BBC Engineering September 1973
From Manual Splicing to Time Code Editing
G. R. Higgs, B.Sc(Eng.)
Video Tape Editor. Television Recording Department
The development of video tape editing from physical splicing of the tape to electronic editing of programme material was straightforward and understandable, but the continued development of electronic editing is perhaps less straightforward. The aim of complete control and total frame accuracy has now been achieved with the introduction of time code as applied in the most recent equipment, and this has inevitably led to greater complexity. Time code has, however, other uses and is applied over a wide range not necessarily concerned with editing. Developments in video tape editing systems are still in progress with an ever-widening choice of hardware appearing on the market, and this article traces the course of these developments in the Video Tape Section at the BBC Television Centre in London.

Physical splicing of tape
Video tape recording of television programmes in the BBC dates from about 1958. The early requirement was rapid turn round of material (e.g. for sports programmes covering more than one event) or recording of programmes at times more convenient than their scheduled transmission hours.
Naturally it was not long before the desire to modify the existing recording before transmission initiated the first steps in 'editing' of the video tape. These first edits were made by cutting and joining the tape during black level - typically a fade down (edit), and fade up - and were not undertaken lightly. With the crude facilities then available an edit could take up to half an hour and was an irreversible operation similar to cutting film negative.
Techniques and equipment improved and editing a picture to make a 'camera cut' became commonplace. On video tape, the sound and vision are physically displaced by 15 frames (0.6sec) with the vision head preceding the sound head as the film gate precedes the sound head in a 'comopt' or 'commag' film projector. The 0.6sec of sound track preceding the splice must therefore carry the sound associated with the picture following the splice; this section of track can be transferred by copying it on to a separate ¼ in. tape and re-laying it on the sound track of the video tape preceding the cut. This is possible because a broadcast video tape machine can record 'audio only' on the sound track while replaying the vision.
This facility has other uses as may be seen later.
There was a period of consolidation and extension of these techniques to a point where complex operations were undertaken by physical cutting and splicing of the video tape. Programmes began to be written, planned and recorded for editing to create production transitions not possible with 'live' or 'non-stop recording' productions.
Physical edits are butt joins made using a specialised metal­foil adhesive tape about ¼ in. wide and only 0.00025 in. thick, positioned across the back of the join. With modern microscope splicing blocks, joins were regularly made with normal accuracies of about 00005 in., positioned during the field blanking period of the recorded signal.
As manufacturers improved the quality of the machines, the video head assemblies became increasingly more delicate and sensitive, and the thickness of the splicing tape at physical joins now became a significant factor. Good edits would replay satisfactorily but problems were encountered with their ability to withstand many playings over a period of time. Slight disturbances were also sometimes noticeable on copies of programmes at points where the original had appeared to replay an edit quite satisfactorily.
The advent of colour increased demands on the system and reduced tolerances still further.
During this period some experience was being gained with an electronic editor fitted to a monochrome machine. This was very much aided by the interest of one or two drama directors keen to make it work and willing to accept the inevitable development problems and set-backs.

Electronic editing
A normal machine switched to record at a given point during replay produces a totally unacceptable disturbance (loss of vision, syncs and colour) for anything from 2 to 10sec. If it can be made to switch from replaying to recording in such a manner that the resulting transition appears as a camera cut, an electronic 'in' edit will have been made. The device added to a machine to achieve this is called an Electronic Editor (Ampex) or Electronic Splicer (RCA), and it controls the transition in such a way that erasure, servo switching, signal switching and audio changeover take place in the correct sequence at precisely the correct times to produce on the tape a 'synthetic' camera cut. The process can be reversed to pro­duce an 'out' edit-a transition front recording to replay of material already recorded on the tape.
Electronic edits immediately dispose of:-
(a) splicing of tape and associated tape handling problems,
(b) the problem of staggered vision and sound heads. Since the electronic edit is time-controlled the vision and sound
changeovers can be initiated simultaneously.
The electronic editor demands first-class performance from virtually all the machine's sub-sections: the signal system, audio system, control system and particularly the servo system. It can be set to control both sound and vision together or 'vision only', leaving the sound unaltered.
Positioning the Edit Manually
Electronic editing is usually carried out during the copying of the originally-recorded material on to a fresh tape so as to assemble it in programme order. Editing from live sources or others not on tape is done, but not commonly. Timing the incoming material is normally achieved by setting back the editing and replay machines the same time from the chosen transition points and starting both simultaneously.
