The Romans and directorial sleight-of-hand
Doctor Who’s 1960s episodes are often referred to as having been recorded ‘as live’, in order, from beginning to end with few if any breaks or retakes. This is only half true. Although that was the ideal method of production in the first few years of the series, when recording time was severely limited and editing difficult and expensive, it was more often an aspiration than a reality. Most episodes used scheduled breaks – usually only a few but, as early as April 1964, Marco Polo’s final episode had around ten planned breaks. Retakes were also common, as seen in both versions of the first episode, An Unearthly Child, for example.
All that said, recording did generally occur in story order so that the post-production editing involved only the snipping out of the wasted tape at the points a recording break occurred, and not complicated reordering of scenes. But still, there were exceptions from very early on in the series, with brief sections recorded out of sequence for The Daleks, The Edge of Destruction and Marco Polo. Slightly later examples include The Dalek Invasion of Earth.
However, the first episode to have a substantial section recorded out of order was Inferno, the fourth and final part of The Romans, directed by Christopher Barry. This was recorded in January 1965 for transmission just a few weeks later in February. Barry scheduled recording in five chunks, ie meaning with four scheduled breaks or run-ons.* The first chunk comprised the last four minutes of the episode, starting with the Doctor and Vicki arriving back at their ‘borrowed’ villa and concluding with the roller captions. There was then a break following which the rest of the episode was recorded in order, starting with the title sequence and ending with Ian and Barbara larking about on their return to the villa, preceding the Doctor and Vicki’s arrival. We’ll return to these two scenes in a moment.
As was usually the case with any deviations from story-order recording, the reason here was not artistic but pragmatic. Even though the series was now recorded in Riverside studio 1, which was notably bigger than the series’ initial home at Lime Grove studio D, space was still inadequate where multiple large sets were required, as here. The camera script gives the instruction “STRIKE TARDIS – SET OPENING” to be actioned during the first recording break. Essentially, the TARDIS interior set was taken down and another set put up in its place. The break also enabled William Russell and Jacqueline Hill to change out of their slave costumes and into more elegant Roman robes for the final scene. A break would have been needed for this costume change even if the episode had been recorded in order, so using out-of-sequence recording did not add to the editing cost.**
There is no existing studio floor plan for this episode, so we don’t know which set was erected in place of the TARDIS. A literal reading of the camera script instruction ‘SET OPENING’ might suggest it was the set needed at the start of the episode, which was the fighting arena, but it could have been any of them. Most likely it was whichever set was easiest to put up quickly given this was taking place within the already restricted recording time, although the camera script does specify this would be a long break. The fighting arena is a large multi-level set with a floor area filled with sawdust and I doubt it would have been practical to create this in a recording break. I would speculate the newly built set would be one of the smaller areas of Nero’s palace seen in the subsequent scenes.
The recording break ostensibly comes between the two villa scenes with each party arriving back after their adventures in Rome. However, Christopher Barry employed a neat bit of sleight-of-hand in his recording schedule, because the break doesn’t fall exactly between these scenes. The two scenes are joined by a dissolve to indicate some passage of time and this may seem the obvious point that the two sections of recording would have been edited together, but it isn’t. It was impossible to create a dissolve in post-production as editing was done by physically cutting tapes, which precluded the merging of images. This could only be done ‘live’ in the studio recording itself by mixing between the output of two different cameras.
Instead, rather than editing the recording directly between these two villa scenes, Barry does so one shot into the second scene. The recording started on the second villa scene (scene 29 in the camera script) with the medium shot of the newly arrived Doctor and Vicki, not the immediately preceding shot of the Doctor’s hand picking a grape from the fruit on the table which opens the scene in the finished episode. Although dramatically continuous with the rest of this scene, this one brief shot was labelled scene 28 and recorded at the other end of the session.
After scene 29, recording continued unbroken with the pre-filmed model sequence of the TARDIS dematerialising, followed by scene 30 in the TARDIS, meaning the actors had only a brief time to move between the villa and TARDIS sets. William Hartnell and Maureen O’Brien leave scene 29 before its end so had a head start. Hill and Russell have the few seconds given by Barry’s shot of the fountain closing scene 29 plus the scheduled 28 seconds of the following film insert. Scene 30 closed with the end of episode roller captions, at which point recording broke for the change of sets already described.
Recording then went back to the start of the episode from which point it was performed in sequence. There were three planned stoppages (one break and two run-ons) to adjust the set and change camera positions. However, additional stoppages occurred as Barry chose to retake three sequences (there were several retakes on all episodes of the story). Recording continued until scene 27, the first of the villa scenes. As this scene ended, Barry recorded the dissolve from Ian and Barbara to the close-up of the Doctor’s hand picking up the grape, which neatly bridges into the scene already recorded at the start of the session. In theory a stand-in for Hartnell could have been used for this shot but in practice there was no reason not to use Hartnell. This shot concluded recording for the episode, with the discontinuous sections of recording later being edited together ready for transmission.
This trick of placing a recording break ostensibly between two scenes but off-setting it by one shot to enable a dissolve between scenes was later used twice by Michael Imison on The Ark in 1966. There may be many other examples of it in the series of which I’m not aware.
The fractured recording process on The Romans illustrates that recording methods used on the early stories were not quite so primitive as fans sometimes think, and highlights the degree of intricate planning needed for the production team to create what may seem like simple effects (a scene transition in this case) with the limited technology available to them.
Images © BBC
Sources
Camera script and production documents for The Romans from the season 2 bluray set.
Thanks to David Gillespie-Pratt for assistance
*A brief note about terminology: A ‘recording break’ was when the studio action stopped and the video recording itself was stopped. A ‘run-on’, also misleadingly called a ‘recording pause’, was a shorter stoppage where the studio action paused but the video recording was left running. For the purpose of this article, the difference is immaterial.
**A charge of £60 per recording break was levied on the production as a contribution to new videotape as each break meant physically cutting the tape in editing, making it unsuitable for reuse.