RESEARCH DEPARTMENT

CO=CHANNEL 1NTERFERENCE BETWEEN TELEV 1S I ON SIGNALS OF THE SAME OR 01 FFERENT STAN'DAROS

Report Moo T=08~

(i... /U)

THE BRIT.ISH BROADCASTING CORPORATION ENGINEERING DIVISION RESEARCH DEPARTMENT

CO - CHANNEL INTERFERENCE BETWEEN TEL~VISION SIGNALS OF THE SAME OR DIFFERENT STANDARDS

( 1982/ U )

G, Fe Newel1, A, t·L L L Eo LW , Taylof1". 1.10 A" ( D, Mau rice) This Report is the propert1 ot the British BroadcastlD& Corporation and aay Dot be reproduced in any form without the written perm18s1on ot the Corporation. Report No" T-084

CO~CHANNEL INTERFERENCE BETWEEN TELEVI SION SIGNALS OF THE SAME OR DIFFERENT STANDARDS

Section Title Page

SUMMARY 0 1

INTRODUCTION 1

THE EXPERIMENTAL PROCEDURE c 2

201 The Test Conditions 0 2 202 Description of Apparatus 3 203 The Subjective Criterion 5

MEASUREMENTS AND RESULTS, 6 301 Co- Channel Interference with Large Offset 6 302 Co- Channel Interference with Small Offsets 10

303 Precision Offset" " 0 G " "" 0' 0 0 0 0 14 304 The Effect of Movement in the Wanted Picture 18 305 Comparison between the 625- line and the 525-line Standards 0 19

SUMMARY OF RESULTS AND CONCLUSIONS 19

REFERENCES 0 21 May 1962 Report No , T-084 ( 1982/14)

CO ~ CHANNEL INTERFERENCE BETWEEN TELEVISION SIGNALS OF THE SAME OR DIFFERENT STANDARDS

SUMMARY

This report describes measurements of the ratios of the amplitudes of wanted and interferin~ si~als which produce a particular subjective ~rade of interference, The measurements were made in order to ascertain the required protection ratios for mutual interference between two stations each operatin~ on any of the three television standards at present in use in Western Europe; 405 lines, 625 lines and. 819 lines all with 50 fields per second,

The results show that? if the frequency difference between the wanted and interferin~ carriers can be "maintained with a stability of less than 2 cls (precision offset), each of the three wanted signal standards require the same protection, In these conditions there is an advanta~e of up to 6 dB in havin~ negative modulation for the wanted signal, If the offset freque~cy cannot be maintained to within a few cycles per second, the required protection is smaller for the systems with the greater number of lines per field,

A comparison between 525 lines, 60 fields per second and 625 lines, 50 fields per second is included,

L INTRODUCTION

Many workers have published the results of measurements of the relative amplitudes of a wanted television signal and an interfering signal for a given subjective grade of interference, Most of these measurements have been concerned with interfering signals which were either unmodulated, modulated with tone or modulated with a television si~al having the same line- and field-frequency as the wanted signal, The measurements to be described here were carried out in order to examine more carefully the effect of the modulation carried by either the wanted or the interfering si~al ; to answer, for example the questions :

(a) What is the effect of the polarity of the modulation of either signal? (b) What is the effect of the number of lines per field? (c) What difference does it make if the two field frequencies are not exactly identical? 2

2. THE EXPERIMENTAL PROCEDURE

2.1 The Test Conditions

The subjeotive assessments were made by sixty observers, men and women, who were not direotly oonoerned with work in the television laboratories. These observers each made assessments for a period of about one half-hour on about twelve oooasions - not more frequently than onoe in anyone week.

Most of the measurements were made with the wanted "programme" oonsisting of a still pioture obtained from a 3! by 3! in (82·6 x 82·6 mm) flying-spot slide­ soanner. Fig. 1 shows the pioture used. A few measurements were m&de with the wanted "programme" oonsisting of motion piotures obtained from a. 35 mm flying-spot teleoine maohine, so that the effeot of movement in the wanted pioture oould be assessed.

'l'he observers were seated in groups of five a.t ~ time in front of a reoeiver and distant from it between four and six times the pioture height. All the receivers had pioture tubes with a. 21 in (53 om) dia.gonal and were adjusted to have a peak brightness of about fifteen foot-lamberts with an ambient illumination of about one

Fig. I ~ Still picture used for tests 8

half foot-lambert, The observers were asked to assess the subjective grade of interference presented to them, using a six-point scale, which will be described in Section 2,8, The assessment.s were individual, no co-opera.tion or discussion of the effects being permitted during the test period, Each observer wrote his or her assessment a.gainst the appropriate test number on a form provided for this purpose, No instruction was given to the observers to assume that the interference was present for only part of the time; they were presented with the interference condition for about a half minute and were asked to assess its subjective effect,

In order to restrict the number of measurements made in this manner, some preliminary tests of an exploratory nature were carried out. These were made by six experienced observers each operating a two-position sm tch which was held in the hand and used to indicate whether the assessment was favou.rable or unfavourable. By means of a meter the test supervisor was able to observe the proportion of favourable assessments. By adjustment of the interference level he could quiokly find the ratio of signal ampli tude.s corresponding to the median assessment. This procedure helped to restrict the range of interference conditions over which the larger-scale tests had to be made. '

A small number of tests were made in which two picture monitors wer~ used and the observers were asked to indicate when the different interferenoe conditions shown on the two monitors were equally visible. These tests were carried out in order to provide some cross-checks on the results of the main series of tests.

