Following on the heels of Raiders and Temple of Doom, the Last Crusade presented the optical effects unit at Industrial Light and Magic and the physical effects team in London with the ultimate challenge, topping their own exemplary work in the two previous films. With such diverse assignments as the construction of an authentic World War I tank, the collapse of an enormous temple, and the aging of a man into dust without the aid of cutaways, Indiana Jones and the Last Crusade would prove to be the perfect vehicle for cinematic one-upsmanship.

In the film's opening sequence Indy must contend with not only the bandits chasing him, but also the train's inhospitable occupants. At one point, he fights with a bandit on top of a boxcar while an angry rhinoceros below pokes its prodigious horn up through the ceiling. Lantieri rigged the roof to break away and then had technicians down below thrusting a horn up through the roof. A cuing system was rehearsed so that at the crucial moment, the horn would burst up under River Phoenix's arm, along his side and between his legs. The massive foam and fiberglass rhinoceros prop had been created by the British special effects crew but three days before filming the scene, Spielberg decided he wanted the rhino to move! John Buechler took on the challenge. The huge prop was delivered to Buechler's North Hollywood studio from the stage at Universal. Buechler and his crew set to work re-inventing the rhino. Over the course of three days, Buechler and company resculpted, re-cast, and mechanized the beast to such a degree, that not only did the rhino blink, snarl, and snort as Spielberg wanted, but it also wiggled it's ears! The director was so happy with the final creature that he allowed Buechler to call "action" and "cut" on the final shot of the film.

Elsewhere in the sequence Lantieri relied upon two mechanical giraffes built in England by mechanical effects supervisor George Gibbs and his crew. "The giraffes were operated by cables and brake handles," Lantieri explained, "and we cut skylight holes in the top of one of the boxcars so the heads could stick up through. One of the first things we learned was that steam locomotives are very loud - so loud that when we were in the boxcar it was hard to hear directions. Ultimately we used radios to communicate with David Tomblin, the first assistant director. We could not stick our heads out of the car because it would ruin the shots, so after a while we started answering his questions with the giraffes themselves. David would ask if the giraffes were ready, and the giraffes would shake their heads 'yes' or 'no' in response. It got to be pretty funny."

Continuing his hide and seek chase onboard the steaming train Indy encounters a ferocious anaconda, in reality an animatronic one.

One of the most unusual effects involved the creation of 1,000 imitation rats, some mechanically articulated. Used for a catacomb sequence, some of the fake rats were used as swimming rats; while others were used for a conflagration effect, so that no live rat was harmed during filming.

For Indy and Elsa's arrival at the Austrian stonewalled castle, an actual castle was photographed in West Germany. When the real edifice was later deemed too small, however, ILM matte department supervisor Mark Sullivan was called upon to expand it with a painting. "We also needed to add a rainstorm over the castle," said Sullivan. "In the past, effects people have had trouble shooting and matting in real water, so we tried something different, we filmed granulated Borax hand soap against black, threw it into a fan to get the effect of sheets of rain and shot it high-speed. Then, since it was a daylight shot, we just barely double-exposed it in. If the Borax had been brighter in the shot, it might have looked like snow. Also, we used a fast speed so it would not appear to be drifting." Mike Lessa in the animation department added lightning bolts to further enhance the establishing shot.

Later in the film, with Henry's diary saved from a book burning rally, Indy and his father race to the Berlin airport to catch the next flight out of Germany. The task of creating a pre-war representation of the Berlin airport fell to the ILM matte department. The live-action basis for the painting was an existing airport facility located on Treasure Island between San Francisco and Oakland. The airport is now part of a military installation, but in the 1930s it was used as a terminal for seaplanes. Also appropriate was the fact that it had an art deco style of architecture. Matte artist Yusei Uesugi added a control tower, Nazi banners, vintage automobiles and a sign that read 'Berlin Flughafen' as a final touch.

