Calibration and accuracy of old test cards

[high_vacuum_house]

Good afternoon,

It occured to me earlier whilst out walking how accurate were some of the old test cards.
With electronically generated ones they should be fairly accurate but with the older monochrome and early colour ones where a camera is pointing at a physical test card on a stand the accuracy could depend on many things.

Such as brightness of light, position of lights around test card, Light spectrum for colour test cards, camera position, quality of camera tube and associated video processing electronics, camera tube scanning circuit linearity and many other variable things.

Were such things as the frequency gratings around the sides of various test cards actually what they say they were on reception on a receiver, and the colours around the edge of test card F actually the correct colour and brightness?

It would be interesting to know if the old 405 line and early colour test cards were subject to any realistic calibration check against a master reference test card set up before transmission?

Many thanks,
Christopher Capener

[murphyv310]

Indeed test cards were occasionally what they say on the box! Monoscopes were mostly used and were very accurate.

Colour cards were not done on monoscopes.

[ppppenguin]

I know that the greatest care was taken with Test Card F slides to get accurate results. I believe this included doing the colour and monochrome parts separately. I assume equal care was taken by the BBC with earlier cards.

Obviously you also had to align the slide scanner accurately and ultimately this could only be done by recourse to waveform and vector monitors.

For very early test cards, using a camera pointing at a card, I think the results were likely to have been as good as you might expect. In other words just about good enough.

[gingpeakin]

As I remember the grey scale test chart most commonly used in studios (and occasionally on OBs when lit) was all that we used. The white was then known and the black was always OK because it was the Gregory's hole (someone else will need to tell me why this was so called) was as black as anything. It was a recess in the chart that was lined with black felt. Everything else fell into place because we knew that the chart was calibrated etc. They weren't cheap.

Ging

[ppppenguin]

I've been told that it's very difficult to make accurate sinusoidal frequency gratings using photographic or printing processes. On some test cards there was no real attempt and they were more or less square wave.

[neon indicator]

Yes, older ones where essentially squarewave.

Cameras of course used lightbox/large slide testcards for alignment. The colour cameras of 1970s very time consuming to adjust, Scope with feature to "dial in" a particular line number and vector scopes used. Two main used: Colour bars to set grey scale and colour and regular test card for general alignment.

Even though I worked in Comms I was trained on how to set up camera and they had me "do" the news studio cameras a few times, no doubt to keep my hand in, in case I was needed to "fill in" or something.

A lot more time spent with colour bars than regular test card, on cameras and monitors. For monitors of course an electronic Colour Bar generator used.

[ppppenguin]

Creating a real life accurate physical version of colour bars is very difficult. I was at BBC Research Dept in the 1970s when they were working to create such a thing. ISTR it involved a number of strips of fabric sewn together. I have no idea if it ever got used in service.

This thread seems to be addressing 2 different questions:

1. A radiated test card, intended for setting up receivers. In this case the accuracy of the physical image and that of the slide scanner or camera are both important. Also used internally for setting up monitors but here a variety of electronic patterns such as bars and grille could also be used.
2. A physical test chart, intended for setting up cameras. In this case only the accuracy of the physical image is relevant.

[neon indicator]

(2) The Colour bars for camera in 1976 or so was a light box. I don't know what the coloured bars where made of.

Later I worked with small non-broadcast studios and also colour cameras for recording surgery in operating theatres. Camera alignment on those cheaper often colour stripe single tubes was a horror in late 1970s (Saticons, Nuvicons, Vidicons?)

My now aged Sony Digital 8 single CCD WS camera never needs any alignment and seems amazing for single CCD compared to the 1970s 3 tube Plumbicon camera. There was an old Mono Orthicon used in corner of a Radio Continuity "room" (very small) used in mid 1970s for the BBC NI spinning world illuminated B&W globe (colour added electronically). I don't remember it ever needing the constant attention the colour cameras needed.

(1) When did BBC switch to Electronic Testcard for broadcast rather than pointing camera at a light box with a large slide?

I presume maybe some testcard used slide and flying spot scanner like a static telecine? I never saw or heard of such a device, but I imagine much more accurate, especially for colour, than pointing a camera at a slide? I know there were "slide scanners" for transmission of regular photographs, though I don't remember seeing one.

[ppppenguin]

My understanding is that TCF was a slide, normally in a Cintel flying spot scanner. This should have been capable of very good results.

The first proper (ie full broadcast grade) electronic testcard was the Philips PM5544. http://en.wikipedia.org/wiki/Philips_PM5544 I think it used magnetic core for part of its storage. The BBC's expert on electronic testcard generation was Richard Russell, also well known for his pinoneering work for the BBC Micorcomputer. He was at DD and I think he had an EPROM based test card working in the late 1970s. As EPROMs became bigger this became cheaper and easier.

PS: Lots more TC info here:
http://www.pembers.freeserve.co.uk/T...al.html#PM5544

This page also confirms my opinion that the colour and mono parts of TCF were separate images.

The original Test Card F comprised two transparencies - one colour and one black and white - sandwiched into a 35mm slide mount.

[brianc]

During my time at the Beeb, I was in the Studio section of the Planning and Installation Department, and "colourised" two of the main studios at TVC (TC3 & TC4. I also worked in OB department but I never ever saw a colour-bar test chart. There were two charts used on a regular basis which were the greyscale already mentioned with its "super black" Gregory hole and the flesh tone chart with the lady in a smock. These were both made by Royale and were indeed very expensive. I knew an American who printed testcharts and he mixed the paint for the greyscale and measured and adjusted the reflectivity for each step and then silk-screened each step onto the chart. I should imagine this was how Royale made theirs too. To manufacture a colour-bar test chart, great care would have to be taken with the colour temperature of the illumination for the pictures to have any sort of meaning, I would have thought.
There was a monochrome chart in use with a couple and a boat - dim and distant memories here! but this chart was used for testing individual camera tubes.

[Patrick Dixon]

One of the products that I did some design work on was the BAL 'Matchman'. The original idea was conceived by us at the ITCA Lab at Thames TV and the design got licensed to BAL to manufacture. Some years later I did a design update on it for BAL.

IIRC the idea was that it overlayed digitally generated patches (or Colour Bars) onto a camera output, so that the camera could be lined up/checked by eye. You needed a Macbeth Colour chart to point the camera at, and for the illumination to be accurate. It also allowed you to capture the colours from an aligned camera, to be compared to another camera (or the same camera later on).

Here's a shot I took of it when I was doing the update (would have been around 1988-ish).

http://www.yatecourtbarn.adsl24.co.u...20Matchman.jpg