TV system line numbers

[Sam]

Why is it that the mechanical scanning systems have an even number of lines, for example Baird's original 30-line and his later 240-line systems, whereas the electronic standards had odd numbers of lines (405, 525, 625, 819(ish))? It doesn't seem to be in any of the books I have built up so far. The mechanical system numbers seem fairly arbitary decisions, but the electronic ones seem a bit odd to just be a number chosen (surely you would have gone for 400 rather than 405 in that case).

Sam

[oldeurope]

Hi Sam, the odd number causes the interlaced picture.
With 404 or 406 or 400 lines you'll get progressive scan.
At the moment there is a big discussion about progressive scan and
Interlace. For digital things it is easyer to make progressive scan,
an even number of lines.
405 441 525 625 gives interlace.

Darius

[ppppenguin]

Darius is correct. With an odd number of lines you get interlace automatically. Each of the 2 fields has a half line in it. So in 625, each field has 312.5 lines. You can have progressive scan with any number of lines, odd or even.

In the early days, frequency division was rather more difficult than now and it simplified things if the number of lines could be broken down into small factors. Hence:

405 = 3*3*3*3*5
625 = 5*5*5*5
525 = 3*5*5*7
819 = 3*3*7*13

[David_Robinson]

But surely you can still have interlace if the number of lines is even. Each field would have the same number of lines. The field sync pulse would be delayed by half a line for the second field, causing the receiver to displace the second field down the screen by half a line and so giving interlace. The only difference from an odd-line system, is that on the second field the field scan is shorter, but a directly triggered field timebase, or a suitably designed flywheel system, will cope.

One of the latest HDTV standards is 1080 active lines, interlaced. I'm not sure if there is a standard for the total number of lines on this system.

[peter_scott]

Originally Posted by David_Robinson

But surely you can still have interlace if the number of lines is even. .

True, but I think you would need to add a small 25 Hz square wave to the frame scan so as to make the displacement between odd and even frames.

As Jeffrey says "With an odd number of lines you get interlace automatically" just because every other (sub) frame commences in the middle of a line and thus a half line pitch further down the frame.

Peter

[ppppenguin]

Originally Posted by David_Robinson

But surely you can still have interlace if the number of lines is even. Each field would have the same number of lines. The field sync pulse would be delayed by half a line for the second field, causing the receiver to displace the second field down the screen by half a line and so giving interlace. The only difference from an odd-line system, is that on the second field the field scan is shorter, but a directly triggered field timebase, or a suitably designed flywheel system, will cope.

One of the latest HDTV standards is 1080 active lines, interlaced. I'm not sure if there is a standard for the total number of lines on this system.

It is possible to interlace with an even number of lines but the field sync pulses will no longer be equally spaced. All the HDTV standards with 1080 active lines have 1125 lines in total. There was a 1250 total line interlaced system that was proposed but never really used.

[oldeurope]

Good evening,
In theory it is possible to get interlace with an even amount of lines
by changing the v pulse space or making every second v pulse longer.
But this is complicated for the receiver.
I tried this when I made the frame reset in my converter.
The main problem is that the V linearity changes a bit if a frame is longer
or shorter. So you don't get interlace over the whole picture.
Conclusion: No interlace with an even amount of lines.

I have an german book written just before 441 started and 180 was in use.
A lot of interesting methodes are discussed there and the british 405 too.
The british 405 interlace system is the simplest and the best methode,
and they used it in the german 441 system too. There is no remarkable
difference between 441 and 405 it is practicly the same.
So you can say the british made the first european HDTV standard.

Kind regards
Darius

[ppppenguin]

The method of interlace used in the 405 line and all later systems was devised in the USA. To be absolutely certain I would have to dig out the published paper but I think it was Kell and Bedford in 1933. Probably in the Procedings of the IRE. This method was available to M-EMI under a patent exchange scheme with RCA and was one of the reasons for accusations (by Baird) of the M-EMI system being non-British.


