In late 1999, the BBC2 Controller, Jane Root, decided to capitalise on the BBC's own back-catalogue and place a series of colour 'Doctor Who' repeats into the 6pm 'genre' slot on BBC2, traditionally the home of purchased-programmes such as 'Star Trek' and 'The Simpsons'. Ralph Montagu of the Restoration Team argued strongly for the need to remaster and restore many episodes from Jon Pertwee's era prior to transmission, and in October 1999 we were given access to budget and technical control over preparation of all the episodes for the repeat season.
The initial run of episodes was scheduled to cover all five seasons of Pertwee stories, which at one episode per week is a clear run of around two years! Unfortunately, a decision was made to move on to the Tom Baker era (starting with 'Genesis of the Daleks') after only two complete Pertwee stories have been shown. Subsequent poor ratings for this story caused the cancellation of the repeats altogether. However, we had already carried out some restoration work on the Pertwees and Baker stories and had plans for many more, as outlined below...
Although every episode exists in one form or another, there are many problems arising from the junkings that took part in the seventies. These problem episodes can be divided into four categories:-
These problems would be dealt with in the following ways:-
Pure Film Episodes
The only 'Doctor Who' story to be made entirely on film (until the 1996 TV Movie, at least) is 'Spearhead From Space', Jon Pertwee's first story as the Doctor. 'Spearhead' was shot entirely on 16mm colour film due to a technicians strike at the television studios where it was scheduled to be made. As the opening story of the repeat season, it was considered vital that it should be presented in absolutely pristine form, so it was agreed that the film would be transferred on the best possible equipment.
BBC Resources recently took delivery of a brand-new telecine machine, its first for ten years. The Philips SDC-2000 'Spirit' Datacine is a state-of-the-art CCD film scanner, operating at a resolution of two thousand lines per frame and capable of outputting in a variety of television and data formats. Coupled with extensive primary and secondary colour-correction from a Pogle Platinum and image-processing and cleanup from a DVNR-1000, the Spirit is a machine capable of giving the best possible pictures.
Television Archives at Windmill Road were asked to assess the condition of the existing prints of 'Spearhead'. Initial findings were not good. Of the four original transmission prints from 1970, only episode one was considered to be good quality, with episodes three and four suffering from film warp and episode two declared unsuitable for transmission due to warp and pyhsical damage. Luckily, the Restoration Team maintain their own internal database of library holdings and were able to identify possible candidates to use instead, in the shape of three film prints made in 1990. These were assessed by the film examiners and found to be in good condition. A/B roll cut negatives of all four episodes are retained by the library and so new prints could have been struck, but it would have meant a large extra expense and a long delay.
The prints were transferred to Digital Betacam and D3 digital videotape in the Spirit suite at Television Centre. The colourist was Jonathan Wood, an expert in feature film transfers for the BBC and the colourist with the most experience of running print film on the Spirit. Jon spent an entire day grading and deblobbing all four episodes and the results were stunning - exceptionally clean, stable and vibrant images. The story has never looked better. Viewers may have noticed a very small amount of 'white crushing', where highlight detail disappears - this was entirely down to overexposure of the film prints and could not be helped. The fact that the crushing is identical on the episode one film, made in 1970, and the subsequent films, made in 1990, suggests that the original negatives may in fact be overexposed slightly.
Problems were encountered with the master sound for episode one, which as with all the episodes, was taken from the archive master 16mm magnetic track. The audio from the film was very poor, suffering from cyclical loss of high frequency and level - rather unfortunate as this was apparently the archive master. Further investigation showed that this was not actually an original magnetic film - it had been made quite recently as the master acetates were in poor condition and had to be junked. Unfortunately, this particular episode appears to have slipped through the quality control net. However, there were at least three other earlier safety copies in the library and so a decent version was laid back onto the transmission videotape before its technical review.
Since the original work to restore episodes of 'Doctor Who and the Silurians', 'The Ambassadors of Death', 'Terror of the Autons' and 'The Daemons' in 1992/3 (see the article on 'Colour Restoration') , technology has progressed somewhat, especially in the fields of colour correction and dirt, sparkle and noise reduction.