The assessments were made with the wanted signal unaccompanied by a sound carrier although such a carrier was sometimes used in order to assist the correct tuning of the receivers,

The amplitude of the wanted signal was adjusted to produce a fringe-area signal-to-noise ratio in the displayed picture, The ampli iiude of the d o a , P4 composite signal was greater than that of the r om. s . noise in the video stages by 27 dB for a 405-line receiver, 29 dB for a 625-line receiver and 81 dB for an 819-line receiver. A number of tests were also made with high sign.al-to-noise ratios. It was found that noise tended to mask the interference pattern. The extent of this masking was such that the amplitude of the interfering signal could be SOme 2 dB greater when the noise was presen t o

2. 2 Description of Apparatus

The picture sources available included:

(a) A 405-line 35 mm flying-spot film-scanner,

(b) A 405-line 3~ x 3~ in [82' 6 x 82' 6 mm i flying- spot slide-scanner, (c) A 625-1ine caption-scanner, (d) A multi-standards flying- spot scanner for either 35 mm film or 3~ x 3~ in {82 ' 6 x 82 "' 6 mm ) slides" This scanner could operate on any of the required standardsc

(e) A 405-line monoscope channel producing a Test Card "C" waveform. 4

Measurements oould be made with anyone of the three standards as the wanted sigaal a.nd anyone of them as the interfering signal with the exoeption of an 8i9-line signal interfering with an 8i9-line signal. In a.ddi tion, two camera channels were available for a short time. These were 4~ in (114 mm) image orthicon channels? one operating on one of the European standards (625 lines? 50 fields per second) and the other on the U,S. standard (525 lines? 60 fields per second). A comparison between these standards was made .

The receivers were:

(a) Two similar? good domestic receivers produced in the U. K. for the 405-line system, (b) Two similar? good domestic receivers produced in Holland for the 625-line system. (c) Two similar? good domestic receivers produced in France for the 819-line system.

These receivers were ~ of course? all of the vestigial-sideband type? but the modulated signals used in the tests were double-si deb and., Some of the tests were repeated with vestigial-sideband filters included in the modulator output circuits, in order to check the effect of the unwanted sidebandso No difference was observed.

The 405-line receivers were capable of operating with either flywheel or hard-lock line synchronization. Comparison of the two methods of synchronization showed that? when a low-frequency component was present in the video interference beat-note, the flywheel synchronizing circuit increased the susceptibility to inter­ ference. On the other hand? when the frequencies involved were too high for the time-constant circuit to follow, the flywheel operation proved an advantage. The differences in the protection ratios required with the two circuits is never greater than 2 dB, and is felt to be negligible in general effect.

Fig. 2 shows the general circuit arrangement. The carrier of the wanted signal was generated by a crystal oscillator and could be controlled in frequency over ± 100 c/s by means of an a. f . c , circuito This enabled the offset (ioe o the frequency difference between the two carriers) to be locked accurately to a sub-harmonic of the line-scan frequency or to a low-frequency oscillator which could be easily and accurately adjusted over a small range of frequencies. The carrier frequencies used for the wanted signal were

61' 75 Mc/s (Channel 4) for the 405-line receiver 62' 25 Mc/s (Channel 4) for the 625-line receiver 65' 55 Mc/s (Channel 4) for the 819-line receiver

The interfering carrier was generated by means of a stable frequency synthesizer for most of the tests but this was replaced by a standard signal generator for certain tests. Considerable care in the layout was necessary in order to reduce crosstalk between the signal sources to an invisible level. 5

WANTED VIDEO SIGNAL

R.F. CARRIER FOR WANTED SIGNAL

WANTED SIGNAL

INTERFERING R.F. CARRIER FOR SIGNAL INTERFERING SIGNAL

SWITCH POSITIONS I. MODULATED INTERFERING SIGNAL 2. C. W. INTE RFERING SIGNAL. D. C . SOURCE

INTERFERING VIDEO SIGNAL

Fig. 2- Block schematic diagram of equipment arrangement

2 03 The Subjective Criterion

The subjective scale used for the tests was :

10 Imperceptible 2 0 Just perceptible 3 0 Definitely perceptible but not disturbing 4 0 Somewhat objectionable 50 Definitely objectionable 6 0 Unusable

The results of the measurements were expressed in terms of the ratio of the signal amplitudes (protection ratios) corresponding to a median assessment about the threshold of disturbance or annoyance? i . e . 5~ of observers returned an assessment which was neither objectionable nor disturbing. This six-point scale has the marked advantage that the grades 1 , 2 and 3 are defined in terms of perceptibility with the proviso that the interference is not disturbing. The other three grades are in terms of disturbance or annoyance. An even number of grades has an advantage over an odd number if the results are required in the form of a median vote based on the middle of the scale; the votes are automatically divided about the mid-pointo If an odd number of grades is used the treatment of votes in the middle grade poses a problem, since these cannot be divided correctly without first examining the distributiono 8 relatively lo~visibility interference pattern which occurs when the offset is near an integral odd multiple of half the line frequency , This offset relationship1 as Behrend2 and others have shown, results in a material reduction of the required pro­ tection ratio. This reduction is about 13 dB at offsets less than about 0"5 Mc/s and drops to 6 dB at about 2-0 Mc/s, Curve C shows the protection ratio for an interfering television signal with the offset adjusted to be near an integral multiple of line frequency to produce the most obvious interference pattern, The protection ratios at small offsets are some 3 or 4 dB less than those required for c ow, inter­ ference. This is because the nominal power of a television transmission is, by convention, taken to be the peak power~ which is much greater than the mean powero Curve D shows the protection ratios for television-modulated interference with the offset equal to an odd multiple of half the line frequency.