The Electronic Editor has a turn on (or turn off) sequence lasting 15 frames. When an 'in' (or 'out') edit is initiated the actual transition on the tape occurs 15 frames later. On an 'in' edit this arises because the existing material must be erased from the tape before new information can be recorded and the video erase head is situated 15 frames upstream from the video record/replay head assembly. Thus initiating the 'in' edit immediately turns on the erase: the wiped tape moves on, and 15 frames later video recording - and sound if required - begin when the wiped tape reaches the recording heads. For an 'out' edit the erase is turned off immediately but the record,/replay head must go on recording until the 15 frames of wiped tape between the two heads have been filled in, and the record/replay head switches to replay at the moment when the 'old' recording reaches it.
Early electronic edits were positioned by marking the back of the tape 15 frames before the required edit point and pressing the record button as the mark reached the vision head. A rather hair-raising experience at the time but nevertheless sequences such as complete intercut telephone conversations were edited in this manner! A rehearsal of the proposed edit could only be made by observing two monitors or by switching the monitoring at the proposed point. After a satisfactory 'rehearsal' the process was repeated but with the record button pressed 15 frames early!
Control of the Electronic Editor

The first 'programmed' electronic editing on BBC video tape machines was 'Editsure'. Both machines were set back about 20sec from the desired points and started simultaneously. Editsure counted reference frame pulses and after a count of 485 started the Electronic Editor. The 'in' edit then occurred exactly 20sec from the start point. The Editsure control panel contained 'Play', 'Rehearse' and 'Edit' buttons. If 'Rehearse' was pressed rather than 'Edit' then only the monitoring (vision and sound or vision only as required) was switched at a count of 500.
This system gave repeatability on each run-up (within the I or 2 frame tolerance of machine lock-up) and was a great step forward. The system also provided that when the Editor was in the 'Vision only' mode it was possible to initiate' sound record' manually. Artistically this was a great improvement. Many of the problems of VT editing are concerned with the sound track and there are numerous occasions when the desired sound transition point is not coincident with the vision. 'Out' edits could still only be made by pressing the editing machine 'stop' button whereupon the 'out' edit occurred 15 frames later.
Editsure was extensively used, particularly on the 405-line system. On this system the video tape cue track was not available because the tape width was required for video information. This arose due to the longer line time of the 405-line system (98usec as against 64usec) coupled with the poorer stability of the video head assemblies and associated servo systems. With the change to 625-lines and more stable machines the cue track became available for clean effects, guide tracks or control signals for electronic editing purposes.
The device called Editec, which was marketed by Ampex Corporation, was the first commercially-produced means for controlling the Electronic Editor to enable the edit point on the edit tape to be made precisely on the desired frame. Editec achieves this by storing edit-point information in the form of edit cues recorded on the cue track. It is important to realise that Editec controls the Electronic Editor; it does not directly control the machine to make the edit. The cues are l0msec bursts of 4kHz tone (i.e. 40 cycles at 4kHz) and are keyed onto the cue track by pressing a cue button. They are recorded by the normal cue track record/replay head at a time coincident with the required edit point. Replayed by this head they would be useless for initiating Editor turn-on since the edit would occur 15 frames later. They are therefore read off by an auxiliary 'cue read-off' head situated 'upstream', just be­fore the video erase head assembly. This is the only practical position but means the cues are now read off 33 frames before the point in programme time at which they were recorded. Therefore after cue read-off Editec counts 18 frames then initiates the Editor turn-on cycle. Fifteen frames later the edit occurs precisely at the required frame, that is just as the edit cue reaches the cue track record/replay head.
As a by-product of Editec having to 'mark time' for 18 frames, the edit point can be altered to occur earlier than the time defined by the edit cue, without changing the cue position on the tape. If Editec is set to minus 7, it counts only 11 frames before starting the Editor, the edit occurring 26 frames after cue read-off instead of 33, that is 7 frames early. The maximum shift possible is obviously minus 18 frames. For operational symmetry Editec can also delay the edit up to 18 frames. There is of course no technical reason why the delay cannot be as long as required.
With Editec a second cue can control the Editor turn-off in such a way that a new piece of vision (and sound if required) can be 'inserted' into existing material.
