Carol Anne’s in trouble on the front cover of Cinefex #26, assailed by a whole new host of ghostly apparitions in Brian Gibson’s 1986 sequel Poltergeist II: The Other Side. The apparitions come courtesy of Richard Edlund’s Boss Film Studios and are thus of superior quality (though most folk wouldn’t say the same about the movie). The back cover features a close-up of the stained glass knight from Barry Levinson’s Young Sherlock Holmes, a family film released the previous year. Two films and two articles, then, within this issue’s total span of 68 pages.
- Poltergeist II – To Hell and Back (article by Nora Lee and Janine Pourroy)
- Young Sherlock Holmes – Anything But Elementary (article by Jody Duncan Shay)
Your present-day VFX fan may well vote for switching the two cover images on this issue, relegating Carol Anne to the back cover and showcasing the stained glass knight on the front. Why? Because that brief sequence from Young Sherlock Holmes (just five shots by my count) is now regarded as a significant milestone in the development of digital visual effects. Even at the time, it seems the Cinefex team sensed something big was happening, since a whopping 20% of this issue is devoted to its creation by ILM’s Pixar Computer Animation Group, under the direction of Dennis Muren and John Lasseter.
Before we get to digital warriors, however, we have some old-school optical ghosts to consider. Lee and Pourroy’s article on Poltergeist II – pleasantly chatty in tone – digs deep into the film’s pre-production, which was so extensive that VFX supervisor Richard Edlund declared it ‘the most storyboarded film in history.’ John Bruno – one of the key storyboard artists – not only analyses this process but also talks about his encounter with Swiss artist H.R.Giger (of Alien fame), who conceived the film’s nightmarish creatures.
Bruno also walks us through the film’s complex cel-animated ectoplasm shots. ‘In the first movie the ectoplasm just moved left to right,’ he says. ‘In Poltergeist II … it is multilayered and dimensional.’ The illusion of three dimensions was achieved by using many layers of animation. For maximum chaos, different layers were drawn by different people: ‘The approach we took,’ says Bruno, ‘… was to scramble the animators so they would not get a rhythm to a scene.’
Chaos reigns, in fact, as the article goes on to talk about dental braces coming to life (wires being pulled through ‘one of the most amazing closeup [gelatin] heads I’ve ever seen’), a tequila-worm-vomit-monster (made from inflatable rubber and inserted into actor Craig T. Nelson’s mouth) and a cloud of glowing butterflies (paper miniatures with flapping mechanisms shot individually and optically comped on an animation stand). The film’s outlandish imagery is summed up by Edlund when he says, ‘None of this stuff makes the least bit of sense, but it works.’
And work it did, well enough to earn Edlund’s team a 1986 Oscar nomination (the top prize was taken by Aliens). In keeping with their broad remit, Lee and Pourroy’s article covers plenty of ground, dissecting the ‘Horrorbaby’ costume worn by amputee performer Noble Craig, the huge ‘great beast’ puppet and the cloud tank imagery employed to visualise the astral dimension known as the ‘other side’.
For the technically-minded, there are descriptions of Boss’s two latest innovations. No doubt they’re both now museum pieces; at the time they were state-of-the-art. The new animation stand is able to move both camera and artwork through all three dimensions under full motion control, while optical supervisor Mark Vargo asserts that the Zap optical printer is ‘unlike any other aerial printer I’m aware of.’ Lee and Pourroy’s detailed breakdown of both pieces of equipment – how they were built, under what tolerances they operated and so on – is another valuable entry in the ever-growing Cinefex databank of visual effects hardware development.
Whatever the dramatic shortcomings of Poltergeist II (let’s face it, there are plenty), its visual effects still represent photo-chemical and optical techniques running at full steam. And, however old-school the creative processes might seem today, the end results frequently worked like a dream. Comparing the model crew’s replica suburban neighbourhood with the real thing, miniature supervisor Mark Stetson concludes, ‘the effects shot looked better than the live-action shot.’
What more could a visual effects artist want?
Young Sherlock Holmes
Moving on to Young Sherlock Holmes, we’re in the safe hands of Jody Duncan Shay as we study ‘a variety of cinematic techniques … rarely explored within the confines of a single film.’ Those techniques – and the images they put up on the screen – earned Dennis Muren and his ILM team an Oscar nomination at the 1985 awards. (Muren’s ILM colleague Ken Ralston walked away with the trophy that year – I wonder how much friendly rivalry there was on that night.)
The early part of the article is devoted not to visual effects but to special effects. In case you didn’t know, the latter includes the sort of practical gags and illusions that are created live on the set, while the former usually involves miniatures, opticals, matte paintings and so on (and of course, to bring things up to date, all things digital). In the case of Young Sherlock Holmes, the on-stage effects included a full-scale flying machine suspended on wires in the grounds of the UK’s Belvoir Castle, an animatronic pheasant and a murderous hat-rack.
Duncan goes on to describe how ILM’s go-motion system was adapted to shoot a pair of small harpies attacking actor Nigel Stock. The go-motion process (a development of stop-motion in which computer-driven motors control both the gross movement of the puppet and the operation of the camera) imbued the wings of the demonic creatures with a natural motion blur, while animator Harry Walton added finesse to their movements and the rotoscope department worked overtime, creating hand-drawn masks to remove the complex go-motion rig.
When it came to the film’s ‘food comes alive’ hallucination sequence, Muren chose yet another technique, abandoning animation in favour of rod puppets ‘because we would be able to do a lot of takes.’ He adds, ‘It’s … something I learned from watching Steven Spielberg direct E.T. – you just get as many takes as possible on the mechanical stuff so that you’ll have a lot to choose from.’