Kell also established that the apparent vertical resolution was less than might be assumed from the number of lines. This fraction, which can only really be derived from testing with a panel of viewers, has always been known as the Kell factor. Progressive scan systems have a higher Kell factor than interlaced ones so the difference between the vertical resolutions of HDTV at 720p and 1080i is not as great as you might imagine.

[oldeurope]

Hi Jeffrey, this is very interesting.
I read about tripple interlaced mechanical sets and so on.
I made some thaughts about 405 and 441. From the schematics of the
441 sets it seems the 441 system used the equilalizing (right?) pulses.
(german: Vorbereitungsimpulse)
This reduces the amount of visible lines. To get no losses against the
existing british 405 they took the next possible higher amount of lines
and this is 441 because of the available dividers.
So german sets may be a bit difficult to adjust the V hold for good interlace
if used with a british signal.
Like you know it from the dual standard sets. Some dual standard sets had a special interlace adjust .

BTW: Today it is very easy to generate the V sync from the H sync.
Double the H frequency and use a modulo counter like the 4040
to generate the V- pulse. I did it in my camera this way.

Kind regards
Darius

[ppppenguin]

The English term is "equalising pulses". As you know, they were not used on the 405 line system. Blumlein was very much against their use. Although they improved the interlace on poorly designed receivers Blumlein argued that they were not needed. As a result, the 405 line system ended up with a lot of interlace problems because set makers were unable or unwilling to design good sync separators.

The story goes that 405 lines was chosen on a Saturday morning meeting at Blumlein's house. One of the divide by 3 circuits in the 243 line experimental system was changed to divide by 5 and the rest was history.

I don't know why 441 was chosen. It's the next sensible number above 405 (441 = 3*3*7*7) so it's an obvious choice at the limit of the available technology.

[Sam]

Aaah, OK. I had wondered if it was something to do with the interlace, as for the 'even' field the frame fly-back is triggered half-way along line 405 (for the Marconi-EMI system). All the Baird systems were progressive scan (from what I have found, anyway!).

You say the line-number could be broken into a number of small prime-factors. Why was this necessary? Was the higher line-frequency generated by multiplying a lower frequency (which would be more stable to generate), or was the line-frequency divided down to get something else (so the two signals would be in synchronism even if the 'main' frequency drifted)?

Does anyone know of a decent book that goes through this level of stuff - it seems to be a higher-level than the 'first course in' books I have go in to.

Originally Posted by ppppenguin

The method of interlace used in the 405 line and all later systems was devised in the USA. To be absolutely certain I would have to dig out the published paper but I think it was Kell and Bedford in 1933. Probably in the Procedings of the IRE. This method was available to M-EMI under a patent exchange scheme with RCA and was one of the reasons for accusations (by Baird) of the M-EMI system being non-British.

I may have to have a dig around in the Uni library basement to see if they have that tomorrow!

Sam

[ppppenguin]

An example for 405:

Start with master oscillator of twice line frequency = 20250Hz

Divide by 2 for line frequency (10125Hz)

Divide by 3,3,3,3,5 for field frequency (50Hz)

Dividers were typically synchronised blocking oscillators which could could quite easily divide by the wrong number, espcially if that number was moe than 7 or so.

In the early days the field frequency was usually locked to mains to prevent beat effects reduce the visibility of hum.

The other way of generating syncs in the very early days was mechanical. Rather like a tone wheel Hammond organ. I wonder if anyone has any fragments of a mechanical SPG.

Please be aware of the ambiguous term "frame". Nowadays we usually say that 2 interlaced fields make 1 frame. We often used to say that 2 interlaced frames made 1 picture.

There have been experiments with higher order interlace (3:1 or more). I think Baird tried it somehow. The movement and/or vertical artifacts get pretty unpleasant.

I recommend Amos and Birkinshaw's set of 4 books titled "Television Engineering". With a bit of luck the uni library will still have them somewhere. They were used in BBC engineering training and went through several editions.

[Ed_Dinning]

Hi Sam, I think the definative book must be Zworykin & Morton "Television" I have the second edition from 1954 that covers colour systems as well. 1st edition was 1940.