Editec has various facilities. Rather than merely altering the edit point relative to the cue one can actually erase the cue and record a new cue at the new edit point selected. Cues can be selected out of a series of cues, selectively erased, or made to operate external equipment, and the machine can be set to record automatically a sequence built up from any number of 'takes' of anything from 1 to 36 frames duration. After the machine stops and is rewinding for the next 'take', the picture determining the vision input to the machine is altered to pro­duce the required effect in the 'animation'. As each pass takes about 20sec, a I min animated sequence made 8 frames at a time takes just over an hour. If the machine is manually re­started each time, of course the time taken becomes indeterminate.

Time Code
In 1968 time code made its first appearance on trial with BBC video tape machines. Time code is a digital signal containing 80 bits per television frame coded as bi-phase mark and re­corded on the VTR cue track. The code uniquely defines (or addresses) every frame on the tape in hours, minutes, seconds and frames. The signal originates from a time code generator and can be distributed to any number of machines. The generator is normally locked to the colour subcarrier of a pulse chain and should strictly be recorded only on machines recording vision derived from the same pulse chain. The generator normally produces real time, resetting at 24.00 hours.
Time Code Editing
Each VTR to be controlled has a programmer with 'start' and 'stop' memories which may be used to control the tape trans­port or (if fitted) the Electronic Editor. The time code used can be the machine's own cue track time code, another
machine's cue track time code or any external time code including the distributed 'time of day' code.
To choose required points the programmers have a display 'Hold' control that 'freezes' the display of the time code being used for as long as required. Facilities are provided for the display of time codes stored in 'start' and 'stop' registers.
On Time
The fundamental difference using time code is that continuous information is available. Circuits can 'foresee' requirements rather than being 'taken by surprise' by the sudden appearance of a command pulse (such as an edit cue). This means that logic systems can detect inconsistent commands and conflicting requirements, and therefore can either indicate these errors or refuse to operate on them, as appropriate.
The first trial system was called 'On Time' and used pre­recorded time code on the editing VTR to: I. control the Electronic Editor 2. control the starting of the replay machine. The programmer on the editing VTR replaced Editec and controlled the Electronic Editor, while the replay machine programmer controlled the replay machine tape trans­port.
The edit point address was entered in the 'start' register of the programmer of the editing VTR and an address say, 10sec earlier in the 'start' register of the replay machine programmer. This machine was then parked manually 10sec before the required transition point. The edit machine was run from a point earlier than 10sec before the edit point, and both programmers were fed the replayed cue track time code from the edit machine. At 'minus 10sec' the replay machine started and locked up under its own servo. Achieving the right 'in' point to the replay tape depended on accuracy in parking the tape at a point 10sec upstream. If the replay tape also carried a time code, accurate parking could be achieved by entering a time 10sec earlier than the replay tape code at the required point into the replay programmer 'stop' register. Playing the tape from before this point would automatically stop the machine at the right point. The replay programmer was then switched to edit-machine time code and the operation continued as for 'manual' parking.
'Out' edits were made by entering the correct code in the editing programmer 'stop' register. Rehearsal was achieved by pressing the programmer 'operate' button only, while an edit was made by pressing the edit machine record button when it had locked up.
A programme control determined whether audio-video or video-only edits were made. 'On time' could be used to control tape transports from 'time of day' code to start and stop machines at precise times,a second machine could be started from the time code of the first machine for changeovers accurate to one frame.
Precise control of edit points
Both 'Editec' and 'On time' suffer the same disadvantage that even with very accurate parking of the replay machine the first frame recorded on the edit machine may commonly vary by ±2 frames. 'This is due to the randomness of lock-up of both machines relative to the same reference sync-pulse chain. In the PAL system this relative lock-up can only alter by multiples of 2 frames, whereas when editing 525-line NTSC, increments of I frame may occur. Thus despite the exact frame accuracy of the edit point on the edit tape, the 'incoming' material does not match this precision. Where 'simultaneous start' is not used and one machine, already running, is used to start the other, there is still a ±2 frame tolerance on its lock-up with respect to reference frames. The relation of the replay to the edit tape on 'in' and 'out' edits is shown in Fig. I


The manufacturers of 'On time' (EECo - Electronic Engineering Company) developed a system whereby the replay machine is control led during run-up to ensure that both replay and edit machines reach their selected frames at the same time. With hindsight this may seem a rather simple, fundamental requirement but its realisation involves logic of considerable complexity. Its eventual development can be ascribed to:
1. The demand for ever-increasing accuracy of video tape editing.
2. The availability of cheap, small, low power logic systems (i.e. digital, integrated circuits).
As with 'On time' each machine has a programmer but its facilities are much extended. The edit programmer controls not only the Electronic Editor but also the edit-machine tape transport. The replay programmer controls the replay machine transport logic and has override control of the replay machine's capstan speed.