Unable to get a standard flight, the Joneses opt for the only available alternative, a commercial zeppelin. When the zeppelin is ordered back to Germany the Joneses, in order to avoid capture, climb down into a small biplane attached to the underbelly of the ship and fly off. Partial full-scale sets and miniatures were used for the sequence. For shots of Indy and Henry lowering themselves into the biplane, George Gibbs and his crew constructed the entire belly section of a full-size zeppelin. The set piece was suspended forty feet in the air between two towers, and a full-size biplane was attached to a large scissor lift that could lower it and move it around in the frame. The full-size setup was used solely for close-up and medium shots. Wide angles of the full zeppelin employed a miniature constructed by ILM model shop supervisor Michael Fulmer and his crew. The airship model, eight feet long and carved out of foam, also featured a like-scaled miniature biplane small enough to fit in one's palm. For scenes of the plane separating from the zeppelin, a larger-scale biplane with a two-foot wingspan was built. The zeppelin and biplane miniatures were then shot separately under motion control and combined in optical.

Though airborne and free of the zeppelin, Indy and Henry soon discover they have not shaken their enemies and are in fact being chased by two Nazi fighter planes. Once again, full-size airplanes and miniatures were combined to realize the sequence. For live-action shots involving the enemy pursuers, Swiss army training planes were dressed to look like German World War II fighters. To simulate machine gun fire, both the fighter planes and the full-scale biplane were fitted with electronic strobe lights and a revolutionary new firing system designed by George Gibbs. "We built what we call 'gas guns'- a new idea we pioneered based on the internal combustion chamber of an engine. The guns run off liquid gas-propane gas and oxygen-that we ignite with a spark plug similar to what is a Honda motorbike. By doing this, we are able to avoid using guns that shoot blanks-which is an advantage, not only because blanks are expensive, but also because you can hurt someone who is standing too close when the blanks go off. With these gas guns, nothing comes out of them except flames. Unfortunately, the technology works only with larger weapons like machine guns. The equipment required is too bulky for handguns."

After father and son crash-land into a countryside shack, they appropriate a car as their next escape vehicle. As the fighter planes try to zero in on the speeding car, Indy barrels around a comer and ducks into a tunnel. One plane veers away, but the other crashes into the mountainside. Its wings shear off completely, but the flaming fuselage continues on through the tunnel as Indy and Henry race to stay ahead of it. There was some discussion of trying to filming the scene full scale. They realized that first of all it would be horrendously difficult to do, especially in a location like Spain where you don't have all the support and materials. But the main consideration was the danger of putting real people in an old tunnel with fire and explosives. So almost immediately it was decided that there was no way to do that practically. So the ball fell in the ILM's court.

"The fire tunnel involved miniatures shot against blue screen and also a complicated miniature set," explained McAlister. "The sequence begins when the plane crashes into the tunnel. We filmed the plane motion control and made wings out of aluminum foil that we stop-motion animated to crumble and break away at the moment of impact. Then for shots of the action inside the tunnel, we made quarter-scale models of the car, the airplane and the tunnel itself which ended up being a hundred and eighty feet long."

Since there was no stage big enough to accommodate the miniature set of the tunnel the filming took place in the ILM's parking lot. It took up about 14 parking spaces for a couple months, which ticked everybody off when they couldn't find a place to park. "But it was worth it," said McAlister. The 210 feet long fire tunnel was built in eight-foot sections, with each section hinged on one side so the top could literally be lifted up like a canopy for accessibility. The camera was affixed to a very complicated sled that was pulled along the left side of the set by a cable underneath the floor of the tunnel. Both the car and the airplane chasing it were also on cables. The models were moving at about twenty to thirty miles per hour and at that speed, things really had to be timed very carefully. The miniature plane was set on fire, sent through the tunnel and filmed at high-speed to help keep the flames in scale. Since the ambient light created by the burning airplane was not enough to fully illuminate the inside of the tunnel, additional lights were concealed within the detail in the over-head rocks.

For close-ups of Indy and Henry inside the tunnel, Ford and Connery were photographed against blue screen in a full-size vehicle. The shots called for a dirty windshield in front of them; but since this would have interfered with the transmission of blue, the glass was removed and later filmed in miniature for optical insertion.

As the action continues, the plane rockets past Indy and his father, proceeds on out of the tunnel and explodes. For that explosion a background plate of an actual tunnel was shot in Spain. The same location was used for the very end of the sequence where a real car with stuntmen drove out and crashed through the airplane debris. For our plate, we shot the exit point of the tunnel with a big explosion going off to give us the correct lighting effect on the hillside around it. Later we shot the miniature airplane motion control as it skids out of the tunnel to the point where the practical explosion on the plate takes place. To get pieces of the plane flying around, we then blew up the model with a miniature explosion. We also shot additional miniature explosions to fill in the gap between the practical explosion on location and the model explosion. To make it appear that the miniature plane was kicking up rocks and dirt before it exploded animated shadows and dirt elements were incorporated into the final composite.