HTH Ed

[Danny]

Hi another good book is basic television by Van Valkenburgh. noooger &neville.inc it covers both the 405 line and 625 line standards also dual standards the sbn 291.39572.4 made and printed by William Clowes & son ltd london,beccles and colchester the book i have is a 1975 reprint .
If you can get any books in this series thay are super for valve stuff .

[tubesrule]

The Sanabria mechanical system in the US was a tripple line interlaced format, although with only 45 lines and a 15 fps rate, it's fairly difficult to say how bad the interlace effects look. It did help cut down on the flicker. I am not aware of any three or more line interlaced electronic systems, but it would be interesting if anyone knows of one.

RCA used mechanical sync generators up through their 240 line experiments. By the time the 343 interlaced experiments started, they where using the blocking osc's as Jeff mentioned.

IIRC didn't the early German 441 system use a mechanical sync generator? I seem to remember hearing that it was some huge size disk spinning at an incredible rate, like upwards of 10,000 RPM. Does anyone else recall this?

Darryl

[Sam]

Hmm, they don't have either of those, but they do have Television, theory and practice by Reyner. I'll fill in a store-request form for it tomorrow.

I will search through the online book-sellers to see if there is any of those books out there in my price-range!

Sam

[ppppenguin]

I had a quick look in abebooks. You should be able to get the Amos and Birkinshaw books for under £10 each.

[oldeurope]

Originally Posted by ppppenguin

The English term is "equalising pulses". As you know, they were not used on the 405 line system. Blumlein was very much against their use. Although they improved the interlace on poorly designed receivers Blumlein argued that they were not needed. As a result, the 405 line system ended up with a lot of interlace problems because set makers were unable or unwilling to design good sync separators.

.

Good morning Jeffrey,
the question is: Do you want an integrator v sync separator or not?
If you want to use one, because of it's advantages, you get interlace
problems without equalising pulses.
If you don't use an integrator you don't need them. Blumelein was right,
look what they made in the later french 819 line system.
The interlace problems took place with the dual standard sets using an
integrator.

Kind regards
Darius

[oldeurope]

Originally Posted by tubesrule

The Sanabria mechanical system in the US was a tripple line interlaced format, although with only 45 lines and a 15 fps rate, it's fairly difficult to say how bad the interlace effects look. It did help cut down on the flicker. I am not aware of any three or more line interlaced electronic systems, but it would be interesting if anyone knows of one.

Darryl

Hi Darryl,
Yes, the output of my standard converter is tripple interlaced with 625
line input("PAL). The output is 416 2/3 and this automaticly generates tripple
interlace.
With a 525 line input("NTSC") it generates a progressive scan picture
the output is 350 lines in this case.
Both signals are visible on a 405 set. The 60Hz causes a little hum inteference
here because of the 50Hz mains. But it is interesting to see.

The tripple interlace makes the V- sync in my converter very easy.
I did not know this when I developed the MK I.
I can take the 625 line V- sync compleatly with all the equalising- and serrationpulses "as is". This makes it much easier.
I wanted to develop a kit "easy conv" but I stopped it because nobody will
build one if they can buy your low cost for 200 Euros

Kind regards
Darius

[oldeurope]

Originally Posted by tubesrule

IIRC didn't the early German 441 system use a mechanical sync generator? I seem to remember hearing that it was some huge size disk spinning at an incredible rate, like upwards of 10,000 RPM. Does anyone else recall this?

Darryl

Yes, of course.
You are talking about the:
Mechanischer Universalabtaster für Personen-,Film-,und Diapositivübertragungen.

mechanical universal scanner for persones, film and Diapositive transmission.

I attached a pic of the Nipkow disc schematic used in this apparatus.
Here some infos about it from Fernseh A.G.
The spinning disc diameter is 720mm and the thickness is 0,2 to 0,3 mm.
It is running in vakuum (5 mm Hg) with 10500 RPM.
This would be a speed of 400m per second.

It is an interesting thing.

Darius