It is necessary for time code to be recorded on both edit and replay tapes. The selected edit point is entered in the edit programmer 'start' register (the 'out' point is entered in the 'stop' register if 'insert' is required). Pressing 'Recue' parks the edit machine 20sec, 00 frames before 'start' time. The selected 'in' point on the replay machine is entered in its programmer's 'stop' register. With the replay programmer in 'slave' mode - recuing the machine will park it l5sec 00 frames earlier. To relate the two programmers the edit-machine edit point (the edit 'start' time) is also entered in the replay programmer 'start' register. Thus this programmer now knows the edit point:
(a) on the edit tape relative to the edit machine time code.
(b) on the replay tape relative to its own machine time code.
When the edit machine is started for rehearsal or edit the replay programmer is also fed edit machine time code. Having parked the replay tape I5sec before the required point ('stop' time) the replay programmer starts the replay machine 15sec before its 'start register' time. (For this operation accurate parking of the edit machine at minus 20sec is unnecessary).
if the replay machine achieved play speed in zero time and all machines always locked up the same way no further action would be necessary. (That however would make 'Editec' or 'On Time' repeatably exact!) The algebraic difference between the replay programmer's 'start' and 'stop' registers represents the relation required between the two tapes as they approach the edit point. This is a fixed figure and by differencing the two time codes every frame (dynamic subtraction) the replay programmer develops an error signal (i.e. the 'difference' between the required and actual differences). The error signal is used to override the replay machine capstan speed until zero error is achieved, the 'synchroniser' lights a 'sync' indicator and releases control of the replay capstan, and the machine servo then completes fully synchronous lock­up, having been left within less than a field of the correct frame.
Since the replay machine starts from rest when the edit machine is already 'at speed' the synchroniser always has to accelerate the replay tape above speed to catch up. The synchroniser only considers frames and 'frame differences'; if an error of one or more seconds is introduced this will be ignored and only the correct frames synchronised. As the tapes are parked accurately by the programmers this should not represent a problem.
If programmed to synchronise frames of opposite PAL switching a 'PAL ERROR' indicator will light. It is necessary to change either the edit or replay tape 'in' point by one frame. If ignored the selected frames will be synchronised but on release of capstan override the replay machine servo will re­frame to a correct PAL frame. This will give an 'in' point error of ±I frame.
It can be seen that the replay programmer is dealing with two different time codes to control the tape replay whereas the edit programmer uses only one to park the edit tape and control the Electronic Editor.
The edit tape either has time code pre-recorded or continuous time code may be built up during editing by use of a second time code generator. This generator slaves itself to the edit machine during replay allowing the assembly of a continuous time code on the cue track as the programme is assembled on the sound and vision tracks. If preset to zero at the beginning of programme it gives immediate and continuous 'programme duration so far' during editing, a very useful asset.
The discerning reader might ask why the same information is entered in both programmer 'start' registers when the edit programmer might start the replay machine at l5sec before the 'start' time. Although possible such an arrangement would reduce flexibility. Firstly both programmers are identical and can be interchanged if necessary. Secondly the 'start' time in the edit programmer may be altered without altering that in the replay programmer. This changes the transition point without altering the relationship between the two tapes. Thirdly the system can be expanded to include more than one replay machine and programmer. In this case separate replay programmer 'start' times determine which machine starts and when.
The BBC Time Code Editing installation also includes the option of an auxiliary programmer associated with the edit machine. This has 'start' and 'stop' registers and allows 'staggering' of the edit by assigning the vision edit (say) to the main programmer and sound edit to the auxiliary. Simultaneous vision and sound, vision or sound only use only the main programmer with auxiliary 'off'. Less critical staggered edits can be made by manual operation of the sound record circuit.
A block diagram of the system is shown in Fig. 2

External operations during editing
Times stored in all the various registers and those in four pro­posed coincidence detectors can he used for starting sound machines, other VTRs not equipped with programmers, video disc or any devices properly interfaced with the existing re­mote start facilities. The flexibility begins to multiply rapidly with introduction of further equipment.