For small offsets the curves apply equally for posi tive or negative mo d u­ lation, Individual pictures produce different levels of interference when modulated negatively or positively on the interfering carrier but the results of tests using a large number of slides and moving pictures show that, in general, the effect of the unwanted modulation is negligible.

The polarity of the modulation used for the wanted signal does, in certain circumstances, affect the protection ratio required for large offsetso Before examining these circumstances it is advisable to consider the subjective effects of the interference.

The interference patterns produced on the screen of a receiver are made up of three main components :

(a) The "principal primary" pattern produced by the frequency difference between the wanted and interfering carriers. (b) The "subsidiary primary" patterns produced by the frequency differences between the wanted carrier and the interfering sidebandso (c) The "secondary pattern" produced by cross-modulation. Of these, the primary patterns are the most important but they can be rendered con­ Siderably less visible by the use of a large offset to produce a fine pattern or by arranging that successive field patterns partially cancel one another, As a result of these primary patterns being made less visible the predominant component can be the secondary pattern caused by cross-modulation, The secondary pattern, which is caused by two sources of non-linearity in the receiver, takes the form of a "square­ la~1 version of the interfering modulation, which is cross-modulated on to the wanted carrier. It can be shown 3 that the two sources of non-linearity, the envelope­ detector and the cathode-ray tube, reinforce each other in producing the secondary pattern if the receiver is designed for use with positive modulation, but partially cancel in negative-modulation receivers, If the primary patterns have been made sufficiently faint for the secondary patterns to predominate, the protection ratios for negative-modUlation receivers can be as much as 6 dB lower than those for positive­ modulation receivers, These effects account for some of the differences between the protection ratios at extreme offsets shown in Figo 4.

Fig, 5 shows the required protection ratio for a 405- line receiver sub­ jected to interference from an unmodulated carrier or from a carrier with audio­ frequency amplitude modulation. Curves are shown for the integral-line, and the 9

52 WANTED SIGNAL: 405 LINES ~ 50 CU~VE INTERFERING OFFSET / SIGNAL CONDITIONS 48 \ --- C.W. AND A.M. INTEGRAL LINE -"- A.M. INTEGRAL LINE --- C.W, INTEGRAL LINE 46 I ~, - '- C,W, AND A.M , HALF LINE ...... A .M . HALF LINE C .W. HALF LINE 44 / 1\ ----- INTERFERING A,M. SIGNAL: LOUD MUSIC WITH FREQUENT PEAKS 42 -'1 OF 100'1. MODULATION DEPTH. I1

40 ~ \

38 I ~, / '\ III .., 36 I 1 ~ O· \ ~ / \. \ et 34 z '. o / I \ , \ 6.... 32 I ..o \ g: 30 i ~ \.'. \ .... w ... . - / \" ...... A.M. INTERFERENCE: POSITIVE ..'" MODULATION RECEIVERS. :i '. :; 26 i \ \ 0 , \ I ~ \ ~'" 24 i 1\ z I 0 C. W.INTERFERENCE: POSITIVE ;;: MODULATION RECEIVERS HAVING U \:, ... __ BLACK LEVEL A.G.C. AND FULL .... 2 2 ~\~~ , D.C. COMPON EN T. .. 0 A.M.INTERFERENCE: NEGATIVE et \ \ Cl. , M ODULATION RECEIVERS. 20 \\ \, C.W.INTERFERENCE : POSITIVE 8 \ MODULATION RECEIVERS NOT VESTIGIAL \ >HAVING BLACK LEVEL A.G.C. AND FULL SIDEBAND \ SIDEBAND \ FULL D.C.COMPONENT;ALSO ALL 6 NEGATIVE MODULATION RECEIVERS.

14 1'0 0·5 o 0·5 1' 0 1· 5 2·0 2·5 3·0 OFFSET FREQUENCY, Mc/s

Fig. 5 - Protection ratios required for C.W. and A.M. sound interfering si gnal s half-line conditions of offset. The measurements were made using an audio signal consisting of a music programme adjusted to modulate to 100~ on peaks and were re­ peat,ed with a 1000 c / s sinewave modulating to 40~. For offsets less than 1·5 Mc/s the primary patterns predominate and there is no difference between the protection ratios required for the c.w. signal and those required for amplitude modulation. This is not surprising because the mean amplitudes of the components producing the primary patterns are the same in both cases. The secondary patterns produced by the amplitude-modulated interference become significant, with positive modulation, at offsets in excess of 2'2 Mc/s for the integral-line condition and 1' 5 Mc/s for the half-line condition of offset. The limiting protection ratio required at large 10 offsets is 28 dB. Negative-modulation receivers are not subject to this limitation by the secondary patterns until the required protection ratio has fallen to 22 dB. This is the same pr'otectidn ratio to which the a. g. c. circuits limit positive-modu­ lation receivers with full d.c. response, even if the interfering signal is unmodu­ lated. Fig. 6 shows an attempt to extend the results to include a colour trans­ mission interfered with by an unmodulated carrier. The curve for a 405-line, positive-modulation, monochrome transmission subjected to c.w. interference at integral-line offsets has been modified by the addition of a correct.ion curve. The correction curve is the result of previous measurements4 to find the increase of protection ratio for a colour signal comparedwith· a monochrome signal. Th~ compo­ site curve shown in Fig. 6 applies for the case where the offset has been a djusted to produce the most obvious interference within a range of ± 5 kc/ s o

3.2 Co-Channel Interference with Small Offsets

Measurements were also carried out to examine in detail the effect of varying the offset from zero to a frequency equal to that of the line scan.

