SD broadcast video: 704 versus 720 pixels (by 576 pixels)

[NY]

Daft question: why is it that some channels and/or some platforms (satellite
versus terrestrial) broadcast standard-definition as 704x576 and some as
720x576. There are even cases where the same channel is 704 on satellite and
720 on terrestrial (or vice versa). It looks as if the picture area is the
same, but differently scaled, rather than the 704 version being the 720
version with 8 pixels either side being cropped off.

Both numbers are exact multiples of 16 (44x16 versus 45x16): I imagine the
compression algorithms work best with widths and heights that are exact
multiples of 16 (*). But why the two different standards which are so close?
I can understand why some lower-budget channels are sub-SD 544x576 because
that reduces bandwidth and therefore transmission costs. But the saving of
704 rather than 720 is minuscule.


(*) Was it pure chance that the analogue 625/25 format had 575 active lines
which is extremely close to 576 (a multiple of 16)? Was the multiple-of-16
factor relevant when 625/25 was specced - eg to derive the frame frequency
by a cascade of divide-by-2 analogue frequency-division stages from the line
frequency.

[Mark Carver]

On 24/02/2022 21:15, NY wrote:
> But the saving of 704 rather than 720 is minuscule.

I think the BBC changed their DTT channels from 720 to 704, as although
tiny, it liberated enough bandwidth to squeeze local radio into PSB1

[NY]

Fair enough, although since then they've had to make much bigger
economies to fit in BBC Three. I suppose they designed the resolution
and bit rates when BBC Three was present, then benefited from a time
when BBC Three was internet-only and the remaining channels could have
higher bit rates, and have now reverted to the original state.

Reducing 720 to 704 is a saving of 16/720 = 2% - but for several
channels I can see that it would be enough to fit in several radio channels.

[Brian Gaff (Sofa)]

I assume that is why its in mono then. Very few of the stations on radio on
Freeview are stereo. The main bbc ones, Asian network, Classic FM and not
much else it seems.

As for the frame refresh rate, Many early 405 line receivers assumed that
the frame was synchronised to the mains, since when this stopped being the
case many needed extra smoothing in the power supply to stop hum bars
rolling slowly over the screen.
When you say active lines on 625, I'm assuming you mean that sync pulses
and other data like teletext were in the ones not officially seen. In the US
they used 525 and their mains was, and still is 60hz of course, I'd guess
for much the same reasons as our relationship.
History of TV is pretty full of strange decisions.
Why did the 405 line use positive video, as this put the sync pulses at
the lowest signal level and hence more prone to am interference like cars
and vacuum cleaner interference.
Brian

--

This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
bri...@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!
"Mark Carver" <mark....@invalid.invalid> wrote in message
news:j7rhb9...@mid.individual.net...

[Brian Gaff (Sofa)]

Why did they bring back BBC 3? Did the streaming fail, I mean did nobody
watch it. To be honest its full of crap now, but some is probably better
crap than the current crap on bbc1.
All depends on your tastes I'd expect. One assumes they have cut back on
the kiddies channels again then.

Brian

--

This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
bri...@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!

"NY" <m...@privacy.net> wrote in message
news:0cKdnRcT3aYtBIX_...@brightview.co.uk...

[Mark Carver]

On 25/02/2022 10:30, Brian Gaff (Sofa) wrote:
> I assume that is why its in mono then. Very few of the stations on radio on
> Freeview are stereo. The main bbc ones, Asian network, Classic FM and not
> much else it seems.

It's not worth having the radio stations in stereo on DTT. I suspect
they are aimed at consumption based on kitchen, bedroom, and hotel
tellies, where listening is a secondary activity.

> When you say active lines on 625, I'm assuming you mean that sync pulses
> and other data like teletext were in the ones not officially seen.

Blanking consumes about 25+25 lines on analogue.