The team decided to go back to the original ungraded Betacam SP recordings of the film recordings combined with the colour signal and completely redo the colour correction and clean-up the dirt and noise problems on the film recordings. Dave Hawley revisited the grading work he had originally carried out seven years ago, but with greatly improved technology at his disposal. The DVNR-1000 was used to digitally remove film dirt and sparkle and to remove electronic noise and fluctuations from the colour signal. Colour correction was carried out on the Pogle Platinum, which includes a facility to rotate the hue vectors to remove problems inherent in the colour signal due to standards conversion from PAL to NTSC and back again. This meant that the colour shifts could be almost neutralised before primary colour correction, resulting in a much better result than could be achieved in 1992.
Ralph Montagu supervised this new work and explains the process in more detail...
"Doctor Who and the Silurians was
originally restored to colour in 1993 for video release. Of the three stories completed
(the others being The Daemons and Terror of the Autons) the Silurians was the most
problematic as there were several short sections on the colour tapes which were affected
by picture loss, and some longer sections suffering from colour patterning. This colour
interference is a milder version of the patterning which has made the colour tapes for
Ambassadors of Death largely unusable.
The decision to re-show some of the early Pertwee stories on BBC-2 provided an opportunity to go back to the pre-graded restoration master tapes and look at how we might improve them. Six years ago our main concern was combining the luminance and chroma; colour grading and other forms of fine tuning took second place. This time around, grading, noise and dirt reduction became the main issue, using technology which was considerably less advanced in 1993.
The work was mainly carried out in Post Production's Grading Suite 1, which is equipped with Pogle colour correction and DVNR-1000. Dave Hawley, who graded much of our work in 1993, was the 'Colourist' (as they are now called) at the controls. The previous master was made on D3, this time we made two recordings, a D3 for transmission, and a Digital Betacam for the archive.
Pogle is able to employ secondary correction, which can give greater scope to adjust the colour. A general setting was arrived at for each episode, but then further adjustments were made where an individual scene or shot which required it. The extra versatility of the system meant that substantial improvements were made, although the relatively poor quality of the original chroma source limited the scope of our work. Compared with the truly superb results achieved on Spearhead from Space, the Silurians looked positively washed out even after our best endeavours!
The original film recordings suffer from a fair degree of both noise and dirt, but DVNR was applied to very good effect and improved the pictures in ways that simply wasn't possible six years ago. However the system couldn't deal with film dirt where this appeared in the same place on more than one frame. Unfortunately this was the case with most of the film inserts, where the film recordings were out-of-phase.
Each frame of a film recording (FR) should correspond to the same frame on the videotape master, which will include film sequences played into the studio. However, each frame on videotape is made up of two interlaced fields, and the absence of synchronisation in the early 1970's meant that in 50% of cases the FR machine took the wrong field as its first. As a result, field one from the tape would be treated as the second field on the FR, which would be combined with field one from the next frame on the tape.
Lack of synchronisation also affected film sequences played into the studio, but this didn't matter provided that the FR followed the same phase. Whilst the law of averages indicates that this should only have happened 50% of the time, sods law dictated that in this case it was considerably more! The result was double imaging on most film inserts (especially apparent where there is movement) and a repeating of dirt spots where these occurred.
As a result, the more noticeable spots had to be taken out individually using the mixer to replace the offending mark with a small area from the previous frame. On average, there were around 25 of these to be removed from each episode. However the last sequence in episode 5 and the first in episode 6 was so blighted by black and white specs that I treated it separately on HAL (a Paintbox which is also able to read and write to a Digital Betacam machine). This allowed me to paint out the offending dirt using the pen and tablet system usually employed for graphics. Coincidentally, this was the scene colourised in the US due to excessive colour patterning.
As mentioned in the article written in 1993, most of the Silurians film sequences suffered from white 'edit-flashes' at each picture cut. These were caused by poor neg cutting when the film inserts were originally put together. In 1993 we left the white flashes as they were part of how the story was finished on tape, but on reflection I decided they were visually irritating and boldly decided to remove them. This did represent a change to the original, something which restoration purists would frown upon, but I took the view that we were only removing what shouldn't have been there in the first place!
Having tackled the first offending scene with Liz Shaw and the Silurian in the barn, we couldn't ignore the other sequences, and it turned out to be the most time consuming part of the whole exercise. The task was made considerably more difficult because of the phase problems, with most white flashes appearing on two if not three frames of the film recording. The flashes were usually removed by taking material from the previous or next good frame, but in some cases movement in the frames prevented this and other forms of repair had to be made. The result was not only the removal of irritating white glitches, but also the creation of clean single frame cuts, rather than the two frame mixes caused by the out-of-phase film recording.
