How were visual effects achieved before the advent of computer generated imagery (CGI)? Most of us know that spaceships used to be miniatures, and monsters used to be puppets or people in suits, but what about the less tangible effects? How did you create something as exotic as an energy beam or a dimensional portal without the benefit of digital particle simulations? The answer was often a combination of chemistry, physics, artistry and ingenuity. Here are five examples.
1. “Star Trek” transporters
The original series of Star Trek, premiered in 1966, had to get creative to achieve its futuristic effects with the budget and technology available. The Howard Anderson Company was tasked with realising the iconic transporter effect which enables Kirk’s intrepid crew to beam down to alien planets. Darrell Anderson created the characteristic sparkles of the dematerialisation by filming backlit aluminium powder being sprinkled in front of a black background in slow motion. Hand-drawn mattes were then used to ensure that the sparkling powder only appeared over the characters.
2. “Ghostbusters” proton packs
The much-loved 1984 comedy Ghostbusters features all kinds of traditional effects, including the never-to-be-crossed particle streams with which the heroes battle their spectral foes. The streams consist of five layers of traditional cell animation – the same technique used to create, say, a Disney classic like Sleeping Beauty – which were composited and enhanced on an optical printer. (An optical printer is essentially two or more film projectors connected to a camera so that multiple separate elements can be combined into a single shot.) Composited onto the tips of the Ghostbusters’ guns were small explosions and other pyrotechnic effects shot on a darkened stage.
3. “Lifeforce” energy beams
This cult 1985 sci-fi horror film, most notable for an early screen appearance by Patrick Stewart, features alien vampires which drain the titular lifeforce from their victims. To visualise this lifeforce, VFX supervisor John Dykstra settled on a process whereby a blue argon laser was aimed at a rotating tube made of highly reflective mylar. This threw flowing lines of light onto a screen where it would be captured by the camera for later compositing with the live-action plates. The tube could be deliberately distorted or dented to vary the effects, and to add more energy to certain shots multiple brief elements of a flashing xenon bulb were added to the mix.
4. “Big Trouble in Little China” portal
A mixture of chemical and optical effects were employed for certain shots in the 1986 action-comedy Big Trouble in Little China. Director John Carpenter wanted an effervescent effect like “an Alka-Seltzer tablet in water” to herald the appearance of a trio of warriors known as the Three Storms. After many tests, the VFX team determined that a combination of green paint, metallic powder and acetone, heated in a Pyrex jar on a hotplate, produced an interesting and suitable effect. The concoction was filmed with a fisheye lens, then that footage was projected onto a dome to make it look like a ball of energy, and re-photographed through layers of distorted glass to give it a rippling quality.
5. “Independence Day” cloud tank
By 1996, CGI was replacing many traditional effects, but the summer blockbuster Independence Day used a healthy mix of both. To generate the ominous clouds in which the invading spacecraft first appear, the crew built what they called the “Phenomenon Rig”. This was a semi-circle of halogen lights and metal piping which was photographed in a water tank. Paint was injected into the water through the pipes, giving the appearance of boiling clouds when lit up by the lamps within. This was digitally composited with a live-action background plate and a model shot of the emerging ship.
Cathode ray tube televisions, those bulky, curve-screened devices we all used to have before the rise of LCD flat-screens, already seem like a distant memory. But did you know that they were not the first form of television, that John Logie Baird and his contemporaries first invented a mechanical TV system more akin to Victorian optical toys than the electronic screens that held sway for the greater part of the 20th century?
Mechanical television took several forms, but the most common type revolved, quite literally, around a German invention of 1884 called the Nipkow disc. This had a number of small holes around it, evenly spaced in a spiral pattern. In the Baird standard, developed by the Scottish inventor in the late 1920s, there were 30 holes corresponding to 30 lines of resolution in the resulting image, and the disc would revolve 12.5 times per second, which was the frame rate.
In a darkened studio, an arc light would be shone through the top portion of a spinning Nipkow disc onto the subject. The disc would create a flying spot – a spot of light that travelled horizontally across the scene (as one of the holes passed in front of the arc lamp) and then travelled horizontally across it again but now slightly lower down (as the next hole in the spiral pattern passed the lamp) and so on. For each revolution of the 30-hole disc, 30 horizontal lines of light would be scanned across the subject, one below the other.
A number of photocells would be positioned around the subject, continually converting the overall brightness of the light to a voltage. As the flying spot passed over light-coloured surfaces, more light would reflect off them and into the photocells, so a greater voltage would be produced. As the spot passed over darker objects, less light would reflect into the photocells and a smaller voltage would result. The voltage of the photocells, after amplification, would modulate a radio signal for transmission.
