Vertical interval signals on 405

[ppppenguin]

A discussion has arisen eslewhere about test signals in the vertical interval on 405.

The background to VITS (Vertical Interval Test Signals) is that they were used by broadcasters to monitor the transmission chain while programmes are being transmitted. A variety of signals have been used but "pulse and bar" was always included to check freuquecy and phase response. This give an immediate and accurate check by using a special graticule on a waveform monitor. Later systems such as the Tektronix 700 could make autpmatic measurements.

The vertical interval was later also used for teletext.

I don't know how much VITS was used on 405. Certainly the 405 and VITS eras didn't have any great overlap. by the time VITS was introduced 405 was very much a legacy service.

All vertical interval signals are a bit of a nuisance on 405 since many sets have poor vertical blanking. While it's entirely possible to put them on the 405 output of a converter (would need a little bit of spare logic in the FPGA, the actual design would be very simple) I would recommend against doing so. They don't have any sensible use outside a trasnmitter network and can cause visible artifacts.

[ppppenguin]

It's possible for FPGA based converters to generate accurate test patterns. The Aurora SCRF can take in and store an analogue full frame image and display it in the absence of signal. Usually Test Card C for 405 lines. This is only as accurate as the original analogue image. It can also generate a couple of specific test patterns, can't remember what they are.

Hedghog has a computed test card. This is 100% accurate within the limitations of the algorithm used to make it. In principle any FPGA based converter can generate any line based test pattern (same of every active line) but the designer has to provide a means to select it and make choices as to which patterns are available.

[Murphyv310]

Having been a trainee station engineer in British Relay we used the VIT signal for setting up the equipment. On 405 I distinctly remember that there was peak white with a negative going 2.5 mhz pulse followed by at least one other pulse positive going at I believe 2.5 mhz. Then followed by a step. The RT Russell TCD can be programmed with accurate VIT for 405.

[ppppenguin]

Before pulse & bar was invented the usual method of measuring a video system was a frequency sweep. This is OK as far as it goes but tells you nothing about the phase or group delay response. These are critical for video.

Constant group delay is equivalent to linear phase. Video measurements usaully specify group delay as it's more closely related to what you see as P&B errors.

The bar of P&B givea an smplitude reference for the pulse and indicates some low frequency distortions.

L. Weaver of the BBC and NW Lewis of the Post Office invented the 2T pulse and bar test as a quick and easy means of assessing a video path. The original work was published by Lewis in 1954. Proc IEE vol 10 part III no 72 pp258-270. The original P&B was conceived as a full frame test signal. I don't know when the idea of VITS was first suggested. This paper shows how P&B was extended to deal with the requirements of colour TV: https://www.bbc.co.uk/rd/publications/bbc_monograph_58

[Terry]

In Europe these signals are known just as Insertion Test Signal (ITS) because VITS is a swear word in one European language - French, I believe

Trevor, your suggestion that the pulse was at 2.5MHz, if correct, would suggest the British Relay weren't quite as critical in their test methods as the broadcasters.

The pulse is correctly known as a 2T raised cosine pulse. T for a 625 line system is 100nS, so a 2T pulse is 200nS wide at its half way point - it is effective one cycle of a 5MHz sine wave raised so the it is all above black level and sits exactly between black level and peak white. The pulse during the bar is, of course inverted, and extends from peak white to black level.

For 405, T = 166nS so 2T = 333nS - effectively one cycle of a 3MHz signal.

There are no sharp transitions in ITS signals - all of the luminance component passes through the same filter that creates the 2T pulse ensuring that there are no signals above 5MHz (625) or 3MHz (405).

[ppppenguin]

Had no idea that VITS was a French swear word.

In practice the 2T pulse isn't quite symmetrical. This is due to limiations in the feasible filters that were originally used to define the pulse. The photo is p25 of Weaver, "Television Video Transmission Measuements". A very useful book published by Marconi Instruments in 1971. My copy is the 1979 edition. Largely now of historical interest as most of the content is not useful in the context of digital TV. You can get one at a decent price: https://www.amazon.co.uk/Television-vide...B000NZDICE

This is another useful book by Weaver. Again largely of hostorical interest: https://books.google.co.uk/books/about/T...5QAAAAYAAJ&redir_esc=y

The -ve going 2T pulse in the middle of the bar was added later.

A 1T pulse is sometimes used for testing studio equipment where the distortions are often too small to be relaibly seen on a 2T pulse. Unfortunately you can't quite accurately generate a 1T pulse with standard 13.5MHz sampling

[`FRANK.C]

That's all very interesting.
It never occurred to me that 405 had VITS. I had thought of VITS as being more modern than that.

Frank

[ppppenguin]

Not sure that VITS was much used on 405. P&B testing with full screen signals dates from the 1950s but VITS is much later.

I can't see any good reason for us to do VITS on 405. It will give trouble with visible flyback and have no real utility.

[`FRANK.C]

Jeffrey
That makes sense.
Darius built a unit to clean up the vertical blanking period on the video feeding his converter.

Frank

[Murphyv310]

In Europe these signals are known just as Insertion Test Signal (ITS) because VITS is a swear word in one European language - French, I believe

Trevor, your suggestion that the pulse was at 2.5MHz, if correct, would suggest the British Relay weren't quite as critical in their test methods as the broadcasters.

BRW used the offair VIT. I was under the impression that the 405 line pulses were 2.5mHz, I stand corrected if I'm wrong.