Fig. 7 shows the protection ratio required for an unmodulated carrier inter­ fering with each of the three standards of wanted signal, 405, 625 and 819 lines. For these measurements the offset was adjusted to produce the most obvious inter­ ference pattern within a range of plus and minus 25 c/s ( "precision-worst" offset) .

60

405 CURVE FROM FIG.3

50 L V '~I CORRECTION CURVE (FIG.4 /'" IN RESEARCH DEPT REPORT T - 06~ ...... / ~ III... 0- 40 / ~ ...... / '\ Sa: V z / ~/ 0 , "." .... ~ '" .... u 1-' ... 30 ...... - .... I- , 0 , Cl: ...... Cl. .... , ...... 20 - VESTIGIAL FULL SIDE BAND "'" SIDEBAND

, 10 1·0 <>5 o <>5 1-0 1·5 2·0 2·5 OFFSET FREQUENCY, Mc/.

Fig. 6- Protection ratios required by ~05-1 ine monochrome and colour receivers with colour transmission and C.W. interference 11

50 - /' ...... ,/ -...... ALL STANDARDS V ~ L " / "'" / V ~ 40 '" '"... " \, / 0' !i "\ / Cl: V 405 ~ ~ / 0 ~ V .... .--- U V ....UI "- r-.... ./ 0 ...... V Cl: Q. ~t--. /62S V 30 I'" LL --~ --- CURVES ARE IDENTIFIED BY WANTED STANDARDS --...... ~ ...... ---I-- -i--f------r 819 k::::: ,/ ~ 819-LlNE RECEIVER SCAN SYNCHRONIZATION/ ---- AFFECTED AT LOWER PROTECTION RATIOS --

20 o 2 3 4 567 8 9 10 11 12 OFFSET FREQU ENCY, ~c/.

Fig. 7 ~ Protection ratios requi red for C. W. interference ( "Preci sion Worst" offset)

With zero offset the interference produces a change of brightness dependent upon the relative phase of interfering and wanted carriers, If the offset is increased, the interference produces, first a brightness flicker and finally horizontal bands of alternate light and dark shading. The most objectionable pattern appears to corres­ pond to an offset of about 600 c/s; it consists of about ten pairs of light and dark horizontal bands moving up or down at such a speed as to traverse the picture in about ten seconds. Increasing the offset results in a greater number of bands each having smaller area and," therefore, lower visibility, The number of bands is deter­ mined by the relationship between the offset and field frequencies and is independent of the number of lines per field until the number of bands becomes comparable with the number of lines,

Fig, 7 shows that for offsets less than 4 kcls the three standards require the same protection ratios, Beyond 4 kcls the curve for the 405-line receiver begins to flatten and reaches an optimum at an offset equal to half the line frequency, Above this offset the required protection ratio increases to a second maximum near to line-frequency offset, Similarly, the 625-line and 8I9-line curves show optima at offsets equal to half their respective line frequencies, The 8I9-line receiver is capable of producing the least visible interference patterns and can therefore operate with the lowest protection ratio, In the testsit was found that the 8I9-line receivers became difficult to adjust for stable line-scan synchronization when the interfering carrier exceeded a level 26 dB less than the wanted signal amplitude, 12

Fig. 8 shows similar information except that the interfering carrier is modulated with a television picture waveform, For offsets less than about 3 kc/s the curves coincide and are some 4 dB lower than the corresponding curve for C o Wo interference, Fig, 7 0 At offsets greater than 3 kc/s the curves diverge and are identified by two numbers separated by a soliduso The first number represents the wanted-signal line-standard and the second represents the interfering standard. Fig" 8 shows that, at offsets near to 5 kc/ s, the 405-line signal requires a greater protection against interference from a 625-line signal than from either a 405-line or an 819-line signaL Similarly, at offsets around 10 kc/s, an 8ig-li ne signal requires a greater protection against a 405-line interference, than against a 625- line interfering signal. These differences are caused by the relative visibili ties of the subsidiary primary patterns.

If the interfering signal is unmodulated, and the offset is harmonically related to the line-sc an frequency, a pattern of vertical bands of alternate light and dark shading is superimposed on the received picture. The widths of these bands are determined by the order of the harmonic relationship~ the visi bility of the bands has a rather flat maximum at widths of about one-twentieth of the picture widtho If the interfering signal is modulated, the resulting sidebands will beat with the wanted carrier to produce subsidiary primary patterns, These patterns have a visibility which varies periodically as a function of the frequency difference between the wanted c arrier and the particular interfering sidebands producing the

50

CURVES ARE IDENTIFIED BY WANTED/INTERFERING STANDARDS

ALL STANDARDS .]I / -...... '\ / -...... ~ /V IV 40 "" "-\ ID... \ V o ~ \ / z I~ 405/625 ~19/405 Q :--- L ~ .... --? U 405/405 /' Id ~ ~ 405/819 V L ---~ L~ b ::::.. V 625/405 Cl: ~ ~ 6251819 l----'::x Q. 30 ...... / V '" ...... V 819/625 ,/ ~ ~"- ,./ -"- r- '" 25/625 -

20 o 2 3 4 5 6 7 e 9 10 11 12 oFFSET FREOUENCY, ke/.

