Measure seven times

By Eugene Brudno

Review contents:

Measure seven times Analog TV transmitters’ parameters measure Are You Producing Legal and Valid Video? “Beta TV Com” measuring equipment Measuring devices for terrestrial and cable television from Broadcast Arsenal Group of companies Measuring TV equipment from IMOS Company TV signals level measurers Measuring tools of Planar Company RAP-TV digital measuring system Kathrein TV signals measuring tools Leader Electronics measuring equipment for broadcast sphere Promax measuring equipment Recom-10 TV signal meter Sunrise Telecom’s measuring equipment for cable TV network analysis New devices from Tektronix Equipment from Wohler Company

At the very beginning I’d like to beg a pardon to attentive reader for allowing myself speaking also on measure theme. an expression can be created that the author is a kind of all-know-person and he writes about studios, OB vans and OB cases and here also about measures. So he is boring!

Really the author is not all-know-person and is far from a thought to teach someone. But many years of work on TV and some observations of what is going on in its technical shell lead to thoughts which should be shared with colleagues. And if colleagues react it will be just fine! Otherwise what the industry journal exists for?

Absolutely occasionally I became an evident of two young people talking which work in TV production sphere:

— It would be good to tune the camera — said one of them — it greens.

— Hey you — replied second guy — they are adjusted for whole working term!

And it was told for sure.

This sentence amazed me and enforced to think about what really happens in our industry? At my view degradation of technical culture happens among those people whom we used call exploiters. Even being strange it is actively assisted by two circumstances: perfection and reliability of equipment and digital technology intervention.

First one is pretty evident. Modern equipment has very high reliability and needs minimal intervention of technical personnel. And even this intervention is necessary it happens most often via menu totally excluding direct intervention into device’s operation. It very simplifies user’s work and cultivates totally customer’ approach. Whole user’ isolation from both physical principles which equipment is based on and physical processes undergoing inside it happens.

Digital equipment aggravates this circumstance. A man has a deal not with signals directly reflecting physical processes but with abstract codes of these signals, e. g. goes even far from physics.

I imagine what can be said in contra to this reason and which theses can be given in support of all described above. I even agree with these theses. Outgoing from evident analogy it can be said: Yes, there are drivers and there are mechanics. So these mechanics an article is addressed to.

The question is pretty legal: Which relation all this have to measures in television? Lets observe measure’ place in TV production.

First of all, what the measure is and what it is necessary for? Measure if a fixation of some being’s (signal’s) parameters and comparison of parameters fixed with defined etalon. Further actions related to correlation or deviations of fixed parameter with etalon are not a subject of present article’s theme.

Pay your attention to that the word “being” was used. Being used with analog environment this definition covers absolutely everything which is going on in production process. In analog environment we always have a deal with a signal which is monosemantically connected with a picture and fully accords to it. I don’t know whether it is possible to illustrate this statement by a fact that the frame waveform diagram of some optical images sometimes totally repeats these images. Those who work a lot with TV waveform diagrams with no doubts observed such “picture” on the waveform monitor’s screen. Or on the contrary. That’s why this signal is called “image signal”. Image signal is present at any point of “light to light” path, e. g. it appears immediately after light stream from object being shot conversion inside a camera and directly precedes to display’ light stream creation.

Whatever happens to signal, whichever conversions it goes through its parameters always reflect image’s parameters and accord to them. The only exclusion is signal’ modulation by different sub-careers when coding according to any color TV system and during video recording. Let’s memorize this circumstance.

Video signals parameters according to GOST 7845-92

Blanking and synch line pulses

Blanking, synch and equalizing fields pulses

H — nominal line duration; a — blanking line interval duration; b — interval between On line start and blanking line interval fall; c — interval between fronts of blanking and synch line pulses; d — line synch pulse duration; e — front (fall) blanking line interval; f — front (fall) line synch pulse duration; i — interval from synch line pulse fron (On line start) to the start of non-modulated color sub-careers (to the end of chroma signal blanking); j — field blanking pulse duration; k — interval between fronts of fields blanking pulse and first equalizing pulse; l — duration of first sequence of equalizing pulses; m — duration of sequence of fields synch pulses; n — duration of second sequence of equalizing pulses; p — equalizing pulse duration; q — field synch pulse duration; r — interval between neighbor fields synch pulses; s — duration of front (fall) of fields synch pulses and equalizing pulse.