Vision mixing facilities are also installed and of course require a minimum of two replay machines to mix pre­recorded material. As yet only one machine (which is used for replaying) is equipped with a programmer; the second is started either manually or by one of the other programmers which may in turn use either edit or replay time code. This means that accuracy is only ±2 frames and some forethought and juggling is sometimes required to arrange that the critical material is on the programmer-equipped machine.
A second Time Code editing complex is to include automatic vision mixing. This has a time code initiated mix/wipe with manually pre-set rate and provides
(a) repeatably exact mixes
(b) simplification of simultaneous vision and sound mixes. Provision is also to be made for later inclusion of a second replay programmer.
The table summarises the degrees of accuracy with which the various editing techniques, mentioned in the preceding pages, can determine the last frame replayed, and the first frame recorded, on both 'in' and 'out' edits.

Other Applications of Time Code

With machines recording 'time of day' precise logging may be done by anyone, anywhere with an accurate (preferably digital) clock. A logged time may be entered in a replay programmer and in 'search' mode the machine will be spooled and stopped at this required point. The time of day logs are useless on machines not equipped with time code apparatus and therefore all other machines (initially the replay machines of 'Editec pairs'), are being equipped with time code readers ('character generators'). These give a display in the form of figures inlaid in vision and one unit is required per machine. Already two 'pluggable' units are in much use for other operations.
Helical scan recordings
Increasingly these are made either during recording or by subsequent quadruplex-to-helical dubbing with time code display inlaid in vision on the helical recording. Detailed study and logging of these recordings can be made and the information brought to an editing session can considerably speed-up the rate of work.
Sound dubbing
For adding large amounts of additional sound material to an edited programme post-edit sound dubbing is usually employed. Originally this consisted of lengthy rehearsal, dubbing in sections to another VTR, and relaying in sections using vision as syncing material, all requiring expensive time on broadcast VTRs. Presently the rehearsal is done to helical scan recordings with inlaid time code recorded from the un­dubbed, edited quadruplex master. If this is an 'Editec' edited programme, time code is recorded on the master tape cue track while dubbing to helical scan. The sound dubbing is done using the master tape with the time code inlaid on the vision monitoring.
Since this is the same as that which all the rehearsals have been made accurate cuing is possible and time is saved. The same time code is copied to the cue track of the inter­mediate sound copy and relaying on the master in sections is greatly speeded up by the use of a time code comparator. This unit displays two time codes and their algebraic difference, enabling one machine to be brought into precise synchronism with another rapidly and positively on non programmer equipped machines by manual capstan override.
A proposed sound dubbing system is for the edited master sound track and time code to be transferred to two tracks of a multi-track sound recording in addition to a helical scan re­cording also containing the time code on its cue track. All sound dubbing can then be done using the other sound tracks, time code being used to keep sound tape and helical scan in synchronism. A final mix track can be transferred back to the master video tape once again using the time code to keep the sound machine in synchronism. This greatly reduces broad­cast VTR time and also makes possible 'track laying' during the editing stage if required.
Intermediate dubbing and editing on film
A technique sometimes employed is to dub video tape sequences to film recording and then use film editing facilities to edit the vision as well as mix and dub the sound. This is then re­dubbed from telecine to videotape. The monochrome vision is replaced by insert editing, vision only, from the original colour video tape. The process is cumbersome and tedious and not to be encouraged, but it may be made more efficient if the original material has time code on the cue track. This is transferred to the optical track of the film recording and is pre­served in segments during the film editing. When retransferred to video tape it can be used to define the exact segments of the original material that have been retained.
Foreign language dubs
Time code has been used to expedite the production of foreign­language masters. When French masters of the Six Wives of Henry VIII were wanted, the programmes were dubbed to film recording with their time code copied on the optical track while a 'sepmag' of the English sound was made as a guide track, The French organisation used full film dubbing facilities for 'looping' carefully-timed and prepared translated dialogue, building up a French master 'mag track'. The completed track could be checked against the film recording and then both returned to the BBC. Using time code the un­touched film and mag track were run in sync with the original master to produce a French-language master tape. Using this technique no VTR capability is required of an organisation specialising in the complex art of preparing foreign-language sound tracks that endeavour to marry up durations, of lip movements.
Carefully-applied time code can be used in conjunction with film recording to solve many problems in sound and vision that arise when handling material received from out­side sources. This is particularly so when old and/or non­standard speed films and sound dubbing are both involved.