Emerging unscathed from the tunnel, Indy and Henry drive on with the remaining Nazi fighter still on their tail. The plane drops a bomb on the road ahead of them, gouging out a huge crater, which Indy cannot avoid. With their vehicle crippled Indy and Henry make their way, on foot, down to a beach, only to find themselves about to be strafed by the low-flying fighter plane. Sighting a flock of seagulls on the sand, the professor uses his umbrella to frighten them away. The birds fly up into the flight path of the airplane, causing the pilot to lose control and crash into a mountain nearby.

To actualize the sequence, McAlister had to create images of a plane smashing into a flock of seagulls without actually harming any real birds."For the scenes where the airplane hits the birds, we made up a whole bunch of crosses with feathers glued on them. Actually, they looked pretty stupid; but because the shots were quick cuts, all we needed moving through the frame were shapes that looked like birds. We took the miniature plane that we used for the fire tunnel, cleaned it up, put the wings back on and hung it up in the air on a long crane arm. Then we placed a painted backing behind it and had the camera pan past the airplane so that it looked like the background was moving and the plane was actually flying. During filming, I had about a half dozen people dropping these cross shaped feather balls onto the airplane from a grid above the stage."

The feather balls served as the base effect, but to simulate a flock of birds in flight, a slightly more sophisticated approach was employed. "We found flying toys called Timbirds that you wind up like you would an old-time glider airplane," McAlister explained. "When you release them, their wings flap and they actually fly. I believe they are loosely based on a Leonardo da Vinci design from centuries ago. Just by chance, one of our cameramen brought a Timbird in one day and released it just before dailies. I thought it was neat, but didn't give it another thought until about a month later when I was having trouble finding real birds to use as a second element on these shots. So we bought several hundred of these Timbirds, wound them up and released them into the air against a black background. Then we double exposed them into the scene to create a denser flock of birds flying through the frame." For one additional close-up shot of the plane's propeller grinding to a halt, a miniature propeller was filmed with ordinary feathers falling towards it.

The use of special effects for shots of birds crashing into a plane was to be expected. However, even the shots of gulls sitting on the beach had to be created by the special effects team. "For three months I had people in Spain trying to get seagulls," Robert Watts remarked, "but no one could come up with any. Usually, if you go out on a beach and throw a few pieces of fish around, millions of gulls will swoop in; but because it was their nesting season, there were none about. So we ended up using dummy seagulls that we cast in plaster and covered with feathers. We had some standing on the rocks, some on the beach, and some out in the water with anchors on them, bobbing up and down on the waves. They looked fantastic. Even standing ten feet away, it was impossible to tell they were not real. Their feathers fluttered in the wind, which gave them movement. Once we had everything all set up, real seagulls suddenly started to appear in the sky - with all those dummies on the beach, they must have wondered what was going on and decided to check it out." For one additional shot of the birds actually flying up through the scene, Watts rounded up hundreds of white pigeons. Since all they needed to do was move through the frame very quickly, the impostor pigeons were able to pass convincingly as seagulls.

For the establishing shot of the fictional Republic of Hatay in the Middle East Mark Sullivan and the matte department created the striking image of a city at dusk. While Sullivan's painting of the German castle was an extension of a real castle, this nonexistent city was created entirely in silhouette with cutout buildings and telephone poles made from photo-etched brass and sheet aluminum. "We grouped the brass and aluminum pieces together in forced perspective about twenty feet deep and sprayed them black," Sullivan explained. "Then we pumped in a lot of smoke and backlit it which created a pretty realistic effect that was shot by Wade Childress and Jo Carson. To bring life to the silhouettes, we added a matte painting that created the appearance of fill light inside the shadows and rim light on the edges of the buildings. I also painted in the sky, and in a couple of passes we put in some smoke coming out of the chimneys and even a few animated birds flying across the scene." Over the city shot appears a title identifying the locale. Usually such a title would be added later in an optical duping situation; but to avoid this extra generation, the entire shot was created latent image with a holdout matte bipacked in the camera. Then after the various silhouette and painting passes were completed, Childress burned the text into the final image.