What follows from that? The following: passing of image signal through any circuits and any devices inevitably leads to signal distortion. Upon level and nature of these distortions it is easy to judge how they’ll impact an image. For example, frequency response’ fall or rise will cause image sharpness decrease or increase.

Because any TV path consists of numerous separate links each of which commits its own distortions than in analog environment constant monitoring of signal’ parameters is needed at every path’s intersection. It forms one of the most important duties for exploitation personnel. Looking to how measure system is organized and being realized one can easy judge about TV production level, personnel’ qualification and technological discipline.

With widely digital technologies adoption the new situation has appeared. Its essence is in following. Image signal as direct accordance to an image doesn’t exist anymore. It is replaced by the code as digital signal is the result of image signal coding. In other words the signal’ nature and its physical characteristics have changed dramatically: value, phase, frequency and timing characteristics which don’t carry information about image characteristics anymore. That’s why distortions of digital signal passing through paths practically don’t impact to image’ brightness. When pulses value is over ADC threshold limit the picture exists and when this value is under limit it doesn’t.

Degradation of path’s frequency response absolutely doesn’t impact to picture till response degrade so much that increase of pulses duration in this case will lead to their overlay and ADC’ fault. E. g. picture’ sharpness doesn’t depend on path’s frequency response at all.

So getting back to measures defining and in relatively to digital environment it is necessary to talk about process’ parameters measure. A-to-D and D-to-A processes are mentioned. And saying strictly it is already not the measures in common sense but detection of particular digital signal’ elements presence which confirm process correctness. These elements are any kinds of control sums, control bits etc. In principle, in digital environment totally different control system is available and must be used and simplest form of it can be (please don’t anger!) control lamps set. If lamp lights than the control bit is present, if lamp off — it doesn’t.

Experienced colleague will immediately say that for analog environment there were error range control systems which work upon the same principle. Whey really were and not only in TV production. But it is not more than particular tool which doesn’t replace measures.

So what we have? It seems like with “massive” as it is used now “digitization” (what a word!) TV signals measures are not necessary due to absence of these signals themselves. No, it is totally incorrect. Just the sphere where traditional measures needed is being narrowed. For example, end-to-end measures of digital signals paths became senseless. Because what transmission path is? It is end pulse amplifier (digital distribution amplifier), cable and receiver usually with re-clocker. What impacts signal’s parameters in such path? What to measure? Probably the pulse’ amplitude only.

Let’s go through the path “from light to light” and try to investigate what and where have to be measured.

The first and only links in terms of modern technology are image sensors. These mostly are cameras converting light from object being shot into image signal. These also are telecines. But in modern TV they become mostly exotic. Film is very rarely used in TV production where it is necessary to transfer and image from film to video tape. But someone and somewhere transfer a movie! So let’s categorize telecines as image sensors.

So let’s talk about cameras. Modern cameras create for young generation of technical staff and camera men an illusion of operation ease. Really white and black balancing takes few minutes and is accessible to absolutely non-qualified personnel. Thanks God went to past such work as raster superposing. CCDs are fixed in color division system and RG & B channels rasters are always superimposed. But knowing this and being sure please don’t hurry to spin your finger near head hearing that CCD camera is discomposed. It rarely happens but upon line only. Of course the reason is not in CCD position but in difference of group time delay between RGB channels. One has pointed a camera to white, pressed ABB and AWB buttons and please — camera is adjusted. But really camera is not adjusted but is ready for work in particular illumination. Do you remember conversation of two guys given above?

Exploitation experience of even most sophisticated cameras shows that once per 3…4 months camera needs complex adjustment, e. g. setup signal levels at control points, or maybe tuning gamma correction indicator, or balance amplitude characteristics of RG & B channels and so on. It is true for all cameras with no depend on their spheres of use: ENG, studio, OB etc. Otherwise appears that what can be heard closely: camera “greens” or “reds” or picture doesn’t bell and so on.