Sundry Problems

Initially trouble was experienced with breakthrough on some machines between cue track and sound track due to the high distribution level required for the time code (approx. + 6dB), but modifications solved this problem.
Reading when spooling
The machines must be equipped with wide-band cue track replay amplifiers if they are to produce a usable output during fast spooling rates of up to 300 i.p.s.
Inclusion of all picture sources
Time code represents a conversion problem in that to make full use of it all likely sources on different reference sync-pulse chains (regions, Outside Broadcasts, etc.) should be equipped with generators to avoid post-recording of time code. Readers must be provided to make use of the code and avoid perpetuating two logging systems.
Discrepancy between mixer and electronic editor cuts
The inability to edit at the same point as a camera cut is beginning to be annoying as the accuracy of editing increases. This problem arises because BBC vision mixers cut in field blanking preceding field one and video tape electronic editors on 625 lines edit in field blanking preceding field two. Thus one must lose at least half a frame on each side of two camera cuts when editing different takes together. As the tapes may only be shifted relatively by two-frame increments this can give rise to unacceptable music edits. Normally a half frame of one or other unrequired shot must be left in to make the music edit work, and the vision cut is only 'dirty' to the trained eye. (The video disc is invaluable in discovering exactly what is happening on these occasions.) The problem is made worse in that these cuts can only be detected after the edit is made because the monitor `bumps' during the rehearsal or take, making it impossible to see the 'rogue' field. The bumping is caused by:
(a) The switch in the machine electronics (taking place at radio frequencies) being demodulated as a spurious signal. (b) The over-sensitivity of the decoders and monitors used.
(c) inability to feed the monitor with external syncs since vision is delayed 9.5µsec with respect to reference sync. This is due to requirements for electronic editing. (It is hoped to experiment with the use of external syncs delayed by 9.5µsec for feeding the monitor).
Films with frames cut out
Problems have arisen with transfers to film recording in that on occasions when the vision should remain 'inviolate' for future resynchronising it has been tampered with making subsequent redubbing etc. difficult
Spool inertia
When time code editing with the replay machine near the end
of a 90-minute tape (when the take-up spool is nearly full) the starting acceleration of the tape was often not great enough so that the synchroniser was I sec out. This of course caused the replay machine to reframe after 'sync' was achieved and gave a rehearsal or edit in error by 24 or 26 frames. This has been overcome by a modification to the synchroniser logic. On 25 f.p.s. the error detection logic was asymmetrical and gave greater range on the `ahead' than the 'behind' portion of its I sec range. As the replay tape is virtually always 'behind' and only 'ahead' due to overshoot the proportions were reversed. The 5 frames gained was sufficient to overcome the problem. Reliability was further increased by removal of an inbuilt 3-frame delay before starting the replay machine, a total of 8 frames gained.
Erasure of programmer information
Entering information in the programmer registers erases previous information. However, plans are in hand to provide print-out of all times that one might wish to recall.
Ergonomic shortcomings
The ergonomics of the equipment is poor and this makes it difficult for an operator's handling of the controls to become automatic. When sufficient experience is accumulated it should be possible to design our own, custom-tailored time code editing equipment.
The very great flexibility of a full-scale time code editing suite tends to result in the complexity of work being expanded rather than the same work done in less time. This is not wrong when the facilities are used creatively to achieve results that could not be realised more realistically in other ways or places but as in all stages of television production the crux of efficient video tape editing is good planning. Ever-widening options coupled with unresolved production requirements demand an excessively high decision-making rate from the video tape editor if reasonable quantity as well as quality of output is to be maintained.
Advantages of Time Code
Time code provides a 'tagging' or addressing of programme information that is fixed in relation to the recorded material and unambiguous, This allows remote logging and decision making, precise duration calculations and exact programme timings to be made.
Accuracy of both edit points
The accuracy of selection of the required point to join the replay material can now be made to the same standard as that to which the edit is made and is also achieved at the first rehearsal.
The repeatability of rehearsal and edit are of very great ad­vantage. Sound work can be very exactly and rapidly repeated, altered or extended during editing.
Reliable synchronisation
The ability to run two VTRs in reliable sync, not only for post sound dubbing but also during editing means that mixes can be 'picked up' at the end of long sequences without the need for long-winded redubbing or unreliable synchronous run­ups.
