The invention relates to a method in the so-called cutting of recorded colour television signals for edi¬ ting purposes, and in particular signals according to PAL and SECAM standards. More specifically the invention relates to a method of the kind set forth in the preamble of Claim 1.

In order to comprehend the present invention, it is necessary to have some knowledge of the aforesaid signal transmission and recording systems, upon which much has been written. A brief disclosure will now be made of the nature of the problem which the invention is meant to solve. The invention will be primarily described with a starting point from the PAL-system, although it will be understood that a similar problem also exists in the SECAM-system, this problem also being solved by the invention, and in the NISC-system.

It is known that a conventional TV-signal reprodu¬ ces 25 frames per second, each divided into two fields each of which sweeps interdigiterly 312J lines, or together 625 lines. The colour information exists in the form of a luminance signal, which is compatible with black and white reception, and a chrominance signal, which exists as a sideband of a suppressed carrier wave ^' at a distance of approximately 4.43 MHz from the lumi- nance signal. In the PAL-system the chrominance signal, from which two scalar colour signals shall be taken in order to calculate together with the amplitude of the luminance signal the three colour signals, is transmitted in phase quadrature wherewith the chrominance signal is displaced in phase through 180° for each alternate line, in order to compensate for phase errors when de-codirig. In the SECAM-system the one component is instead trans¬ mitted in each alternate line and the second component in

each other line, wherewith calculation of the three colour signals is effected with the aid of the one component directly and the other component taken from a preceding line, collected from a delay line. Thus a fundamental principle of both the PAL and the SECAM systems is that the information is not similarly coded in two mutually sequential lines, but that said lines have opposite chrominance coding and must each be decoded in a manner peculiar thereto. Now, if the number of lines in a TV-frame (two fields, 625 lines) had been an even number, then the problem solved by the invention would never have existed. It has not been possible, however, to avoid an uneven number of lines, because of the problem of interlacing between the fields. Decoding of the chrominance signal in the PAL- system is effected by means of a reference signal (colour carrier wave f ) obtained from an oscillator which is ^" normally crystal controlled and which is phase locked to a series of short carrier-wave transmissions, which take place over horizontal and vertical retrace periods, or .blanking intervals. During these burst phases the trans¬ mitted carrier-wave frequency is phase-displaced alter¬ nately through + 135° and alternately through -135° in relation to the ideal oscillator frequency. The task of signal "cutting", in which a series of frames with rele¬ vant control signals are placed one after the other in immediate sequence on, for example, a video.tape (or during transmission) can be beset with certain problems since the periodicity of the control signals does not coincide exactly with the frequency of the frames.

For example, in the case of the PAL-system the periodicity is equally divisible by a period correspon¬ ding to four frames (eight fields) , at least if the horizontal synchronizing (sync) signals f„ and the colour carrier-wave f _Q originate from one and the same oscilla¬ tor frequency according to the formula

4 f ^f H = 1 135 + 4/625 which is recommended by CCITT. This equation can be written as

- f ^f H "

709 379 o or

4 A . 5 C ⁴ f

709 379 o

It will be seen immediately that the denominator and the numerator lack common integer factors (the denominator is an odd number and does not end in 0 or 5) , and that the two frequencies are equally divided after 4 . 625 periods for f„π., corresponding accordingly to four frames. Thus, in this respect four types of frames can suitably be discerned, and can be designated I, II, III and IV. Two frame sequencies can be spliced with no problem whatso¬ ever from the control signals, provided that there is no interruption in consecutivity in this respect, i.e. provi¬ ded that one proceeds from I to II, II to III, III to IV or IV to I from one sequence to the next. (The periodi¬ city is therefore four frames or eight fields.)

For the sake of simplicity this type of splice will be referred to as splice-type A. Splice-type A can always be practised, although when one is bound to this type of splice it places compulsory conditions upon the cutting procedure. "Cutting" of film in the classical mode is, after all, effected to an accuracy of one frame, and in many instances it may be of actual importance that so is the ..case. It can be noted that a conventional feature film is composed of 800 - 1200 spliced scenes, and hence simple and problemfree cutting is a necessity.

It was seen in the aforegoing that both the PAL and the SECAM systems employ a line-to-line signal switch providing some form of symmetry with the period two frames (which accordingly together have an.even number of

lines) . Consequently it is conceivable to splice _ between two frames from even to odd numbers or from odd to even numbers according to the same designation (that is from I to IV, II to I, III to II or IV to III) . This type of splice is referred to here as splice type B.