A viewer’s mechanical television set would consist of a radio receiver, a neon lamp and an upright Nipkow disc of a foot or two in diameter. The lamp – positioned behind the spinning disc – would fluctuate in brightness according to the radio signal.
The viewer would look through a rectangular mask fitted over the top portion of the disc. Each hole that passed in front of the neon lamp would draw a streak of horizontal (albeit slightly arcing) light across the frame, a streak varying in brightness along its length according to the continually varying brightness of the lamp. The next hole would draw a similar line just beneath it, and so on. Thanks to persistence of vision, all the lines would appear at once to the viewer, and it would be followed by 11.5 more sets of lines each second: a moving image.
A number of people were experimenting with this crude but magical technology at the same time, with Baird, the American Charles Francis Jenkins and the Japanese Kenjiro Takayanagi all giving historic public demonstrations in 1925.
The image quality was not great. For comparison, standard definition electronic TV has 576 lines and 25 frames per second in the UK, twice the temporal resolution and almost 20 times the spatial resolution of the Baird mechanical standard. The image was very dim, it was only an inch or two across, and it could only be viewed by a single person through a hood or shade extending from the rectangular mask.
The BBC began transmitting a regular mechanical TV service in 1929, by which time several stations were up and running in the USA. An early viewer, Ohio-based Murry Mercier Jr., who like many radio enthusiasts built his own mechanical TV from a kit, described one of the programmes he watched as “about 15 minutes long, consisting of block letters, from the upper left to the lower right of the screen. This was followed by a man’s head turning from left to right.” Hardly Breaking Bad.
Higher resolutions and larger images required larger Nipkow discs. A brighter image necessitated lenses in each of the disc’s holes to magnify the light. Baird once experimented with a disc of a staggering 8ft in diameter, fitted with lenses the size of bowling balls. One of the lenses came loose, unbalancing the whole disc and sending pieces flying across the workshop at lethal speeds.
Other methods of reproducing the image were developed, including the mirror screw, consisting of a stack of thin mirrors arranged like a spiral staircase, one “step” for each line of the image. The mirror screw produced much larger, brighter images than the Nipkow disc, but the writing was already on the wall for mechanical television.
By 1935, cathode ray tubes – still scanning their images line by line, but by magnetically deflecting an electron beam rather than with moving parts – had surpassed their mechanical counterparts in picture quality. The BBC shut down its mechanical service, pioneers like Baird focused their efforts on electronic imaging, and mechanical TV quietly disappeared.
“These are small,” Father Ted once tried to explain to Father Dougal, holding up toy cows, “but the ones out there are far away.” We may laugh at the gormless sitcom priest, but the chances are that we’ve all confounded size and distance, on screen at least.
The ship marooned in the desert in Close Encounters of the Third Kind, the cliff at the end of Tremors, the runways and planes visible through the windows of Die Hard 2’s control tower, the helicopter on the boat in The Wolf of Wall Street, even the beached whale in Mega Shark Versus Giant Octopus – all are small, not far away.
The most familiar forced perspective effect is the holiday snap of a friend or family member picking up the Eiffel Tower between thumb and forefinger, or trying to right the Leaning Tower of Pisa. By composing the image so that a close subject (the person) appears to be in physical contact with a distant subject (the landmark), the latter appears to be as close as the former, and therefore much smaller than it really is.
Architects have been playing tricks with perspective for centuries. Italy’s Palazzo Spada, for example, uses diminishing columns and a ramped floor to make a 26ft corridor look 100ft long. Many film sets – such as the basement of clones in Moon – have used the exact same technique to squeeze extra depth out of limited studio space or construction resources.
Even a set that is entirely miniature can benefit from forced perspective, with a larger scale being used in the foreground and a smaller one in the background, increasing the perceived depth. For example, The Terminator’s “Future War” scenes employ skulls of varying size, with background ruins on an even smaller scale.
An early cinematic display of forced perspective was the 1908 short Princess Nicotine, in which a fairy who appears to be cavorting on a man’s tabletop is actually a reflection in a distant mirror. “The little fairy moves so realistically that she cannot be explained away by assuming that she is a doll,” remarked a Scientific American article of the time, “and yet it is impossible to understand how she can be a living being, because of her small stature.”
During the 1950s, B movies featuring fantastically shrunk or enlarged characters made full use of forced perspective, as did the Disney musical Darby O’Gill and the Little People. VFX supervisor Peter Ellenshaw, interviewed for a 1994 episode of Movie Magic, remembered the challenges of creating sufficient depth of field to sell the illusion: “You had to focus both on the background and the foreground [simultaneously]. It was very difficult. We had to use so much light on set that eventually we blew the circuit-breakers in the Burbank power station.”