No physical marking of tape
In an endeavour to achieve repeatability, 'random run-up' editing requires physical marking of the tape with 'china­graph'. The elimination of this requirement increases the likelihood of a tape being reusable.
Emphasis on programme requirements
With automatic recuing, the chosen edit points (not necessarily those which prove artistically satisfactory) can be achieved without 'trial' rehearsals, and centralised control of the whole editing process becomes more civilised. The operational requirements while being more complex are shifted to the programming stage of the operation and allow more concentration and attention on the editing of the programme material.
In development is an application of time code to improve operation in editing pairs. These channels have two machines, are equipped with Editec, and can be operated by one man as either a record pair or an edit pair.
During editing considerable tedium is involved in constantly recuing two VTRs and a sound machine. With Ediplace the 'in' point on the replay machines will be chosen by entering the time code address in a register. The machine will automatically rewind and park itself 13sec earlier. The editing machine will be parked manually l0sec before the edit point as at present. After starting the replay machine the edit machine will be started at a time of I0seconds: 00 frames earlier than the chosen point in the register. This reduces the `parking variables' to one, makes marking of the replay tape unnecessary and speeds up recuing. There will be a facility to adjust the stored 'in point' time by plus or minus a number of frames.

Use of Video Disc in Editing
The facility of slow motion (forward or reverse), freeze and (with certain limitations) up to double-speed replay from the video disc make it a powerful addition to video tape editing capabilities. The BBC Video Disc control logic has been extended in such a way that it can be operated by external command signals. These are distributed by the local remote control facilities system. With Editec this means starting the disc at any pre-selected speed by commands originating at either cue read-off or edit-point time. Start time can be varied over a range of up to 10seconds by a variable delay incorporated in these modifications.
A time-code editing channel can, by nature of its equipment, originate commands through the remote control system at any time, independently of edit points.
A second development of the BBC video disc is a 'field counter'. The fields recorded can now be precisely selected and the video disc can always be recued accurately to any one of the 1800 fields (36sec at normal speed) recorded. These features mean that the disc can match the accuracy of time code, making possible repeatably precise tape/disc edits. The facility to make subsequent changes in the disc operational mode is now under development. These commands maybe originated from an editing channel or the field counter itself to reverse or freeze.
The channel also has a black edge caption facility which might be extended from its present manual 'cut in', 'cut out' operation to external control as above.
Imaginative use of the video disc and facilities allied with mixing during video tape editing allows a wide range of effects and sequences to be achieved.

Future Possibilities
As stated earlier the 'continuous information' aspect of time code gives rise to many benefits and may be expected to give rise to more in the future.
Already centralised time-of-day code can be used to start programme transmissions precisely. In the United States time code is used to control programme junctions with automatic slide changing, mixing, overlay, TK & VT running etc., to create complex presentations far too involved for manual, real-time operation.
Systems available now dispense with pre-wired programmers and fixed options by using computer control of equipment to instructions given by keyboard and light pen on a visual display unit. The computer editing programmes are on computer tape, thus extending operations means only playing-in an extended programme. The unit can ask for decisions and point out errors as required. Additionally all edit decisions made can be stored and recalled. This at present is the limit of 'on-line' editing.
'Off line' editing involves making programme decisions using low-cost equipment such as helical scan machines or monochrome `disc packs'. The decisions made are stored on punched paper or magnetic tape that is used later to control assembly of the programme on broadcast VTRs. This moves the decision making time off broadcast VTR channels. The disadvantages of having to pre-load these units with material, the obvious need to rehearse or simulate completely the intended edit coupled with the great flexibility to which producers are accustomed (particularly with respect to sound track editing) give rise to much duplication of effort and equipment. They are perhaps more suited to short, intensively­edited sequences, such as titles, special effects, and commercials.
Video tape editing can be seen to be reaching a stage where production requirements are no longer restrained by equipment capabilities but by equipment capital cost. Limitations are no longer intrinsic in the medium; only in the cost of the time which it is realistic to expend on a programme.
The development of ever-improving small, light, high quality cameras, recorders and less expensive format machines promises to widen the horizons of electronic programme production. It seems likely that video tape editing in the BBC will evolve further in the mixture of 'on' and 'off' line techniques that has already begun. This suits the nature of our requirements and makes possible many specialised services and flexible combinations of these to master the occasional unique problem' programmes.
Inevitably video tape must cease to be regarded only as a technical production facility and come to be considered a creative medium.