Since, as before shown, the colour carrier wave is evenly divided by four frames, but not by two frames it is obvious that four frames contain an odd number of periods of the colour carrier wave, and hence it will be seen that there is a phase difference between the spacing of two frames which reaches to 180° or one half of a period. This can be corrected in a tape machine in a known manner, with the aid of a time base corrector which delays the line signal for a suitable length of time. There is then obtained a displacement corresponding to 1/567 of a frame line occurring as a lateral shift in the frame. This has no disturbing influence when the scenes are dissimilar, but can have a disturbing effect when concerning animated film or cartoons. These types of splices have hitherto been conside¬ red the only ones possible. It has been considered impossible by all those skilled in this art to cut a frame at just any point, because of the structure of the control signal. An object of the present invention is to eliminate this limitation and to enable a frame to be cut at any desired point, i.e. to make it possible to "splice" the signals at any selected frame sequence.

Table I illustrates all conceivable frame sequen- cies in a. splice. As disclosed in the aforegoing the splice types A and B are known and practised.

TABLE I

Last Picture First picture after splice-splice type

B '1

I II IV I III

II III I II IV il IV II II I IV I III IV II

The problem is now the splice types C- and C ₂ shown in the Table, this problem, as it has appeared, being that the correspondence between chrominance signal and reference phase in the. colour carrier wave is so poor that after a cut the colours are initially wrong (this is equally true for both PAL and SECAM) . If the splice, is carried out in spite of this, the subsequent distur- bance in picture reception will be quite noticeable and the origin of this disturbance can be readily discerned by one knowledgeable in these matters.

This problem is solved by means of the present invention with the aid of a method of the kind set forth in the introductionhaving the characterizing features set forth in the characterizing clause of Claim 1.

Thus, a fundamental feature of the invention is that there is ensured a consistent change in the coding of the chrominance signal even in those signal "cuts" where the coding is naturally wrong. Although it is admittedly possible to cut when this is not done, the receiver or monitor receiving the signals will "get out of step" and it will take a longer or shorter period of time for the error to compensate itself, in the case of the PAL-system a time corresponding to half a frame, since the colour carrier wave oscillator in the television apparatus must have time to re-phase. Accordingly, in the case of an editing splice of the aforesaid kind, the first line in

each f ame is "discarded" when proceeding in accordance_ with the invention.

When viewing the result on a screen this means that there is obtained a vertical displacement equal to one line spacing - the picture is displaced vertically on the screen through a full line spacing, which corres¬ ponds to 1/300 of the picture height. This presents no disadvantage in motivational changes. In the case of ani¬ mated pictures, however, such displacement is generally unsuitable, as it is also with the aforesaid splice type B (Table I) since in these cases there is also a lateral shift which corresponds to half a period in colour oscillator frequency.

In practice the invention can be applied with the aid of conventional studio-type tape machines with but few complicated auxiliary devices. These machines, equipped with a ^" time base corrector, operate in principle with digitalized signals. In playback modes, the signals are digitalized line for line and are read into FIFO-type memory stores from which they can be later read-out and re-converted to analogue signals and recorded on video¬ tape. Such machines are well known and can be acquired from AMPEX, Marconi or SONY. The change of signal sequen¬ ces for different lines as envisaged with the invention can be readily effected due to the fundamental construc¬ tion, which ^" is designed to enable transcription to be effected from one tape to another with compensation for all eventual speed variations while maintaining synchro¬ nization. The invention will now be described with reference to an embodiment according to the PAL-system and with reference to the accompanying drawings.

Figure 1 is a conventional diagrammatic illustra¬ tion of the signal sequence in the PAL-system. Figure 2. shows how the invention functions. Figure 3 is a block schematic. Figure 4 shows part of a block schematic taken

from a manual and illustrating a separate, non-limiting embodiment of the invention.

Figure 1 , which is taken from a standard manual on the PAL-system, shows how the fields alternate in the system and how the various line signals and burst phases are placed in relation to the vertical retrace period or vertical blanking interval, which has a duration of 25 lines, and of which there are two per frame. Of the 625 lines which make up a frame period, there is thus only 575 visible lines, starting with the latter part of line

23. The first field is terminated with line 310 and there¬ after follows line 336, which after a vertical retrace is placed between the sites for lines 23 and 24, line 337 etc. to line 623, which is only a half. This is also illustrated in Figure 2, which illustrates the line placement symbolically.

As mentioned in the introduction, when splicing in accordance with the invention under certain ^' conditions one line is discarded. One such case can be illustrated throug Figure 1.

