is Birmingham still the disaster recovery site for analogue
> TV?
No, and its new location appears to be a state secret. (Although
everyone I know in the broadcasting industry knows where it is :-) ).
I could tell you, but then I'd have to kill you !
Dickie mint
> By the way, is Birmingham still the disaster recovery site for analogue
> TV? If so, would that be the Mailbox or what's left of Pebble Mill?
>
No, that distinction has rested on a famous site above the M25 for over
a year now. There is some involvement at the mailbox.
Now swallow this ........
Pebble Mill was supposed to be demolished from 27th Dec :-) It's still
up, though there is no beeb presence there anymore, it's not owned by
Advantage West Midlands.
DB
> MPEG final emission domain ?) after analogue switch off would it not make
> sense for the Beeb (sorry BBC Broadcast :-)) to mirror ITV's arrangement
> using London (and Manchester?) for all regional playout ?
Actually it's nothing to do with BBC Broadcast. The network arrangements are
still handled by the BBC themselves. BBC Broadcast only provide the feeds.
Not that BBC Broadcast will still exist after switchoff anyway.
![[JC]'s profile photo](http://lh3.googleusercontent.com/a-/ALV-UjWm4GsdhZlMosjaVQjl0vUYQRgxiBcVoQWzukpIVeRpsOZWuQ=s40-c)
[JC]
> Listen lad I remember the days when TV-AM used to switch over to Anglia
> at 9:25 am in 't mornings and prior to the spinning night and the
> oh-so-missed music there'd be a pop on the sound and the frame would
> actually roll, like the vertical hold on the set was briefly messed up.
>
> You kids these days have got it easy, with your minor 6 frame stutter!
Ah but we had the frame rolls even when they were starting up - or at
9:25 after TV-AM played some music.
Mikeapollo
news:slrnd19j...@thinkpad.nowster.org.uk...
> In article <37jnd1F...@individual.net>,
> Mark Carver wrote:
> > Paul Martin wrote:
> > No, and its new location appears to be a state secret. (Although
> > everyone I know in the broadcasting industry knows where it is :-) ).
>
> > I could tell you, but then I'd have to kill you !
>
> So, they're not going to get rid of New Broadcasting House on Oxford
> Road anytime soon? :-)
It's not us up here 'gov, honest!
Mikeapollo
Mike
Paul Martin <p...@zetnet.net> wrote:
>BBC2 used to have some very definitely copyright music on their TC
>tapes in the late 1970s. Artists such as Donovan and Bread were
Would that have been "Guitar Man"? -- I remember hearing that on TV
as a youngster, I thought it was associated with one of the school
countdown clocks. But it may have been on the testcard/in-vision Ceefax
now I think about it.
I also recall Cat Stevens "Days On The Old School Yard", was that over on
ITV schools? or BBC schools?
--
--------------------------------------+------------------------------------
Mike Brown: mjb[at]pootle.demon.co.uk | http://www.pootle.demon.co.uk/
m
Gosh what memories - I was in Switching in about 1967 when we used to do
regular feeds back to the regions during local news.
Nice to see from Andy that it was still happening into the 90s.
Incidentally, having just worked on the Election coverage with all it's
complex hub system, I remember being in Switching for the 1967 election
and THAT was the London hub.
It consisted of me at a table at the back with the special 25way yaxley
switch that operated the first transistor switching matrix.
Switching sources when phoned by the studio. Simple life!!
(If anyone is still interested, the yaxleys were special manufacture. It
was impossible to get 25way wafers and so they were 24way with a 2way
wafer at the back and a concentric operating spindle - Spose that is
better than the first bodge-up which controlled a solid state matrix
from the old vision levels on the Ledex vision switches)
Mike
PS talking of supervisors, it was MFR who would always wait till after
Magic Roundabout before going to supper but at least sent the engineers
home early.
DMac
>>>>regional news times
Ah bring back Bangor telecine (looked like a 16mm projector pointed at a
wall
Soft as putt-tee-ey! to quote one of our ex TM1s)
And many a young engineer forgetting to come off genlock..... prior to the
switch
...or forgetting to re genlock