Digital platforms do not transmit the blanking period, it would be a
pointless waste of bandwidth

> Why did the 405 line use positive video, as this put the sync pulses at
> the lowest signal level and hence more prone to am interference like cars
> and vacuum cleaner interference.

Wasn't the reason negative mod was adopted for 625/UHF because it pushed
the sync pulses (rather than high luminance picture content) into the
non linearity 'zone' of klystrons ?
Anyway negative mod any better though for reception and impulse
rejection; overshoot and all that ?

[Brian Gaff (Sofa)]

Well on the old dual standard set from Redifusion I had it was clear that
the negative video was better and more stable, of course to find out, both
would have had to have been transmitted on vhf. You could however find that
as Europe went 625 negative, except France, that negative was better at
rejecting impulse noise that otherwise gave rise to picture jitter on the
French stations. It was pretty bad on their old 819 line service too, which
we modified the set to handle though the audio needed a second receiver as
did the am sound on 625 of course.
I remember that when I could see, looking at colour TVs in France one
could not really tell it was not negative mod and used Seam colour. You
occasionally heard am interference though on some stations.

Brian

--

This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
bri...@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!
"Mark Carver" <mark....@invalid.invalid> wrote in message

news:j7rqlo...@mid.individual.net...

[Roderick Stewart]

On Fri, 25 Feb 2022 10:30:53 -0000, "Brian Gaff \(Sofa\)"
<bri...@blueyonder.co.uk> wrote:

>Why did the 405 line use positive video, as this put the sync pulses at
>the lowest signal level and hence more prone to am interference like cars
>and vacuum cleaner interference.

And why didn't 405 line use equalising pulses?

Rod.

[NY]

"Mark Carver" <mark....@invalid.invalid> wrote in message

news:j7rqlo...@mid.individual.net...

> Wasn't the reason negative mod was adopted for 625/UHF because it pushed
> the sync pulses (rather than high luminance picture content) into the non
> linearity 'zone' of klystrons ?
> Anyway negative mod any better though for reception and impulse rejection;
> overshoot and all that ?

I'd always understood that the main reason for negative mod was better
rejection of impulse interference from poorly-regulated car and motorbike
(especially motorbike!) ignition systems - probably more of an issue with
spark-gap/coil ignition than with modern electronic ignition.

[NY]

"Mark Carver" <mark....@invalid.invalid> wrote in message

news:j7rqlo...@mid.individual.net...

> On 25/02/2022 10:30, Brian Gaff (Sofa) wrote:
>> I assume that is why its in mono then. Very few of the stations on radio
>> on
>> Freeview are stereo. The main bbc ones, Asian network, Classic FM and not
>> much else it seems.
> It's not worth having the radio stations in stereo on DTT. I suspect they
> are aimed at consumption based on kitchen, bedroom, and hotel tellies,
> where listening is a secondary activity.

Interesting assumption. I don't have any FM radio, apart from a) an
all-in-one hifi system and b) a separate tuner, neither of which I CBA to
set up, especially since the tuner requires a 300 ohm aerial and doesn't
have a telescopic one.

I use DTT or DSat for recording radio programmes (mostly Radio 4), in
exactly the same way as for TV programmes.

[NY]

"Roderick Stewart" <rj...@escapetime.myzen.co.uk> wrote in message
news:9e4k1h9ndt62ghled...@4ax.com...

What was the purpose of equalising pulses? I could never find a really good
explanation.

[MB]

On 26/02/2022 21:37, NY wrote:
> Interesting assumption. I don't have any FM radio, apart from a) an
> all-in-one hifi system and b) a separate tuner, neither of which I CBA to
> set up, especially since the tuner requires a 300 ohm aerial and doesn't
> have a telescopic one.

Never had any problem with just a piece of wire connected to the back of
the VHF FM tuner though use DAB most of the time?

[Ian Jackson]

In message <sve6n4$gl7$2...@dont-email.me>, NY <m...@privacy.invalid> writes

For several reasons, positive modulation is a real PITA. One of them is
that there is no easily-obtainable measurement of the RF signal level.