Fig. 8 ~ Protection ratios required for interference f rom TV signal ( ~ P recision Worst n offset) 13 patterns; a peak of visibility is obtained when this frequency difference is near to being harmonically related to the wanted line-scan frequency. When the two signals have the same line-scan frequency all the subsidiary patterns have a maximum visibility at those offsets wh ich cause the primary pattern to have maximum visi­ bility. If the signals have different line-scan frequencies the subsidiary patterns may, as in the case of 625-line interference with a 405-line signal, have maxima which are so displaced in offset tha t there are always flome components near to maximum visibility. This accounts for the fact that the minimum protection ratio for this case, shown in Fig. B by the curve 405/ 625, is greater than for the case where the interference is a 405- or even an Bl9-line signal. The case in which the interfering signal has a line-scan frequency exactly twice that of the wanted signal could apply to an B19-line signal, locked to a mains frequency half per cent low, and interfered with by a 405-line signal locked to a mains frequency half per cent high. The minor peak in the B19/405 curve of Fig. B at 10' 25 kc/ s offset can be explained with the aid of the simplifying assumption that a television signal has a spectrum consisting essentially of components spaced at multiples of the line-scan frequency. The subsidiary patterns produced by the even-order sidebands of the interfering signal will all have peaks of visibility at the same offsets as produce Feaks of visibility of the primary pattern. The odd-order sidebands will differ in frequency from the wanted carrier by integral mUltiples of the wanted line-scan frequency only when the offset is an odd mUltiple of half the line-scan frequency. This will cause all the odd-order sidebands to produce patterns that have maxima of visibility whenever the patterns produced by the unwanted carrier and its even­ order sidebands are at minimum visibility. The facts that the two nominal line­ scan frequencies are not in exact two-to-one relationship and the spectra are more complex than was assumed, do not ~ffect the result materially.

Small-scale tests conducted with large offsets were carried out using a 405-line signal locked to mains fr~quency interfering with an B19-line signal h aving crystal-controlled scan frequencies. This made it possible to maintain a stable relationship between the offset and the wanted line-scan frequency. In these conditions the minor peak was obtained at each offset, up to about 1 Mc/s, that was an odd multiple of half the wanted line-scan frequency. The required protection ratio was about 10 dB less than that at the nearest offset that was a multiple of the line-scan frequency, Offsets that were odd multiples of one-quarter of the wanted line-scan frequency produced a further reduction of 6 dB in the required protection ratio.

The tests were repeated with the Bl9-line signal locked to mains frequency instead of to a crystal-controlled signal. The reduced stability of the wanted line­ scan frequency caused the minor peak to become less obvious as the offset was increas­ ed. The ratio

Protection ratio required for an offset of (n + 1 / 2) times the line-scan frequency

Protection ratio required for an offset of n times the line-scan frequency

is shown in Table 1 for various values of n. 14

TABLE 1

Protection Ratios for Bl9-line wanted and 405-line interfering signals

Offset frequency half-line protection ratio

Line-scan frequency integral-line protection ratio

(n) (dB)

5 -12 32 -13 50 -14

It is of interest to note that the modulation components, or sidebands, of the wanted signal do not materially affect the interference patterns. This is because, in general, the amplitude of the wanted signal is at any instant larger than that of the interfering signal and therefore does not modulate the beat ,frequency component. Even if the amplitude of the interference were affected by the amplitude of the wanted signal, any changes in the vision interference pattern intensity would be coincident with changes in the brightness of the wanted picture, and would not be obvious. The results for a 405-line signal interfering with an B1g-line signal, shown in Fig. 8, are in agreement with those of Vigurs,5 although his results were obtained using comparison measurements made with a somewhat lower level of inter­ ference.

3.3 Precision Offset

It is well known that the visibility of an interference pattern is con­ siderably affected by the relationship between the difference-frequency (offset) and the field frequency of the wanted signal. If the offset contains an exact number of half cycles in each wanted field period, a stationary pattern will be produced. The pattern can be arranged so that succffisive fields reinforce each other or so that they tend to cancel each other. There is a very material difference in the visibility of the pattern in these t~o conditions. In determining the offset conditions for the partially cancelling sequence of patterns ("precision­ best" offset) it must be remembered that the start of successive fields is displaced in the raster by half a line. If half a line period contains an even number of half cycles of the offset frequency, the successive fields must commence with opposite polarities of the difference (offset) waveform. If half a line contains an odd number of half c~cles, the successive fields must begin with the same polarity. Hopf6 has expressed this in mathematical .f.orm as fo] lows:

Offset for minimum visibility = mfL ± (2n + 1)fp for (2n + 1)fp < fL/2 where m and n are positive integral numbers and fL and fp represent t-he linfrscan and. picture repetition frequencies respectively. The offsets midway between the yalues for minimum visibili ty produce a station ary pattern of high ~isibility. During the tests, which were made with still pictures for the wanted programme, it was found that this stationary pattern could be made more obvious by changing the offset by some 4 c/s. Thi s caused the pattern to move slowly up or down the picture. 15

Fig. 9 shows the required protection ratio for c . w. interference with each of the three standards as the wanted signal. The continuous lines show the results of measurements made with the offset always adjusted to produce the most obvious interference patterno The broken lines represent the precision-best offset conditions. It was found that these values of offset had to be maintained within one or two cycles per second of the precise frequency to prevent the interference patterns from becoming visible.