Having freed Henry and Brody from the belly of the steel beast, Indy fights it out atop the tank with Nazi officer Vogel. Caught up in their scuffle neither man is aware that the armored vehicle is heading straight for a cliff. When the tank goes over the precipice, it appears that Indiana Jones has indeed made his last crusade. "For obvious reasons," noted Mike McAlister, "the scene could not be done with a real tank and a real cliff. Not only would they lose actors if the tank did not stop at the right moment, but they also did not want it to look like the tank was slowing down as it got closer to the edge, which it would have had to do if it were being shot for real. So we created the illusion of a cliff by combining the full-scale tank footage with miniatures and a matte painting. We started by picking a suitably flat spot over in Spain where we could simulate a cliff edge. We shot the scene with the actors on the tank just as if the cliff were really there. During filming of the live-action, the tank got closer and closer to our imaginary cliff line and then continued right over." To complete these wide-angle shots, a matte painting was added to create the precipice itself. Then for other angles of the vehicle actually tumbling over the edge and crashing into a ravine below, a quarter scale miniature tank built by the Gibbs unit was filmed going over a fifty-foot high cliff in Spain. Additional shots of the tank plummeting into the chasm employed matte paintings and a tank miniature shot against blue screen at ILM. Final shots looking down on the chasm featured the miniature tumbling into a rock quarry located near the effects facility.

Amazingly, Indy once again cheats death and rejoins his friends at the edge of the cliff. With barely a moment to catch their breath, the determined foursome race off to find the Grail temple and block the Nazi pillage. Their quest takes them to the Canyon of the Crescent Moon where they discover a mighty temple carved into the face of a cliff. The initial long-distance view of the full canyon was yet another Mark Sullivan matte painting, augmented with a miniature set built by Paul Huston.

Once inside the temple Henry gets shot in the stomach and Indy has no choice but to face the temple's three challenges. The first, The Breath of God, requires a 'penitent man' to pass. Indy grapples with this cryptic phrase as he walks down a temple corridor and realizes a penitent man would get down on his knees in prayer. Indy does so just in time to avoid being killed by an array of circular blades that slice into the passageway from the walls and ceiling. The blades themselves were a combination of fully operational mock blades built into the temple set and miniature blades blue screened in by the ILM effects team.

Indy quickly disables the device and then moves on to a passageway of stepping stones marked with individual letters. For this second challenge, the Word of God, Indy deduces that he must spell the name of God in order to proceed. However, when he steps down on the stone marked with a letter 'J', for Jehovah, the rock collapses beneath him and he nearly falls into an enormous chasm below. Initially, the scene was filmed with Harrison Ford stepping on the wrong stone and being attacked by a big spider but the result wasn't satisfactory enough and they came up with the a chasm underneath. In the scene, the chasm was realized in a matte painting showing a view from the bottom of the cavern looking up at the place where Indy's foot has pushed through the floor. To create the shot, a stuntman was filmed on a set built thirty feet above the floor of the ILM main stage. The set represented a portion of the ceiling of the cavern, or the underside of the floor Indy is walking on, and was made to look like it was constructed from inlaid stones. The camera was positioned on the floor of the stage looking up to get the correct angle on the stuntman's foot crashing through. "On the wall behind the set," Sullivan said, "I placed a background painting depicting the wall of the cave so that as the stuntman's foot fell through, we could also film debris and dirt falling away from him. The painting was done on a large canvas, fifteen feet wide and forty feet high, that was erected vertically in the main stage. This shot was actually added late in our schedule and I had to do the painting in a hurn, so I literally threw buckets of paint onto the canvas and smeared it around with brooms. Since the background was supposed to be dark and mysterious, it did not need a lot of detail." Once this live-action plate was filmed latent image, it was incorporated into a matte painting of the rest of the cavern. In separate passes, foreground miniatures were shot to add cobwebs and support pillars for the 'safe' stones in the ceiling. To complete the illusion, matte paintings of aerial haze and light streams were double-exposed into the shot.

Indy manages to catch himself before he falls into the cavern. Quickly he realizes his error, spells out Jehovah on the correct stones and then proceeds on to face the final test. The Path of God involves a narrow tunnel that leads out to a ledge overlooking a chasm, with no apparent way to cross over. The riddle tells Indy that he must make a 'leap of faith' so he slowly, and with great trepidation, steps out into the void. To his astonishment his foot lands on something solid, yet he appears to be standing in midair. As Indy proceeds to walk across the abyss to a ledge on the other side, the camera swings to the right and reveals the trick, Indy is walking on a bridge painted to match the rock on the opposite cliff face, thereby making it invisible from his original vantage point. The camouflaged bridge was a Spielberg inspiration. "I thought it would be interesting if somebody hundreds of years ago painted a false perspective on a bridge that matched the terrain two hundred feet below in color and texture. Of course, thinking the idea up and having it sketched was the easy part. We never knew for certain if it would work until ILM got involved and made it happen."

