System A Front Porch Timing

[SwanseaSteve]

This is a bit trivial, but does anyone know the aforesaid timing value please?

The best I know is that the line-sync pulse is 9us wide (plus or minus 1.0us), and that the timing from line-sync datum to the end of line-blanking is 16.5us (plus or minus 0.5us). Looking at the scale of the front porch as drawn on the diagrams that I've seen, I'd say it was about 0.5us to 0.75us.

Can anyone do better please?
Steve

[FERNSEH]

Hi Steve,
I set the front porch to 1 microsecond in my own build standards converter.
The line sync pulse and the back porch are 9 microseconds each. That leaves slightly less than 80 microseconds for the active video.

DFWB.

[peter_scott]

http://www.scottpeter.pwp.blueyonder...0line%20System

Sorry the fractions aren't very readable but are as follows:

Line front porch (not highlighted on diag) 1/200 line period
Sync width 1/10
back porch 1/20
Signal 17/20

Line pulses in frame syncs:
4/10 : 1/10 4/10 etc

Peter

[FERNSEH]

In the original 405 line system, that is 1936 to say 1964/5 and perhaps as late as 1969 the active video period was 83 microseconds.
The converted 405 waveform has a active video of about 79 to 80 microseconds. The reason for the slightly less 405 active video time is because the standards converters had to accommodate the characteristics of the 625 line system.
Way back in the mid sixties I noticed that many older sets which were initially adjusted so that the picture just filled the mask displayed slightly less picture width. I never realised at the time this was a result of the converted 405 video waveform.

DFWB.

[SwanseaSteve]

Originally Posted by peter_scott

Sorry the fractions aren't very readable but are as follows:

That's the original 1934 draft spec - I've never seen that before. Very interesting, but evidently the BBC weren't strictly using those timings even by 1962. If you look at the attachment in https://www.vintage-radio.net/forum/...073#post623073 from 'Synchrodyne' in a different thread, you can see an "EID" containing what the BBC considered "standard" before they added experimental NTSC color encoding on top.

Then we have:

Originally Posted by FERNSEH

In the original 405 line system, that is 1936 to say 1964/5 and perhaps as late as 1969 the active video period was 83 microseconds.

Assuming this is true, and that the BBC EID's value of 16.5us from sync-datum to end of back porch is true, then the front porch would have to be 98.765 - 83 - 16.5us (i.e. about -0.7us!!). Which is obviously wrong!

I think Fernseh has mis-remembered that 83us value for the visible line-scan time. It can obviously not be less than 82us, and 81.5us would give 0.765us for the front porch.

Originally Posted by FERNSEH

The converted 405 waveform has a active video of about 79 to 80 microseconds. The reason for the slightly less 405 active video time is because the standards converters had to accommodate the characteristics of the 625 line system.
Way back in the mid sixties I noticed that many older sets which were initially adjusted so that the picture just filled the mask displayed slightly less picture width. I never realised at the time this was a result of the converted 405 video waveform.

That's interesting. It would rather imply that the 625->405 converters of the day resampled the whole 625-line system's scan-line (64us) and output it 98.765us long for the 405-line system, blanking and all. But they must have done something to stretch the line-sync pulse in all of that, because just re-timing the 625-line line sync pulse (which starts off at 4.7us) would result in an appx 7us line-sync pulse which is a bit short for the 405-line system (should be 9us).

And then again - maybe they didn't bother. Tests might have shown that every 405-line TV in existence at the time worked happily with 7us line-pulses. This could easily be checked though: there are several original BBC converters in the hands of enthusiasts, some of them are even members of this forum. It wouldn't take long with a 'scope.

Steve

[FERNSEH]

Hi Steve,
625 > 405 converters generate their own system A type sync pulses.
The 625 syncs are stripped out before conversion. We're only interested in the active video portion.
The duration of a 405 line is 98.8 microseconds. Front porch 1uS, Line sync, 9uS, Back porch 9uS. Active video 79.8uS. That's the figures in my standards converter and I've used them since 1985.
The relationship between the the two systems is 81/125.