The phasing of the burst signals is illustrated in Figure 1 by designations +135 and -135° respectively. Assume now that the end of the frame A which terminates in. (the black) the line 625 uppermost, which falls in a vertical retrace period, is to be spliced with frame C in the diagram. In frame A we see that the phase of the last burst phase (in line 622) is +135° and that- the plus phase consequently lies in lines having an even number. In frame B in the diagram the first burst phase lies in line 7 and has the phase +135°. In this case the plus phase thus lies in an odd-numbered line, and this adaption is correct since the chrominance coding changes therewith. If we now wish to pass over frame B in the diagram we see that in frame C the burst phase lies in line 6 and.has the phase +135 .If such a splice were to be made a phase error of 180° would result between the burst and the colour carrier-wave oscillator in a receiver. However, if

it were possible to adjust the oscillator immediately to the correct timing, no problem would exist; no such possibility exists in practice, however. (A corresponding adjustment in the SECAM-system would be still more troub- lesome) . As a result the coding will be wrong when colour signals arrived from and including line 23.

When proceeding in accordance with the invention phasing is synchronized and a jump is made over one line in the system. When reconstituting the video signal, which is effected during the aforesaid transcription, the burst phasesare permitted to continue in the same manner in the preceding and the following section, meaning that frame C obtains a positive phase in even numbered lines. When the site on the video tape corresponding to line 23 is located in front of the head there is placed at said ^• location a black level, whereafter the signal sequence belonging to line 25 is placed in the site for line 24, etc.. As will be seen from the continuation of the diagram, this means that information intended for line 310 lands in line 309 and that line 310 is empty of information and extinguished. The scanning functions continue in a normal fashion and the signal sequence intended for line 324 lands in line 323 in place of the information intended for line 323, which latter arrives in the vertical retrace. Ultimately the signal sequence intended for line 623 will fall in line 622 and line 623 will be devoid of informa¬ tion. In other words, when shown on the screen the frame will be displaced through a vertical distance corresponding to two lines. It has been noted in the introduction that the colour carrier-wave frequency is divided equally over four frames (8 fields) but over this period consequently has an odd number of periods. It will be understood that since the colour carrier-wave has one and the same phase at both ends thereof, the carrier-wave must have been - displaced through one half period in the centre .of such an interval. A splice which is to give the correct phase

must take this into account. This is done by controlling the transmitted digitalized frame signal through the time • base corrector with a corresponding time shift of one half period. This corresponds to roughly 1/567 of a line, or roughly in the order of one millimetre. It will also be seen that a division of the 8-field interval into four parts will result in the need to compensate for phase shifts relative to the ends of the interval of τr/2, π and 3r/2, if coding is to be correct. These are thus corresponded by cases C. , B and C ₂ respectively in Table I. Thus, there is obtained in all of these cases a hori¬ zontal shift of varying magnitude, due to the influence of the time base corrector.

The invention is normally carried into practice with the aid of a data, processor arranged to control the two tape machines, as illustrated in Figure 3.

In the following there is described by way of example a non-limiting embodiment of a method according to the invention as applied to an apparatus designated AMPEX VPR 2B, which records on one-inch tape type C. When recording ton tape of the kind meant here, there is recorded in addition to sound and video signals also a control track. This track contains information as to how the signal on the tape is recorded and is operative during play-back to lock the tape feed to a separate station reference (crystal controlled) generated in the machine.

This control signal contains a negative pulse at the beginning of all odd-numbered fields, a positive pulse at the beginning of all even-numbered fields, and a pulse train of particular appearance (negative + positive + negative) at the beginning of each eigtht field (field 1 in 8-field sequencies) . When the tape is played back there is- generated at the correct playback speed an internal 8-field pulse (one for each fourth frame, i.e. 6.25 Hz), a 4-field pulse (12.5 Hz)and a 2-field pulse (25 Hz). At the same time, signals corresponding to these

three signal series are generated from the synchronizing, equipment of the station, and tape movement is locked-in to these signals. In addition there is generated a so- called PAL-phase-signal having a pulse for each alternate line, i.e. with the frequency 7.8 kHz, which indicates when the colour signal shall be inverted or normal.

The actual locking-on of the tape machine is effec¬ ted in the following manner: the servomechanism responsib¬ le for tape speed operates at 95% of the nominal speed, at which the control signals of the station and the tape are accordingly displaced progressively towards each other until there is coincidence between the two, where¬ upon the tape-speed servomechanism is caused to lock-in the correct speed. With commercially available machines it is then possible to lock-in 8-field correct, although it can also be decided to lock-in 4-field correct, where¬ upon the 6.25 Hz-condition no longer applies.

In one example of how the present invention can be applied there is used a plurality of tapes upon which film has been recorded and which are therewith provided with control tracks. Also present in one of the sound channels (at times in the vertical-sweep component of the video signal) a frame-marking in hours, minutes and seconds and frame-number (25 frames per second) . Prior to cutting, the editor chooses which scenes shall be present on the final tape and lists the frame-markings successi¬ vely for the first and second frame of each scene. This can be done in a variety of ways. One of the least expensive methods in this respect is to work with a copy transferred onto a simpler VH5-cassette, where each frame is provided in a known manner with a visible frame marking as the film is copied from one tape to the other. The list is then used in the final editing or cutting stage.