In total, the work undertaken took at least as long as the original restoration. Whilst the results are still not as good as some of the other colour restorations, I feel we achieved as much as is possible, and probably ever will be possible unless we can find some better colour originals. Although it isn't the greatest story, the repeat did give us the opportunity to improve on our original work, and for this to be seen by viewers of terrestrial television!
At the time, we believed that the other stories restored to colour would be scheduled for transmission in 2000, so we next tackled Terror of the Autons. Technically this restoration had always been the best, but Pogle and DVNR-1000 enabled us to make further improvements, especially where areas of bright red triggered high levels of noise on the original.
We were also able to compare the original two-inch recording of Jo extinguishing the Doctors 'steady state microwelding' experiment with the restored version. This short clip was discovered after the original restoration was completed and we thought that the original might give us a clue as to what the correct grading should be. To our surprise, the original two-inch source looked strangely flat compared with the FR, and there was no significant difference in the colour. This is probably explained by the fact that the process of recording onto film tends to increase contrast and add grain. Both these characteristics can create a more arresting image, albeit lacking the definition and steadiness of the videotape original!
Work on Terror of the Autons was only completed one day before we heard that Tom Baker stories would follow after Doctor Who and the Silurians. Work on The Daemons will have to wait until another time!"
Many of Jon Pertwee's episodes only exist as NTSC standards conversions recovered from overseas TV stations over the last few years. The quality of the PAL to NTSC conversions is fairly poor, being made in the mid-seventies on equipment that is considered extremely crude by today's standards. However, the crude nature of the conversions may actually turn out to be a blessing...
Around five years ago when we first looked at this problem, we handed it to the BBC Research Department at Kingswood Warren in Surrey to see what they could make of it. Image specialist Jim Easterbrook was able to work out exactly how a sequence of PAL video fields had been converted to a new sequence of NTSC video fields - and was able to demonstrate via a software simulation how it could be possible to reconvert material back to PAL almost seemlessly. This would eliminate the blurring and judder effects whenever anything moved on the screen. We dubbed this technique 'Reverse Standards Conversion', but unfortunately the funding for the project dried up and it was never implemented. Now that we have money available, we are hoping that Jim's research can be resurrected in time to be applied to at least some of the affected episodes. Failing that, they will be reconverted using the best possible modern standards converters.
Two stories, 'The Ambassadors of Death' and 'The Mind of Evil' have very little colour material available and two other stories, 'Planet of the Daleks' and 'Invasion of the Dinosaurs' are each missing one colour episode. The technology is available to recolourise these episodes from scratch using a specialist computer system, but the cost (approximately twenty thousand pounds per episode) is prohibitive. At this point it is unlikely that the first two stories will be considered for transmission in either monochrome or mixed format - sadly the ratings plummet, as modern viewers appear to have little tolerance for monochrome programmes. It is hoped that we will be able to afford to colourise the two remaining episodes however, releasing two complete six-part stories for transmission.
Tom Baker Stories
Prior to the cancellation of the Baker repeats after 'Genesis of the Daleks', the team had already completed remastering and restoration work on half of this story, plus the two following stories, 'Revenge of the Cybermen' and 'Terror of the Zygons'.
Originally, we didn't intend to do any work on the Baker stories, but after seeing lots of problems with episode one of 'Genesis of the Daleks' during transmission, we successfully argued that work needed to be done to improve the technical quality to make it more suitable for the expectations of a modern viewing audience. The sort of problems that needed rectifying were dirt and sparkle on film sequences, overall video noise throughout the episode, and tape problems such as dropouts, offlocks etc.
Starting with episode four of 'Genesis of the Daleks', most of this work was carried out using the DVNR-1000 and vision mixer to fix any problems. Additionally, a private collector held film inserts for half of episode two and all of episodes three and four of 'Terror of the Zygons', so these were transferred on the Spirit telecine and edited back into the programme. The new transfers were much more stable and had higher resolution than the original telecined inserts, which also suffered from incorrect interlace and field flicker problems.
Copyright Steve Roberts, Ralph Montagu, March 2000