As mentioned before, the 819-line receivers required very careful adjust­ ment, of the synchroni ring controls if the interfering signal exceeded the amplitude corresponding to a protection ratio of 26 dB o

Fig. 10 shows the protection ratios required for the precision-best and "worst" offsets for each of the wanted standards interfered with by television signalso

The differences between the protection ratios required for the precision- best and -worst conditions are shown as functions of offset in Fig. 11. These differences are extracted from Fig. 10.

Figso 12. 13 and 14 show the information for three wanted standards, 405, 625 and 819 lines respectively, extracted from Fig. 100

lIO

,.. -...... ",,- '-'.- CURVES ARE IDENTIFIED BY ' WANTED STANDARDS /' t'-.,. " PRECISION WORST» CURVE ~ _ ..... - --"PRECISION BEST"CURVE / fAU. STA~ DAI1D i---l ~ / , , / / / / " '\ / ;/ ...CD \. / / , '" I o \ '\ / / ~ \.. 7 •Cl: ~ / /1 Z ~ ./ 7 " \ / ....~ X v ~\ V ...... I ...... [7 '\ 6250- / ~30 I ~ ~ Q. j....-- , "Oll If ~ "-...... ~ K ~~ ...... 819 ... ' ... f"< , // ,-1--- _. -t-.. ~ --~r=:;: .,-~ .. :- .. -+- •• :----. .-- , // V l' V 625 r---~ ~ ...... 819-LINE RECEIVER SC/toN SYNCHRONIZATION/ 819 - ~-p AFFECTED AT LOWER PROTECTION RATIOS ""-t--r- i-- .------20 o 2 3 5 6 7 8 9 10 11 12 " OFFSET FREQUENCY, kc/s

Fig. 9 ~ Protection ratios required for C.W. interference with wPrecision Worst w and "Precision Best" offsets 16

so

CURVES ARE IDENTIFIED BY WANTED 1 INTERFERING STANDARDS (L) FIELD FREQUENCIES LOCKED TO SAIo4E Io4AINS FREQUENCV(BY SEPARATE "SPOPjGY' LOCKS) /' ...... (U) FIELD FREQUENCIES LOCKED TO ~ DIFFERENT MAINS FREQUENCIES "PRECISION WORSTHCURVES ------40SWANTE:rRECISION BEST" '\. ------6 2S WANTED 40 '\ ID 4OS(62S(U) -----}SI9 WANTED CURVES " , r\ ------ci , j: ".--1\., " , \ •Cl: , , , , ~s{U) ~ , , '- l- \ ',- t------{40s/40s(UAND(u') V :"" ...... ~ ~ ..... ---- 40S/819 (l)ANo(U) ~ I-'" --- ~ 0 405140$)}~ ~ ",""-- 62 S/",?S(QAND(U) V B19/40S(Q AND(U) ...... ca.Cl: \~ , 40S(819(U) lie;: 30 ~ ~ ., ---- ~~ ~ ... \.:'~ r"-,_ ~~ ; ; N I"''''''''''' V -. \ ~ '" '"'-:, , /_--! ~ - \ 'I!'--~.. 'I- f-- Ir':-- ,....- \ ~:- --- ~" \ oL , '--.--::::' .:;..: SI9/40S(U) ./ _/0.: ~~ r---. S19/40s(l.,) " ~ SI9-LINE RECEIVER LINE SCAN 62S/405(u) 40s(~~ " SYNCHRONIZATION AFFECTED "'- 62S/40s(L,) " AT LOWER PROTECTION RATIOS

20 o 2 3 4 s 6 7 S 9 10 12 OFFSET FREQUENCV, ke/. " Fi g. 10 - Protection ratios requ i red for modul ated (TV) interference wi th "Precision Worst" and "Precision Best" offsets

11

10 / ~

CURVES ARE IDENTIFIED BV !lI 9 L. \ WANTED 1 INTERFERING STANDARDS rj ~05/40S(L) ~ (L) FIELD FREQUENCIES LOCKED TO ~ 8 I \\ ~ IX SAME MAINS FREQUENCY (BY \ \~, \ 62S/40S(L} ~ 7 / SEPARATE 'SPONGY" LOCKS) .... (U) FIELD FREQUENCIES I.DCKED TO / BI9i405(U) ll! ./ ~ \' .... 6 -..... " '\ DIFFERENT MAINS FREQUENCIES. ~ ~ -,~ \~ IL 5 o "'" z 4 40S/~S(~ ~ :~ o ""-.... j: ~ ~ ~405(L) ~ 3 '\'" ~ 2 4OSI62~(U~ \\'\ ","~~ "\ " \\ I\.~ r-...."'"~ ~ o ~ 62~ " '- o 2 3 4 5 6 7"" B 9 10 11 12 OFFSET FREQU·ENCY. kc/I Fig. 11 - Reduction of protection ratio for "Precision Best" offset relative to "Precision Worst" offset 17