DFWB.

[peter_scott]

Originally Posted by SwanseaSteve

the BBC weren't strictly using those timings even by 1962. If you look at the attachment in https://www.vintage-radio.net/forum/...073#post623073 from 'Synchrodyne' in a different thread, you can see an "EID" containing what the BBC considered "standard" before they added experimental NTSC color encoding on top.

Hi Steve,

I didn't see a front porch spec in the EID but it just looks as if they compromised the sync pulse to allow for the burst signal. Looking in Hawker's "Television Engineer's Pocket Book" of 1962 he quotes the same ratios as I quote earlier.

Amos & Birkinshaw in "Television Engineering" 1963 quote different figures, namely:
Front 1.5 to 2.0 us
Sync pulse 8 to 10 us
Overall 17.5 to 19 us (Front + sync + back)

Peter

[emeritus]

The attached is an extract from a booklet distributed at Marconi's 1954 demonstration of their implementation of 405 line NTSC colour TV. Fig 6 shows the synch and colour burst waveforms. The paragraph commencing at the foot of page 28 explains the modifications that had been made to the contemporary monochrome 405 line waveform used by the BBC

[SwanseaSteve]

Originally Posted by FERNSEH

Front porch 1uS, Line sync, 9uS, Back porch 9uS. Active video 79.8uS. That's the figures in my standards converter and I've used them since 1985.

No-one's knocking you nor your timings, Fernseh. Indeed, it would seem that a typical 405-line TV will work perfectly well even with some wild variations to the so-called standard timings! I was just wondering what the "standard timings" were supposed to be, and couldn't ever seem to nail down the front porch.

If the NTSC disgrams are remotely to scale then I would say that the front porch was roughly 0.75us to 1.0us by comparison with the stated 1.5us gap from the rising edge of the sync pulse to the start of the burst.

Oh, BTW - extra street cred for having managed to build a 625-405 standards converter in 1985 (at all) and still having it running now is just showing off!!

[Doctor Hue]

Originally Posted by FERNSEH

...The reason for the slightly less 405 active video time is because the standards converters had to accommodate the characteristics of the 625 line system.

I'm not sure what I'm missing here. Could you elaborate please. I'm still trying to figure out exactly what active line length I need for my pc output.
The original 1938 ref I have seems to be 169/200 ( 83.456uS ) but the 1961
colour specs are about 81.77uS. How does 625 conversion create 79 to 80 ?

[Synchrodyne]

Referring to the original question, some timings are available in the atached item from Wireless World 195208, although I suspect that they might add to the apparent confusion rather than clarify.

Cheers,

[Synchrodyne]

And some more numbers from Kerkhof & Werner (1952).

Cheers,

[Doctor Hue]

Originally Posted by Synchrodyne

although I suspect that they might add to the apparent confusion rather than clarify.
Cheers,

Thanks very much for those. The WW CCIR shows a back porch of 6 to 9 uS
- rather a large variation !

The original 1936 spec called for "at least 10 lines of vertical blanking"

When was 14 lines standardised ?

I still don't understand FERNSEH's point:
" the standards converters had to accommodate the characteristics of the 625 line system " that caused the active line to drop to " 79-80uS ".

[Doctor Hue]

Originally Posted by Doctor Hue

I still don't understand FERNSEH's point:
" the standards converters had to accommodate the characteristics of the 625 line system " that caused the active line to drop to " 79-80uS ".

I think I've figured it out now. The ratio of the active line lengths to the total
lengths of each standard ( 405 & 625 ). 405 is around 84.5% ( 1938 )
625 is around 81.2 % which would translate to about 80.2 uS in 405. Seems a shame to 'waste' approx 3.2 uS but makes the converters of the time easier.