With negative mod, it's simply sync tip level, which is steady and
constant, and easy to measure with a peak-detecting meter, or as might
be displayed on a spectrum analyser.

With positive mod, it can vary between 30% (black level) and 100% (peak
white). OK, there might be a few lines of VITs, but these are
essentially transient, and make peak level difficult to 'catch'.
--
Ian

[Woody]

If your tuner does not have a 75R FM aerial socket just connect the
aerial with core to one of the 300R terminals and the screen to the
earth (or an earth) terminal - may be one of the AM aerial connections
if the tuner is AM/FM. You will loose a little sensitivity but it will
match well and certainly work.

[Roderick Stewart]

Equalisation between fields of the residual charge on the integrating
capacitor in the field sync separator. Without them, the line sync
pulse that immediately precedes a field sync pulse will precede it by
a whole line on alternate fields, and only half a line on the others.
The timing of the start of field scan depends on the voltage on the
integrating capacitor, and if it's not equal between field scans, it
can trigger alternately late and early, and cause the lines displayed
on the screen to bunch in pairs. ("Early" translates to up, and "late"
to down). Using two or three lines of half width pulses twice as often
doesn't alter the average signal voltage, so it won't interfere with
anything else, but it gives enough time for the capacitor to discharge
to a level that doesn't matter, so when the broad pulses begin, the
integrating circuit has the same initial conditions on every field.

There are other ways of detecting field sync, but most early TVs used
fairly simple circuitry (essentially a resistor and a capacitor) that
didn't always work perfectly. The extra complication of adding
equalising pulses in the broadcasters' sync pulse generators obviates
the need for extra care or complication in millions of receivers, so
you might expect that as soon as the idea had been thought of it would
have been implemented everywhere, but our 405 line system never did.

Rod.

[Roderick Stewart]

On Sun, 27 Feb 2022 08:52:43 +0000, Ian Jackson
<ianREMOVET...@g3ohx.co.uk> wrote:

>For several reasons, positive modulation is a real PITA. One of them is
>that there is no easily-obtainable measurement of the RF signal level.

You can sample the black level on the back porch (as some of the
better homemade receivers did to obtain an AGC voltage), but you only
get a measure of 30% of peak, which is not as good as a 100% reference
that can be sampled with simpler circuitry. Designers of domestic
electronics being cheapskates is not a new thing.

Rod.

[Ian Jackson]

In message <t0km1htg218ft7suu...@4ax.com>, Roderick
Stewart <rj...@escapetime.myzen.co.uk> writes

There were also some commercial sets that did have a 'proper' gated
black-level AGC system and black-level clamp (especially in the earlier
days), but an awful lot had that ghastly mean-level AGC, with AC-coupled
video drive to the CRT. [Around 1967 the Wireless World carried an
article on how to do a simple add-on modification to at least stabilise
black level. I did the mod to my Ferguson 1500, and it did sort-of
work.]
--
Ian

[Brian Gaff (Sofa)]

Yes well another oddity of old black and white sets, not the early ones but
those around the time of dual standard was the AC coupled video that made
blacks lighter when not much white was on the screen.
I did modify one with a dc restore circuit once, but then found I needed to
modify the eht to stop the regulation affecting picture size and focus when
you got a sudden flash from something.

Brian

--

This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
bri...@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!

"Roderick Stewart" <rj...@escapetime.myzen.co.uk> wrote in message

news:giim1hlosd2gi4ifd...@4ax.com...

[Brian Gregory]

On 25/02/2022 10:30, Brian Gaff (Sofa) wrote:

> Why did the 405 line use positive video, as this put the sync pulses at
> the lowest signal level and hence more prone to am interference like cars
> and vacuum cleaner interference.

I assume they simply made the decision too soon, before they knew all
the full implications of it.

--
Brian Gregory (in England).

[Mark Carver]

Indeed, there weren't that many vehicles around in 1936. Although those
that were had absolutely no electrical suppression.