The video-tape which is to receive the final edition of the video-film is then prepared in a tape machine, by providing the tape with a continuous control signal and with frame markings in running order in the sound channel

intended therefor. The film is then copied from the original tape onto a copy tape. The now prepared copy tape is placed in a first tape machine and a first origi¬ nal tape is placed in a second tape machine. It is assumed that it is desired to copy from the original tape a section with the first frame 0 hrs, 3 mins, 8 sees, frame 11 and the last frame 0 hrs, 3 mins, 52 sees frame- no 2, this section to be copied at a commencement time of 0 hrs, 1 min, 16 sees, frame-no 2. The second tape machine has been set to omit the

8-field synchronization. The tapes are adjusted by being wound forwards or backwards as the case may be to a point corresponding to six seconds before cutting. It is now established whether or not 4-field synchronization is possible. This is effected by adding together in both instances the values of both seconds and frames, thereby obtaining in this case 19 and 54 respectively, the one thus being an odd number and the other an even number. Four-field synchronization is not therefore possible. This is equal to one of the cases C _j or C ₂ in Table I above. (If both values had been even or both odd the problem solved by the invention would have been non¬ existent in this case) . Consequently, when transcribing in the case of the present example a signal is inverted in the control assembly of the second tape machine - as explained in more detail hereinafter. The tape machines are started and are driven synchronously during the start period found available and synchronism is achieved, where¬ after the video signal is transcribed at the correct time in troublefree fashion,all control signals in the video signal being correct. The procedure then continues with a new scene, the tapes being adjusted to respective posi¬ tions corresponding to six seconds prior to cutting, calculations are made, the tape machines started and ^"' transcription initiated, etc.. The sequence between transcriptions from original to copy can be effected advantageously while avoiding long search times, since one

has complete freedom of choice in the order in which copying is made. When all material to be copied- from the first original tape has been transcribed, the procedure - is continued with the next original tape, which is placed in the second tape machine.

We now arrive at the modification necessary in the play-back machine, the second tape machine, in accordance with the illustrative embodiment, in which a machine designated AMPEX VPR 2B is used. Reference is made in this respect to the manual entitled "VPR 2B Theory and Maintenance 1809477".

In order to be able to apply the method according to the invention it must be ensured that the phase of the signal corresponding to half the line frequency is in- verted. Figure 4, which is taken from Figs 2-54, Color Frames" of the aforesaid manual, shows part of a block schematic incorporating a reference generator which 'receives on its input horizontal reference signals (one per line) which are divided by two in a flip-flop or like device. As will be seen from the Figure, this flip-flop has two outputs which can be selectively coupled through a switch J3. In principle this switch can be used to effect re-phasing in accordance with the invention by hand. It is more natural, however, to effect this re- phasing automatically, which can be accomplished by replacing an integrated circuit 74 LS 157 (Texas Instru¬ ments), which can be controlled with an external signal from a microprocessor, in which case said processor can also be used to carry out the aforesaid "parity calcula- tions". The principle establishment is thus the same as that illustrated in Figure 3.

By then changing the phase of the 7.8 kHz-signal at an "erroneous frame sequence", the inventive effect is obtained by which it is possible to carry out the splice types C _j and C_ set forth in Table I.

This operation is carried out in the machine in which the original tapes are read-off, this machine being

to to to to .-_ __. tn O tn o tn o tn

o ro ft

3" & ro 0 ti P. μ-

3 3

Φ ro

P 3

3 rt

01

^

3 ^*

01 o β

Ω z ro

P <! ro

P H to _^ o tr o β

0 rt

< ro Ω

3 P rt 3 μ- o tr

3 ro

P y TJ β rt rt μ-

3 μ- ro 3 rt p. o ro y ro

P Hi

Hi ro

Ω Ω μ- rt ti

Ω Z β y- μ- et rt P

01

_* _

whereby it is possible, although with less simple expe¬ dients, to carry out the invention without needing to digitalize the video signals.

Although the detailed example hasonly illustrated a single embodiment adapted to the PAL-system and to a given conventional tape machine designed for C-standard tapes, it will be obvious to those skilled in this art that the method can also be put into effect with the aid of other types of machine,- and used for recordings in the SECAM-system. The method is probably of minor interest in respect to the American NlSC-system, since this system uses a frame frequency of 30 frames per second and consequently, as a result of the absence of any direct compatibility with the cinematographic frequency of 24 frames per second, uses field (half-frame) splices, where¬ with the problems encountered are quite different.

It is a fact, however, that in the USA films are recorded at -30 frames per second to obtain an adaptation to the 30-frame system of television when transcribing to video, wherewith the same problem is met and the invention can also be applied to the NISC-system.