Randall William Cook was inspired years later by Ellenshaw’s work when he was called upon to realise quarter-scale demonic minions for the 1987 horror movie The Gate. Faced with a tiny budget, Cook devised in-camera solutions with human characters on raised foreground platforms, and costumed minions on giant set-pieces further back, all carefully designed so that the join was undetectable. As the contemporary coverage in Cinefex magazine noted, “One of the advantages of a well-executed forced perspective shot is that the final product requires no optical work and can therefore be viewed along with the next day’s rushes.”
A subgroup of forced perspective effects is the hanging miniature – a small-scale model suspended in front of camera, typically as a set extension. The 1925 version of Ben Hur used this technique for wide shots of the iconic chariot race. The arena of the Circus Maximus was full size, but in front of and above it was hung a miniature spectators’ gallery containing 10,000 tiny puppets which could stand and wave as required.
Doctor Who used foreground miniatures throughout its classic run, often more successfully than it used the yellow-fringed chromakey of the time. Earthly miniatures like radar dishes, missile launchers and big tops were captured on location, in camera, with real skies and landscapes behind them. The heroes convincingly disembark from an alien spaceship in the Tom Baker classic “Terror of the Zygons” by means of a foreground miniature and the actors jumping off the back of a van in the distance. A third-scale Tardis was employed in a similar way when the production wanted to save shipping costs on a 1984 location shoot on Lanzarote.
Even 60 years on from Ben Hur, Aliens employed the same technique to show the xenomorph-encrusted roof in the power plant nest scene. The shot – which fooled studio executives so utterly that they complained about extravagant spending on huge sets – required small lights to be moved across the miniature in sync with the actors’ head-torches.
The Aliens shot also featured a tilt-down, something only possible with forced perspective if the camera pivots around its nodal point – the point within the lens where the light focuses. Any other type of camera movement gives the game away due to parallax, the optical phenomenon which makes closer objects move through a field of view more quickly than distant ones.
The 1993 remake of Attack of the 50ft Woman made use of a nodal pan to follow Daniel Baldwin to the edge of an outdoor swimming pool which a giant Daryl Hannah is using as a bath. A 1/8th-scale pool with Hannah in was mounted on a raised platform to perfectly align on camera with the real poolside beyond, where Baldwin stood.
The immediacy of forced perspective, allowing actors of different scales to riff off each other in real time, made it the perfect choice for the seasonal comedy Elf. The technique is not without its disadvantages, however. “The first day of trying, the production lost a whole day setting up one shot and never captured it,” recalls VFX supervisor Joe Bauer in the recent documentary Holiday Movies That Made Us.
Elf’s studio, New Line, was reportedly concerned that the forced perspective shots would never work, but given what a certain Peter Jackson was doing for that same studio at the same time, they probably shouldn’t have worried.
The Lord of the Rings employed a variety of techniques to sell the hobbits and dwarves as smaller than their human friends, but it was in the field of forced perspective that the trilogy was truly groundbreaking. One example was an extended cart built to accommodate Ian McKellen’s Gandalf and Elijah Wood’s supposedly-diminutive Frodo. “You could get Gandalf and Frodo sitting side by side apparently, although in fact Elijah Wood was sitting much further back from the camera than Gandalf,” explains producer Barrie Osborne in the trilogy’s extensive DVD extras.
Jackson insisted on the freedom to move his camera, so his team developed a computer-controlled system that would correct the tell-tale parallax. “You have the camera on a motion-controlled dolly, making it move in and out or side to side,” reveals VFX DP Brian Van’t Hul, “but you have another, smaller dolly [with one of the actors on] that’s electronically hooked to it and does the exact same motion but sort of in a counter movement.”
Forced perspective is still alive and kicking today. For Star Wars Episode IX: The Rise of Skywalker, production designer Kevin Jenkins built a 5ft sand-crawler for shooting in the Jordan Desert. “It was placed on a dressed table at height,” he explained on Twitter, “and the Jawa extras were shot at the same time a calculated distance back from the mini. A very fine powdery sand was dressed around for scale. We even made a roller to make mini track prints! Love miniatures :)”
Next month, Terminator 2: Judgment Day turns 30. Made by a director and star at the peaks of their powers, T2 was the most expensive film ever at the time, and remains both the highest-grossing movie of Arnold Schwarzenegger’s career and the sequel which furthest out-performed its progenitor. It is also one of a handful of films that changed the world of visual effects forever, signalling as it did – to borrow the subtitle from its woeful follow-up – the rise of the machines.