In fact, wasn't ignition  suppression mandated in the 1950s, when both
cars and TVs increased in number and it was all becoming a nuisance ?!

[Brian Gregory]

On 24/02/2022 21:15, NY wrote:

> Daft question: why is it that some channels and/or some platforms
> (satellite versus terrestrial) broadcast standard-definition as 704x576
> and some as 720x576. There are even cases where the same channel is 704
> on satellite and 720 on terrestrial (or vice versa). It looks as if the
> picture area is the same, but differently scaled, rather than the 704
> version being the 720 version with 8 pixels either side being cropped off.

It's worse than that. There are three different formats.

1) 704x576, all pixels used, scales to 16:9 with correct aspect ratio.

2) 720x576, only the middle 704 pixels are needed to make 16:9 in the
exact correct aspect ratio. The extra 8 pixels each side may have
picture info in them from just outside the 16:9 shaped picture, or they
may be black.

3) 720x576, all pixels used, scales to 16:9 with correct aspect ratio.

[Brian Gregory]

On 01/03/2022 15:58, Brian Gregory wrote:
> It's worse than that. There are three different formats.
>
> 1) 704x576, all pixels used, scales to 16:9 with correct aspect ratio.
>
> 2) 720x576, only the middle 704 pixels are needed to make 16:9 in the
> exact correct aspect ratio. The extra 8 pixels each side may have
> picture info in them from just outside the 16:9 shaped picture, or they
> may be black.
>
> 3) 720x576, all pixels used, scales to 16:9 with correct aspect ratio.

Oh and it's worth mentioning that there is the same mess with 544x576
and 528x576.

[Robin]

yes; well remembered :)

--
Robin
reply-to address is (intended to be) valid

[Indy Jess John]

On 01/03/2022 15:23, Mark Carver wrote:

As someone who was driving 1950s cars in the 1960s, I can confirm that
the early cars had copper conductors in the HT leads, and that would
have radiated the spark current. Later 1950s cars had HT leads with
carbon cores and they didn't create interference. One early car I
bought had copper cored leads terminated in "antiference" spark plug
connectors and that didn't interfere with TVs either.

I never explored what was in the spark plug connectors because they were
a sealed unit and I couldn't see what was inside without wrecking it,
and I had no idea where I might find replacements.

Jim

[NY]

"Brian Gregory" <void-invalid...@email.invalid> wrote in message
news:j86u9b...@mid.individual.net...

I've never heard of 528x576. Is that used for any sub-SD channels on DVB-T
or DVB-S in the UK? Or is it more commonly found outside the UK? Unlike
704/720 full-SD, I've only ever seen 544 sub-SD.



Thinking of "[the pixels] may have picture info in them [...] or they may be
black", why is it that some SD broadcasts (particularly BBC regional news)
have a black or white half-line at the top right of the picture? I could
understand a black half-line at the top left or the bottom right of an
archive programme made in analogue, but why would modern all-digital
production introduce a half-line?

[Brian Gregory]

On 01/03/2022 20:51, NY wrote:
> I've never heard of 528x576. Is that used for any sub-SD channels on
> DVB-T or DVB-S in the UK? Or is it more commonly found outside the UK?
> Unlike 704/720 full-SD, I've only ever seen 544 sub-SD.

"That's TV (UK)" Freeview channel 91 is 528x576.

The satellite version of it seems to be the same 528x576 padded with
black out to 544x576.

[Brian Gaff (Sofa)]

Who knows the logic of all this. Back in the early days of flat displays,
some screens had oblong pixels which made the setting up of a picture on
them rather interesting.

Brian

--

This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...
bri...@blueyonder.co.uk
Blind user, so no pictures please
Note this Signature is meaningless.!

"Brian Gregory" <void-invalid...@email.invalid> wrote in message

news:j87l20...@mid.individual.net...

