BBC Video's tradition of issuing a special VHS boxset to celebrate the show's November anniversary continues in 2002, with The First Doctor Boxset, containing one previously unreleased story from each of the first three seasons.
The stories contained in the boxset are 'The Sensorites', 'The Time Meddler' and 'The Gunfighters'.
'The Sensorites' and 'The Gunfighters' exist as film recording negatives in fairly good quality. All ten episodes were transferred to digital videotape on the Spirit telecine with a light DVNR to remove the smaller video problems prior to sending them up to Peter Finklestone for deblobbing. Episode four of 'The Gunfighters' suffered from vertical instability and jitter caused by a faulty clawbox mechanism in the film recorder camera. This episode was digitally stabilised using a Snell & Wilcox 'Shakeout' realtime image stabiliser before being sent off. Both of these stories were treated with Peter's VidFIRE videoising effect to try to recreate their original studio video look as much as possible.
Unfortunately, 'The Time Meddler' is in a much poorer state and so the decision was made not to VidFIRE it at this time - both the quality of the prints and the short timescale mitigate dagainst this. The BBC holds edited prints of the story, returned from a TV station in Nigeria. Additionally, film collector Ian Levine holds a set of unedited prints of the first three episodes, which have previously been wet-gate telecined by the team.
The Sensorites
All 6 episodes of The
Sensorites exist in the form of 16mm film recording negatives, recovered from BBC
Enterprises in 1978. All are stored field recordings with excellent detail and greyscale
reproduction for a 1964 production. The
episodes have minimal levels of film damage and the original 2 videotapes seem to
have been in good condition when the films were made. Despite this, there were still
around 3000 defects to be corrected on each episode, and this was done in 3 passes.
Experience shows that about 90% of defects are spotted and fixed on the first pass, this
increases to 99% with the second pass and, hopefully, 99.9% on the final pass;
unfortunately experience also shows that a handful of small defects will slip through
not bad though when an episode has over 35,000 frames.
Episode three presented a problem,
with a small but very noticeable hair in the gate of the film recorder appearing in the
bottom-right of every frame for about the first 3 minutes. Although it was a laborious
task, this was painted out on a frame-by-frame basis prior to the main deblobbing.
Throughout all the episodes, there
were dirty vision mixer cuts and single frame mixes where the picture changes from one
camera to another and these were all repaired with paintbox techniques.
Episode 6 suffers from the most faults. At one point there is a 2 frame off-lock as illustrated here, along with the repaired frames which were achieved with a combination of retouching and Re-Vision Twixtor to generate replacements:
Throughout episode 6 there are 4
phosphor holes on the film recorder screen. These appear as static black dots on the
picture and are especially noticeable when the camera pans or people move across the
screen. Exploiting the rock-solid stability of the Spirit transfers, these were
successfully removed by recording delicate brush-strokes with Commotion Pros wire
removal tool and automatically applying these to every frame in the episode. Where the
repair coincided with highly contrasting tones, new artefacts were produced, but these
could be painted out as with dirt and sparkle.
Most of the episode openings
dissolve from the title sequence into a re-enactment of the previous weeks
cliffhanger. This means that it was not feasible to use a good quality generic
version of the Hartnell titles as it would be difficult to match the geometry to provide a
seamless join. Consequently each title sequence had to be deblobbed individually, and most
of the episodes feature out-of-phase titles. Oddly enough, the flowing clouds of the
sequence are uniquely suited to hiding the worst aspects of this artefact so they do not
look too bad at all. Episode 6, however, uses
a 35mm film recorded insert of the end of episode 5 for its reprise. Because the quality
is correspondingly lower (being a film recording of a film recording) it was replaced from
the new transfer of episode 5 and the titles were lifted from another Hartnell
episode.
As before, with The Aztecs, the roller captions for the end credits have been electronically reproduced to reflect more closely how they would have looked on original broadcast.
It was noted that there was no
director credit on episode 4 and Rex Tuckers interview in Doctor Who Magazine noted
that he had his name taken off the production when he refused to make further edits. Innes
Lloyd made the edits himself. Archive documentation lists episode 4 as a first edit (/ED
suffix) but the episode as transmitted was at least a 2nd edit (/ED/ED suffix).
Research by Richard Bignell has revealed that in 1967 BBC Enterprises retained the
VT of episode 1 and 35mm film recordings of parts 2-4 for the purpose of foreign
sales. This is slightly unusual in that 35mm was a gauge usually reserved for UK
rebroadcast and 16mm was for sales. Furthermore, the 35mm film recording of episode 4 was
still known to be in existence in 1976 (and might well have been the source for the clip
used in William Hartnells obituary on the BBC news in 1975). Episode 2 in particular is of low quality, very
noisy, with lots of artefacts printed in, consistent with printing from a dupe neg;
episode 4 will also be at best a reduction print from the 35mm. The exact details of why
the prints were made and when they were junked remains a mystery.
Episode two required the most
work, with 5 passes required to remove damage and artefacts. It wasnt easy to keep
count of the number of repairs, but it was well over 12,000 for the one episode. The other
episodes required around 6-8000 repairs each.
The reprises from the previous
episodes are clearly film recorded inserts, but are different takes so could not be
replaced; presumably they were recorded at the end of the previous weeks episode to
allow for an easier start the following week. Episodes 1 and 4 do not have a repeat insert.
Despite the stabilisation
performed on part 4 by the Snell & Wilcox Shakeout box, there was a one second period
of severe instability, during a film sequence near the start of the gunfight, which could
not be corrected automatically as it lay outside the 2% correction window used with
Shakeout. To increase the Shakeout range to
the next setting (5%) would have resulted in an unacceptable amount of the picture being
lost throughout the episode. This sequence was therefore corrected using the image
stabilising tool in After Effects.
The end credits for each episode of The Gunfighters are particularly poor and barely legible, but have been recreated for the video release.
Both The Sensorites and The Gunfighters have been treated with the VidFIRE process to approximate the original video look of these studio-based stories.
The sound on 'The Gunfighters' was quite hissy, but episode four was particularly poor. This severly distracted from the illusion of the 'video' feel created by VidFIRE, so the soundtracks were all processed by Mark Ayres to significantly improve their fidelity.
The Time Meddler
As noted in the introduction, the prints of 'The Time Meddler' are relatively poor compared to the other two stories in this release. The BBC copies of episodes one, three and four all contain edits, notably the entire introductory sequence where the Doctor and Vicki discover Steven in the TARDIS at the beginning of episode one. Fortunately, collector Ian Levine has unedited copies of episodes one and three, so these were used to patch in the sequences missing from the BBC prints.
There is a large film recorder off-lock in episode one, unfortunately occurring at a point where Hartnell is lunging forward and gesturing. This required much more specialised repair than was available in the telecine suite, so the sequence was given to Dave Jervis in 3D and Video Effects. He repaired it on an Adobe AfterEffects system, using multiple layers of backgrounds, mattes, morphs and motion-estimated video to recreate the original shot.
Cleanup work for the rest of the story was largely done in real-time using the DVNR, with some manual retouching afterwards. Because of the poorer quality of the images, it was decided not to VidFIRE this story at the present time.
Copyright Peter Finklestone / Steve Roberts, 19 July 2002