so CURVES ARE IDENTIFIED BY INTERFERING STANDARDS (Ll FIELD FREOUENClES LOCKED TO SAME MAINS FREOI£NCY (BY SEPARATE ·SPONGY· LOCKS) (U) FIELD FREOUENCIES LOCKED TO DIFFERENT MAINS FREQUENCIES. /' ...... 'PRECISION WORST' CURVES I'.... -_._-_._-- "PRECISION BEST" CURVES /V "" ALL STANDARDS V ~ / 40 \ 1\ V ~ , --.. - , , .... , ",". , ---, , , , , -- Q , , , , \ lL / ~ , " a: , "' , , , , , , 625 ~\ ..,...... - , , , I",\: I'-- ~ ~ ...... -- / G , " - -' -, / ~/ , '" "-; ~ --- - / \ ...... k:::: ;--:;;, d405 (L) AND (U) ,( \, 625~( :::;:-, -\819 lU) ~ \, ,'/ /1 30 , ' , \ , / , , , 405 (\J} f-;7' , , ' , " 819 .M / , \ " \ , /' , , , \ ,,' , , "- I , ,,'. \ I , //

/['405 (L) " , , , , " "

20 o 2 3 4 5 6 7 8 9 10 II OFFSET FREQUENCY, kef. Fig. 12 - Protection ratios required by a 11-05-1 ine trMsmission for modulated (TV) interference wi th "Preci sion Worst" and "Preci sion Best" offsets

50

CU~VES ARE IDENT! FlED BY INTERFERING STANDARDS

(J..) FIELD FREQUENCIES LOCKED TO SAME /' MAINS FREQUENCY (BY SEPARATE f"-.. 'SPONGY" LOCKS l l'-.. CUl FIELD FREQUENCIES LOCKED TO DIFFERENT MAINS FREQUENCIES 40 f\. 'PRECISION WORST" CURVE ------'PRECISION BEST" CURVES 'Il I\. " , 0 " ... \ ~ , a: , v 40S (L.) AND (IJ) z , ;::0 1rl... .. '" ...... ~ '~" ..... I ...---V '" , 30 , , ,- .. ' --- ", "' . -- - - ". . , 40! , -. -- . , , (u)" "_ ...... " -. " " , ". , , , ...... " , , , , -, ~ ..... - -. """" , , " -" , , , , , , , , , , ,.4- 405 (L) ---- "

20

o 2 3 4 S 6 7 8 9 10 II 12 OFF SET FREQUENCY, kef.

Fig. 13 - Protection ratios required by a 625-1 ine transmission for modulated (TV) interference with "Precision Worst" Md "Precision Best" offsets 18

50

CURVES ARE IDENTIFIED BY INTERFERING STANDARDS

(L) FIELD FREQUENCIES LOCKED TO SAME ,,- MAINS FREQUENCY (BY SEPARATE ...... 'SPONGY" LOCKS) "'- (U) FIELD FREQUENCIES LOCKED TO DIFFERENT MAINS FREQUENCIE S. 40 "PRECISION WORST"CURVE. \ dl... [\. -----·PRECISION BEST" CURVE. 0' /' ...... '" ~ '. \ "Z 405 (L) AND (U) 0 I'\. I- U \ W V "i'--. I- 0 ~. "-.., /' "IL 30 ...... "./ "'"- 405(U) I' ~ ~' r---~ r--- . --: ?"" ~-. I-- ' ...... -- ./0/ , -- -- '-f-- l..-Y'" -- --~ b;::- 819-L1NE RECEIVER SCAN SYNCHRONIZATION 405(L)- f-- '- -- -- AFFECTED AT LOWER PROTECTION RATIOS

20 o 2 3 5 6 7 8 9 10 12 OFFSET FREQUENCY, kef, "

Fig , !~ = Protect ion ratios requ i red by an 8!9 ~ 1 ine transmission for modulated (TV) interference with "Prec i sion Worst" and "Precision Best" offsets

304 The Effect of Movement in the Wanted Picture

The measurements described so far had all been made with a still picture for the wanted programme. In order to assess the effect of having both some interest and some motion in the wanted programme) a small-scale set of measurements was made with the wanted programme consisting of a short excerpt from a motion picture. These excerpts lasted for a few minutes and consisted of self-contained sequences of events e . g , parachutists preparing to jump and then leaping into space; a man approaching a motorcycle) starting the engine and driving away.

The assessment was made on the complete excerpt , The results indicate that for coarse patterns) or for brightness flickering caused by very low beat frequencies) the effect of movement in the wanted picture is negligible. As the pattern becomes finer ) the effect of movement in the wanted picture is to mask the interference slightly. The results are shown plotted in Fig. 15. The improvement shown only applies when the interference pattern is largely determined by the inter­ fering carrier, If) for instance? a 405-line signal were interfered with by another television signal with an offset of 5 kc/s and the level were such that coarse patterns due to cross-modulation effects were present? the reduction would not apply to these coarse patterns" 19

10

9 Cl).., THI5 EFFECT DOE5 NOT PERMIT PROTECTION RATl05 L.OWER THAN o 8 TH05E IMP05ED BY 5YNCHRONIZATION TROUBLE5 OR SECONDARY I PATTERN5 DUE TO MODUL.ATION OF THE INTERFERING SIGNAL. I a:~ 7 I I I i z I Q 6 i -- r------+-I - I-- - ! 0- U Id 0- =1 I o a: "- ... 4 r-- o , ~ ~~OPllMUMt - , z 3 o f..--~625 OPTlMU~""--- '----~ t-- 2 Ir-- - ti ::> ....- t-- a I--- --r--t-- f..-- Id 1\ - I--- ! I - a: V ...... r---- I V - 11405 OPTIMUM r--~ o V I '---- I o -- 2 3 4 5 6 7 8 9 10 11 12 OFFSET FREQUENCY, kc/s