Work in studio and within OB is connected also to continuous measure of signal level for keeping them conformed to nominal value by black and white. What does it means? It means that camera engineer or technical person sitting on the camera control must have a possibility to observe at any moment on the waveform monitor a luma signal of each camera and upon necessity output to monitor the triad of RGB signals and chroma signal. Last one is necessary if the scene being shot contains fine details with saturated colors; In this case iris stabilization only by luma signal leads to picture distortion. Really:

y = 0.3·R + 0.5·G + 0.11·B

If blue tones dominate in the frame and you control the scene only by luma signal than you’ll open an iris trying to pull the signal up to 100%. During it the other, non-blue details became displayed with limit loosing their coloration.

That’s why camera engineer’s working place must contain waveform monitor and other devices which form so called parade — timing sequence of RGB signals. It is very convenient on a single scan to observe luma signals from all cameras simultaneously. It very much simplifies camera engineer’s operation especially in shooting stories with complex lighting mode or with unprepared lighting algorithm.

Field color synch signals (’k’ letter defines red line and ’c’ letter defines blue line)

Monitoring signals of main colors is possible using color difference signals, representing them in vector form, e. g. feeding for instance R-Y signal to X input of waveform monitor, and B-Y signal to its Y input. Truly said this method is less spread because it needs special waveform monitors.

Absolutely necessary attribute in work with cameras is grade pattern — “gray wedge”. Efforts to adjust the camera without it sometimes give a result but it is rather shamanism then technical operation. Why I describe the work with cameras so deep? It’s very simple. Within budget of any TV facility the “Shooting equipment” chapter is most expensive. And it is natural because cameras are the sources of native signals for further production. They establish native picture quality and create its redundancy for all further processing which can only make the quality worse.

So starting development of production facility, OB or mobile set you don’t save on measure and monitoring tools. It is not wasted money — one just need correctly and insistently use measuring equipment.

Let user doesn’t have an impression that ENG is simpler. Absolutely no. Rather it is on the contrary. If in studio and within OB cameras and their signals are always in a field of view of technical staff than ENG cameras get in their hands pretty rare. Main operation in ENG camera exploitation is automatic balancing. By the way ENG cameras more often bring bad picture and it explains by field shooting conditions. Commonly it is true and that’s why ENG cameras need periodic whole adjustment what means that waveform monitor and “gray wedge” are necessary.

Everything told about cameras (unfortunately far not everything was said) is related also to telecines. Where gathered material goes to? Yes, to editing facility. Each editing facility must have waveform monitor for recorded signals monitoring. And how to be in network modes, in which…

The thoughts

First one

Compression is not eliminating picture redundancy in common leading to unrecoverable loss of frame content but eliminating or more correctly counting of redundancy for simplifying signals transmission. Lost of content is not inevitable result of compression; transmission in this context means any move of information: via paths of within disk arrays. Inside systems with compression as however also inside any digital systems we deal with not signals but with their codes. That’s why senseless became question about parameters and characteristics of these systems in terms of image signals which are natively analog having totally different characteristics and absolutely other results caused by parameters of circuits which these signals passing through.

Second one

Very often people which deal with compressing systems experience temptation to project compression impact onto final result, e. g. onto picture. There are two ways for it. First one presumes use of special test images and their estimation before and after compression. In general this is not measure but expert estimations method. Result is evident but how to interpret it numerically? In world practice a device has already appeared which calculates errors number according to principle “find ten differences” in picture been compressed. But while place of error appearance relatively to image content is unknown this is rather device for equipment developers than for practice. It is good if grass structure in picture disappeared and what of main hero’s eye went off? I exaggerate deliberately to show how sophisticated is everything related to digital systems quality estimation.

Second way is an effort of use traditional testing signals and make on their base numerical compression impact estimation. Usually a question appears: How looks frequency response diagram of for example NLE workstation? They feed frequency packets to its input and get them on output in their nature. What this result means? Absolutely nothing. Really they get response of “Input ADC — output DAC” frequency response. Why? According to compression systems working principle. They catch redundancy either inside a frame (YRGB) or between frames (MPEG). And what are we doing? We feed a signal with 100% redundancy. Recall the frequency packets testing signal structure. All the lines are similar and so are frames. So the system use as reference either one line (YRGB) or one frame (MPEG) and during back processing the native picture fully restores and therefore native signal restores two. At my view this is the one of most convincing examples showing senseless of traditional measures for digital systems.

Third one

There where objective picture control is needed analog signal forms and its measures are necessary.