"The leap of faith shot was an example of a scene that could not exist without the use of special effects," McAlister said. "When the shot was first conceived, there was great discussion in England on how to do it and Douglas Slocombe, the director of photography, was pretty sure he could come up with something there that would work. But Elliot Scott was very worried about whether or not they could accomplish it practically. Also, it would have been much too expensive to build a full-size set because the sequence only involved four or five shots. So the buck was passed to us early on, and I was sure, that if anyone could accomplish it, we could, partly because we would have more control over the elements."

ILM's solution to the leap of faith dilemma was to use a combination of matte paintings and miniatures. The first shot in the sequence was a matte painting by Yusei Uesugi. It was the basic bottomless pit shot, looking down the side of an enormous rock cliff that just goes on and on into darkness. Since this was supposed to be Indy's point of view, his feet are visible in the frame. The feet and the rock that they are standing on were miniatures built by Paul Huston. They found some small boots about one-third the size of the real boots used by Harrison Ford, and positioned them over a miniature rock placed in front of the painting. Then they puppeteered them a little bit so they would not look like stiff shoes.

The remainder of the leap of faith shots involved a model of the bridge, nine feet tall by thirteen feet wide, that was carved by Huston out of green styrofoam. To paint the bridge, the camera was placed in the starting position of the shot so that Huston could view the model from the same angle as the camera. "In the start position of the shot," said McAlister, "the camera was at a very high angle looking down at the bridge and the cliff below. Paul looked through the camera and first sketched on the bridge the detail that he saw below. If, for instance, there were certain rock formations way down low in the abyss, he painted those same formations on the bridge so that through the lens the bridge appeared to blend in with those formations. As long as we photographed the bridge from the same angle that Paul painted it from, it was impossible to tell that the bridge was there." To facilitate the painting process, 35mm Kodacolor print film was loaded into the Vistavision camera. As Huston painted, stills were shot every hour and developed at a nearby one-hour photo store, enabling an on-going assessment of the work in progress.

The illusion of invisibility is broken immediately when the camera moves off its initial axis. Once all the lines and textures on the bridge no longer match up with those on the cliff wall behind it, the nature of the bridge becomes fully evident. This camera move, crucial to the success of the illusion, was actually determined many months earlier during live-action shooting on the ILM blue screen stage. "We shot Harrison Ford against a blue screen long before we built the miniature bridge," McAlister noted. "We did a shot of him stepping out into supposed midair, realizing he is on something solid and then starting to take a few steps forward. I had to imagine as well as I could the move that we would ultimately do on the miniature and then make that move on Harrison. It was important that the perspective change on Harrison match the perspective change we would later do on the miniature. In a way, it made sense to shoot Harrison first because then we could conform our miniature shoot to whatever restrictions we had on the live-action."

Since Ford was filmed in a blue screen environment, he did not have a shadow, which was clearly needed since a strong shaft of light was streaming into the finished scene. To create the shadow, a miniature puppet figure was photographed on the miniature bridge. A stop-motion puppet was positioned on the bridge right where the blue screen element was going to be composited and then imitated Ford's movements. They did one pass with the puppet on the bridge and then one of the bridge itself. Later they split-screened the puppet out of the shot, leaving only its shadow. Then they matted Harrison in and he suddenly had his shadow. Indy successfully reaches the ledge on the other side of the abyss and tosses a handful of dirt onto the bridge to aid those who are following him. "This shot was created by first filming Harrison on a small mockup of the ledge that was constructed at the studio in England. Then on a black stage we shot handfuls of dirt being tossed onto a cloaked platform built in the shape of the bridge. We found the take that best matched the plate of Harrison pretending to throw gravel, then projected the live-action into the miniature and matted in the dirt falling onto the bridge shape."