Spielberg (the film’s producer) clearly approved of Muren’s approach, and reportedly loved art director Dave Carson’s concept for ‘a potato leader and cream puffs jumping off walls and little jellybeans coming out of Watson’s pockets.’ Director Barry Levinson was less enthusiastic about the silliness. ‘There was an ongoing struggle,’ says Carson, ‘… [with] Barry wanting to cut the sequence way, way down and Steven wanting to pump it way, way up. Eventually, Steven stepped back.’
Matte painting fans will be pleased to know there’s a section on their favourite subject, interesting not only for its breakdown of the matte shots but also for the remarks made by Chris Evans about the challenges he faced in ‘resolving technical problems that wouldn’t have been there if we’d been given the opportunity to offer our expertise at the beginning of the production rather than later, after the plates had already been shot.’ (Compare his frustration here to the enthusiasm he expresses in issue #25’s article on Enemy Mine, a film for which Wolfgang Petersen included the matte team in the creative process from day one.)
All this is great stuff, but it’s really just a series of sideshows leading up to the main event. So, now we’ve knocked down some coconuts and won ourselves a big cuddly bear, it’s time to enter the three-ring circus that stands at the very heart of this particular fairground. The sign over the circus tent reads Pixar.
‘My main worry,’ says ILM visual effects supervisor Dennis Muren, talking about his decision to realise Young Sherlock Holmes‘s stained glass knight through computer imagery, ‘was that I had never seen computer imagery that looked real.’ In order to give Pixar a fighting chance of getting past that sticking point, he gave them a lot of time and space to develop the necessary tools to achieve his vision. Even so, he left himself an escape route: ‘We went into the project with enough end time … [to] get the effect through a more traditional approach.’
Muren’s comprehensive account takes us through a difficult conceptual phase (how to make a character that was essentially two-dimensional look threatening) and on to the various advances Pixar made to make their procedures more interactive. These included improved preview programs ‘so that we could see results in five minutes rather than in two hours,’ and a new laser scanner that meant they no longer had to read images off a cathode ray tube.
Technical supervisor Bill Reeves lists his four steps of computer imaging – modelling, animation, rendering, output – while artistic supervisor John Lasseter introduces us to the Polhemus three-space digitiser he used to gather coordinates from a clay model of the knight and convert them into vector data. The hardware and software being used here is bespoke – or close to it – from Eben Ostby’s program designed to smooth the noisy data from the digitiser to the Evans and Sutherland system used for the animation itself. There’s not a Mac or PC in sight, and none of the software carries a logo from either Adobe or Autodesk. This is computer imagery pulling itself up by its bootstraps and inventing what it needs as it goes along.
Some of the procedures sound impossibly clumsy: ‘We used [a] Mitchell camera to shoot RAR “pencil tests” off the screen,’ says Lasseter. Others are so forward-thinking that they still form the basis of many animation systems used today: ‘The articulation controls [of the knight’s skeleton] were set up something like a tree … the technical term for it is “hierarchical”.’
Muren’s mission to create something ‘real’ runs through Duncan’s account from beginning to end. At all stages, he’s pushing and pushing for maximum believability. For example, the sequence’s ambitious hero shot, where the camera tracks all the way round the knight, involved an early form of match-moving that relied heavily on tape measures and blueprints of the set. Lasseter’s response to the challenges of this difficult shot is a phrase that may well be tattooed on the hearts of every visual effects artist out there: ‘We really wanted to pull it off because no one had ever done that before.’
If you asked an informed cinemagoer to name a film that advanced early digital effects, it’s my guess they’d say Tron, Terminator 2 or Jurassic Park. All good answers, and while we’re at it let’s not forget Star Trek: The Wrath of Khan, The Last Starfighter and The Abyss. The stained glass knight from Young Sherlock Holmes, I suspect, is largely forgotten except by those in the business and fans of the art. Perhaps that’s not surprising, given that it makes only a brief appearance in a film that was never a big hit (less than $20 million domestic gross according to boxofficemojo).
Duncan’s account of this tipping point in visual effects is thirteen pages of solid gold, made all the more shiny by the contemporary comments of the various artists involved. Optical supervisor John Ellis, standing firmly on the traditional side of the ILM fence, has this to say about his Pixar colleagues: ‘They’re typing things in and talking about things like they’re from another planet … but then you see it the next day and you think, “Wow …”‘ And Pixar’s Bill Reeves has this to say about Muren: ‘We learned by him egging us on … stretching us so that we had to develop some of these techniques.’
At this point, any commentator worth his salt would mention briefly what Pixar went on to achieve after separating from its ILM roots.
But I don’t really need to do that, do I?
Flipping back through this issue to review the pictures, I see that the Poltergeist II section is replete with everything you’d expect from a Cinefex article: behind-the-scenes shots of monster-makers and miniatures set alongside colour stills and, in this case, some extraordinary Giger concept art. My favourite image is on page 26 and shows actress JoBeth Williams in nonchalant mood while standing knee-deep in a quagmire full of zombies.
While there are some great photos in the Young Sherlock Holmes article showing go-motion puppets in front of bluescreens and, yes, an entire cast of dancing cream puffs, pride of place goes to the shots of a young John Lasseter and his team wielding digitising tablets while staring at enormous 14″ CRT screens. As a vision of the future, it’s now looking pleasantly quaint.
Did you enjoy this Cinefex retrospective? If so, click here to read the others in the series.
2 thoughts on “Revisiting Cinefex (26): Poltergeist II, Young Sherlock Holmes”
Thanks, Graham! Another excellent essay — though, for the record, you’re a few zeros short on the boxoffice take for “Young Sherlock Holmes.”
Oops! Thanks, Don. Counting was never my strong point. Corrected now 🙂