Fig, 15 ~ Reduction of protection ratio resulting from masking of the interference by motion in the displayed picture

3. 5 Comparison between the 625-line and the 525-1ine 2tandards

A small number of measurements was made to compare the protection ratios required at close offsets by a 625-line, 50 field si~al subjected to c . w. inter­ ference and a 525-line~ 60 field si~al also subjected to cow. i nterference" The results are shown in Fig. 16. The measured points are shown in addition to curves derived by interpolation of other results" 1 It will be observed that the 525/60 receiver requires a larger protection at offsets greater than 1 kc/so The inter­ ference patterns consist of horizontal bands. The 525/60 receiver, having a higher velocity of vertical scan ~ reproduces these bands as a more coarse and therefore more obvious pattern than does the 625/50 receiver" In addition to this effect, the vertical dimensions of the horizontal bands at the optimum offset, half line fre­ quency, are a function of the number of lines in each picture because each half cycle of the interference is reproduced by only one scanning linea The pattern on the 525/ 60 Iicture is therefore coarser than that on the 625/50 picture: this accounts for the 2 dB increase of protection ratio required by the 525/ 60 si~al at 7"8 kc/s offset"

4 . SUMMARY OF RESULTS AND CONCLUSIONS

The primary patterns are the most obvious components of interference effects. They can be reduced in visibility by using as large an offset frequency as possible in order to produce a pattern structure as fine as is possible with the wanted system. Even with close offsets a considerable reduction in visibility can be achieved if full advantage is taken of the fine pattern obtained by using offsets near to integral multiples of half the wanted line-scan frequency.

If the offset cannot be contr olled to maintain a stability of less than one or two cycles per second, the lowest protection would be required by the system having the largest number of lines per field" The effect of i ncreasing the field frequency is to increase the vertical dimensions and therefore the visibility of any particular pattern. 20

50

CURVES ARE IDENTIFIED ~ BY WANTED STANDARD '"""'" K" (LINES/FIELDS PER. SEC) f.-- EXPERIMENTAL RESULTS ~ '"",- • 625/50 f.-- ~ " o 525/60 " 40 '\. "- '\ - III i'_, 525/60 ." I\. 0 '-{ I-« a: 0" z ""~ 2 l- 625/50 "'-... V ~ ,~ 1&1 ~~ r- "- I--~ 0 -0 a: " Il. 30 .... '" i--

3 4 5 6 7 8 OFFSET FREOUENCY, kefs

Fig. 16~Protect i on ratios requ i red by 62S/S0 and 52S/60 displaysforC. W. interference The best results obtainable with close offsets occur with the use of preclsl0n offsets near to 1/3, 2 / 3 . 4/ 3, etco of the wanted line- scan frequency. This enables advantage to be taken both of the vertical resolution and of the storage of the human optical system, which permits partial cancellation by successive reversal of patterns.

When full advantage of these effects has been taken, the predominant components of the interference patterns are the secondary patterns caused by the inherent non-linearity in the receiver. In these conditions the use of negative modulation for the wanted system has some advantage. The limiting protection ratios required for the secondary patterns in conditions where the primary patterns have been reduced to a low visibility are;

(a) With positive modulation on the wanted system, 27 dB i f the signals have different line-scan frequencies or are not locked to the same mains frequency, 25 dB if the signals have the same line-scan frequency and are locked to the same ma i ns frequency ; 28 dE if the interference is an amplitude-modulated sound signaL

(b) For negative modulation of the wanted s ignal, these ratios can be re duced by as much as 6 dB " 21

Advantage c an be taken of these limiting protection ratios only when t h e p rimary pattern has been made relatively invisible by the use of a large offs et or, in some cases , by a precision offset. If the interfering signal is the a .m. sound of a television trans mi s sion, it is u s ual to quote the protection rat ios in terms of the two vision-carrier amplitudes. In this case, the ratios for the a .m. sound s i gnals quoted above should be r e duc e d by the number of decibels by which the interfering c arrier (p eak) exceeds the interfering sound carrier.

5. REFERENCE S

1. "Pattern Vi s ibility in Telev ision Co-Channel Interference", Research Department Report No. T-091.

2 . Behrend, H.L., "Reduction of Co- Channel Television Interference by Precise Frequency Control of Television Picture Carriers", R.C.A. Review, December 1956.

3. "Television Co- Channel Interference: An Advantage of Negative Modulation", Research Department Report in p reparation.

4. "A Compar i son of C. H. Interfer e n ce Effects on Colour and Monochrome Televis i on Transmissions", Research Department Report No. T-0'31, Serial No. 195'3/21.

5. "Co-Channel Inter fer ence in Te lev ision Broadcasting: Effect of 405-Line Interference with Reception of 819-Line Transmission", Research Depart­ ment Report No. T-071, Serial No . 1959/22.

6. Hopf , H. , "Experiments on the Operat i on of Television Transmitters with Pr eci sion Offset Carrier Frequencies", Rundfunktechnische Mitteilungen, December 1958.

MV