[SwanseaSteve]

Thanks, Synchrodyne for that very useful WW scan. I think we can assume the data to be pretty much "correct" for 405 lines at the height of its ascendency during the 50's, but notice that the frame-sync timing on 625 lines was different by the time the UK adopted it in 1964. By then, there were 5 equalisation pulses either side of 5 broad pulses. And the odd/even frame labelling (on 625) had been reversed too.

Full marks to the French 819 line system for having the worst frame sync signal in all history!!

[Synchrodyne]

I am fairly sure that the change in the 625-line waveform, to use 5 rather than 6 equalizing pulses was made soon after that WW article was published, and for example before the Belgian Systems C (625-lines) and F (819-lines) variants came into use in September, 1953. That the Belgians chose the 625-line pattern for their 819-line synchronizing signal (albeit with 7 equalizing pulses) would seem to argue the affirmative on your proposition.

I cannot right now find definitive information on the 625-line synchronizing signal change, but a 1955-published Philips book simply stated that the 625-line system used 5 equalizing pulses whereas the American 525-line system used 6. It does provide some clarifying comments that aren’t really, but it will be easier if I scan and post the couple or so pages rather than paraphrase.

One imagines that the CCIR change from 6 to 5 equalizing pulses was also reflected into the Russian (OIRT) 625-line system soon thereafter, although the timing of that might be tough to track down. The Russians did all of the early development on 625-lines, based upon the NTSC 525-line work. But it seems that back in the late 1940s and early 1950s, the Russian origins of 625/50 were not mentioned in polite company. One needs to use deduction to work out what the Gerber (compromise) was all about.

As far as I know, the Argentinean 625-line version (System N) retained 6 equalizing pulses. This was essentially the European 625-line system squashed into a 6 MHz NTSC channel, with NTSC detail parameters in the sound channel and vis-à-vis vision signal levels, and used since mid-1951, in fact before many of the European System B countries started regular broadcasts. Apparently the initial development work was done by and/or for the Japanese, who with a mix of 50 and 60 Hz electricity supplies could have opted for either 625/50 or 525/60, and where there was a debate as to whether to use 6 or 7 MHz channels. Whilst the exact details of the deliberations are unknown, one may read between the lines and see that 6 MHz channels were favoured (more of them in a given spectrum allocation) but that 625/50 proponents did not want that to exclude their system. It seems not unlikely that the system choice involved other than technical considerations, as well. If anything, System N was less mentioned in the literature of the 1950s and 1960s than System D.

Cheers,

[Synchrodyne]

The footnote on p.129 of the attached excerpt from W.A. Holm, "How Television Works", Philips Technical Library provides a brief comment on the choice between 5 and 6 pulses.

Cheers,

[Synchrodyne]

And some more waveforms, this time from G.H. Hutson, "Television Receiver Theory, Part 1", Edward Arnold, 1966.

[Doctor Hue]

Originally Posted by Synchrodyne

The Russians did all of the early development on 625-lines, based upon the NTSC 525-line work. But it seems that back in the late 1940s and early 1950s, the Russian origins of 625/50 were not mentioned in polite company. One needs to use deduction to work out what the Gerber (compromise) was all about.

Found this in EBU Tech Review 1996

[FERNSEH]

That's right the active video part of a line of the 625 standard is 81.5%.
In the original 405 system it was approximately 84.5%. When I set about to design my standards converter in late 1984 it soon became clear that the active video portion of the converted waveform would be 79 - 80 microseconds. The clock pulse generator frequency is common to both standards. Separate divider chains are employed to create the 15,625HZ reference for the PLL and the twice line rate of 20,250Hz for 405 lines.
The number of samples during the active video part was 526 on both standards and this resulted in the fact the 405 active video also occupied 81.5% of the 98microsecond total line period.
A higher master clock frequency of 12,656,250Hz will result in 658 samples during the active video portion.
There is a clock pulse generator for 405 lines and this references to the 10.125Hz rate generated from the divider chain supplied from the 625 clock generator. So it follows both standards have a common reference. This is essential to maintain good interlace because the 405 frame sync pulse train is actually triggered by the input standard frame sync.

DFWB.