[Mark Carver]

I'm struggling to remember the death of Queen Victoria however

[NY]

"Brian Gaff (Sofa)" <bri...@blueyonder.co.uk> wrote in message
news:svn6up$3k8$1...@dont-email.me...

> Who knows the logic of all this. Back in the early days of flat displays,
> some screens had oblong pixels which made the setting up of a picture on
> them rather interesting.

SD, especially in widescreen, has very oblong pixels. For 576 rows of
pixels, you'd expect either 4/3 * 576 = 768 or else 16/9 * 576 = 1024 pixels
across the width. I believe the reduction from 768 to 720 is because of the
Kell Factor in analogue days: an interlaced analogue picture looks less
sharp than a non-interlaced one (assuming you're not distracted by the extra
flicker of non-interlaced) and they reduced the bandwidth of the signal to
give comparable resolution in the horizontal direction to the perceived
resolution in the vertical direction. And then that carried over into
digital standards and widescreen stretched everything out even further. In
contrast, HD is 1920x1080 which is 16/9, so that has square pixels.

[NY]

"Brian Gregory" <void-invalid...@email.invalid> wrote in message

news:j87l20...@mid.individual.net...

Ah, I can't get That's TV on terrestrial because the signal is too weak
where I live, due to a combination of factors: distance, a hill in the way,
wrong direction (it's aimed at Hull or Grimsby, rather than north to
Bridlington), possibly wrong aerial (*), possibly different propagation and
cable attenuation with higher frequency. I have occasionally received the
LOC multiplex when reception conditions have been abnormal: for That's TV, a
TSReader scan shows 128 kb/s audio but the video stream could not be
decoded.


(*) I suspect my aerial dates from analogue days, maybe even before Channel
4, when everything was lower than 530 MHz. Certainly COM4 (546 MHz) is
noticeably worse reception than COM 5 (490) or COM6 (514) even though all
three are almost the same power, and COM7 (746 MHz) is undetectable (though
the lower power is probably a big cause too). Maybe one day I'll get it
looked at, though I tend to use satellite for most TV, only using
terrestrial as a fall-back if there are two or more things I want to record
at the same time.

[NY]

"Mark Carver" <mark....@invalid.invalid> wrote in message

news:j88og1...@mid.individual.net...

Was unsuppressed ignition only a problem for TV? Did it not cause problems
with radio reception, especially VHF which is more similar to 405-line TV?
Or is VHF radio less susceptible because it is FM rather than AM/VSB?

[Mark Carver]

I recall being driven mad by next door's electric kitchen knife when
trying to listen to FM radio,  it didn't affect MF/LF reception at all.

Also didn't some early (1980s vintage) FM car radios have impulse noise
filters (by use of a delay to mute the audio when there was a pulse ?)
Seem to recall the same devices were available for record decks too ?

It was far more of a problem at the bottom end of UHF Band IV, than at
the top of Band V too

[Brian Gregory]

On 02/03/2022 09:54, NY wrote:
> Ah, I can't get That's TV on terrestrial because the signal is too weak
> where I live, due to a combination of factors: distance, a hill in the

>...

It's marginal here. On my aerial some receivers can manage it most of
the time, others can't.

[Rink]

Op 27-2-2022 om 9:52 schreef Ian Jackson:

And positive modulation needs more power, because almost the whole line
is between 30% and 100%.
While with negative modulation almost the whole line is between 0% and 70%.
Only the syncs are at 100%.

[Ian Jackson]

In message <t005cj$phh$1...@dont-email.me>, Rink
<rink.hof.ha...@planet.nl> writes

No. With PAL I and B/G, with -ve mod sync tip is 100% of envelope
amplitude, black is 76% (-2.4dB) and white is 20% (-14dB). Fully
saturated yellow takes the downward excursion down to 12.5% (leaving
just enough to facilitate the inter-carrier mix to give a sound IF of 6
(5.5 for B/G) MHz. [NTSC is very similar, but slightly different.]
--
Ian