The original Terminator, a low-budget surprise hit in 1984, launched director James Cameron’s career and cemented Schwarzenegger’s stardom, but it wasn’t until 1990 that the sequel was green-lit, mainly due to rights issues. At the Cannes Film Festival that year, Cameron handed executive producer Mario Kassar his script.
Today it’s easy to forget how risky it was to turn the Terminator, an iconic villain, an unstoppable, merciless death machine from an apocalyptic future, into a good guy who doesn’t kill anyone, stands on one leg when ordered, and looks like a horse when he attempts to smile. But Kassar didn’t balk, granting Cameron a budget ten times what he had had for the original, while stipulating that the film had to be in cinemas just 14 months later.
Even with some expensive sequences cut – including John Connor sending Kyle Reese back through time in the heart of Skynet HQ, a scene that would ultimately materialise in Terminator Genisys – the script was lengthy and extremely ambitious. Beginning on October 8th, 1990, the shooting schedule was front-loaded with effects shots to give the maximum time for CGI pioneers Industrial Light and Magic to realise the liquid metal T-1000 (Robert Patrick).
To further ease ILM’s burden, every trick in the book was employed to get T-1000 shots in camera wherever possible: quick shots of the villain’s fight with the T-800 (Schwarzenegger) in the steel mill finale were done with a stuntman in a foil suit; a chrome bust of Patrick was hand-raised into frame for a helicopter pilot’s reaction shot; the reforming of the shattered T-1000 was achieved by blowing mercury around with a hair dryer; bullet hits on the character’s torso were represented by spring-loaded silver “flowers” that burst out of a pre-scored shirt on cue.
Stan Winston Studio also constructed a number of cable-controlled puppets to show more extensive damage to the morphing menace. These included “Splash Head”, a bust of Patrick with the head split in two by a shotgun blast, and “Pretzel Man”, the nightmarish result of a grenade hit moments before the T-1000 falls to its doom in the molten steel.
Traditional models and rear projection are used throughout the film. A few instances are all too obvious to a modern audience, but most still look great and some are virtually undetectable. Did you know that the roll-over and crash of the cryo-tanker were shot with miniatures? Or that the T-800 plucking John off his bike in the drainage channel was filmed against a rear projection screen?
Plenty of the action was accomplished without such trickery. The production added a third storey to a disused office building near Silicon Valley, then blew it up with 100 gallons of petrol, to show the demise of Cyberdyne Systems. DP Adam Greenberg lit 5.5 miles of freeway for the car chase, and pilot Chuck Tamburro really did fly the T-1000’s police helicopter under a 20ft underpass.
Chaotic, confusing action scenes are the norm today, but it is notable that T2’s action is thrilling yet never unclear. The film sends somewhat mixed messages though, with its horrific images of nuclear annihilation and the T-800’s morality lessons from John juxtaposed with indulgent violence and a reverence for firearms. “I think of T2 as a violent movie about world peace,” Cameron paradoxically stated. “It’s an action movie about the value of human life.”
Meanwhile, 25 person-years of human life were being devoted by ILM to the T-1000’s metallic morphing abilities. Assistant VFX supervisor Mark Dippé noted: “We were pushing the limits of everything – the amount of disc space we had, the amount of memory we had in the computers, the amount of CPUs we had. Each shot, even though it only lasted about five seconds on the screen, typically would take about eight weeks to complete.”
The team began by painting a 2×2” grid on a near-naked Patrick and shooting reference footage of him walking, before laser-scanning his head at the appropriately-named Cyberware Laboratory. Four separate computer models of the T-1000 were built on Silicon Graphics Iris 4Ds, from an amorphous blob to a fully-detailed chrome replica of Patrick, each with corresponding points in 3D space so that the custom software Model Interp could morph between them.
Other custom applications included Body Sock, a solution to gaps that initially appeared when the models flexed their joints, Polyalloy Shader, which gave the T-1000 its chrome appearance, and Make Sticky, with which images of Patrick were texture-mapped onto the distorting 3D model, as when he melts through a barred gate at the mental hospital.
The film’s legacy in visual effects – for which it won the 1992 Oscar – cannot be understated. A straight line can be drawn from the water tendril in Cameron’s The Abyss, through T2 to Jurassic Park and all the way on to Avatar, with which Cameron again broke the record for the highest-grossing film of all time. The Avatar sequels will undoubtedly push the technology even further, but for many Cameron fans his greatest achievement will always be Terminator 2: Judgment Day, with its perfect blend of huge stunts, traditional effects and groundbreaking CGI.