For the death scene of the film's villain, Donovan, Lucas and Spielberg played with an idea that was originally explored for the opening of the ark in Raiders. The idea was that all the Germans at present would be disintegrated, with their skin being rotten and falling off their faces and then with their bones being turned into dust. Unfortunately, the technology at the time could not support such a venture and the idea was dropped. Inspired by this, Boam wrote a scene where Donovan was supposed to age in a matter of seconds and wither away until he is only skeletal dust. The effect, tagged at ILM as 'Donovan's Destruction', was considered common enough for the special effects of the late 80s but was given a different wrinkle by Spielberg. "I would not agree to Donovan's destruction unless it could all be done on camera in one continuous shot," he stated. "I just did not want to do a series of cutaways so that the person could be advanced in makeup. We've all seen that, and I think people do have a level of expectation with these movies. They expect to see certain things that have not been done before. Also, the minute they know ILM is involved; they have a very high expectation that the people up there will somehow top themselves. So things like having the decomposition in one shot are important, I think the ILM crew would be disappointed in me if I did not offer them a challenge like that."

To meet the Spielberg challenge, Mike McAlister and his team had to incorporate three different approaches. First, Julian Glover spent three days to shoot the needed scenes. It began with six takes with progressively older make up, then another day with inflating pads taped to his forehead and cheekbones. As these were pumped full of air, his eyes seemed to recede into their sockets. To make him grow instant long gray hair, he was fitted with a wig into which the hair was drawn back mechanically. Then the film was run backwards at higher speed. Three motion-controlled puppet heads were filmed showing Donovan in advancing stages of decomposition. Then to blend the heads together so they appeared to age seamlessly, a variation of the computer 'morfing' technique pioneered on Willow was employed. "When Donovan first starts aging," said makeup effects supervisor Stephan Dupuis, "you see him from behind as his hair grows long. That initial shot was done in England. Then they cut back to a reverse shot over Elsa's shoulder. That's where we came in with Donovan already aged slightly. We took a head cast of Julian Glover and made a latex head built on top of a torso. Inside this head were motion control mechanics designed by Kelly Lepkowsky. The first head went through programmed moves that made Donovan's cheeks suck in and his nose go back so that he looked a little like the Phantom of the Opera. When that head reached its most decayed position, we took a cast of it and made a second head. We attached that head to the same motion-controlled rig and made it age even further. Donovan's nose went completely in, the eyelids shriveled up and went inside his head, and the mouth curled back even more. At that point he looked like a mummy, and from this we cast the third head."

The first two heads had simple foam latex skins, but the third one was more complex because it had to shrivel down to a skeleton. "We started with a gray skull underneath," Dupuis continued. "Over that I needed material that would shrivel up under heat like saran wrap. I looked at different materials and finally just took styrofoam cups and melted them in a solvent until I got liquid. Then I painted this liquid on the inside of the mold for the third head. When that dried I had a very thin plastic skin that served as the outside skin on the face. Between this skin and the gray skull underneath we placed pieces of Shrink hard-a plastic that shrinks when you heat it. We cut out pieces in worm-like configurations and pasted them on the inside of the skin so that when the face was heated up, those things would shrivel and curl and open up the skin. They looked basically like a bunch of maggots, pretty gross, but on the screen you do not really see that much of it."

All three heads were placed on the same motion control rig so that they would go through the exact same motions in the frame. "We needed to have the three heads in perfect register for editing," McAlister explained. "As Donovan disintegrated, we wanted to be able to cut from the first head to the second at the best possible point, and then from the second head to the third. By having all three heads go through the exact same motion in the frame, we could then choose any point to make the transitions between them." Once the puppeteered heads were filmed, the next step was to blend them together seamlessly using digital imaging processing. "We first used the morfing technique during Willow for the transformation of Raziel from a goat to an ostrich to a turtle to a tiger and then to a human being. On Willow they had the benefit of cutaways, plus the animals were so dissimilar in shape that no one could say how they would actually transform. But for Indiana Jones we had to transform a human face, something that people are very familiar with, and we had to make our shots look as realistic as possible. The other major difference was that on Willow they only morfed individual elements that were then composited optically. But for Indy we set out to create an image that was almost completely composited inside the computer so that what we scanned back out was essentially the final shot. As far as I know, this is the first time anyone has ever accomplished this type of digital compositing for a full-screen, live-action image."

To perfect the transition, McAlister first tested optical cross- dissolves at the points he thought the head changes should take place. Once the transition points were firmly established, the original footage was then scanned into a Pixar image-processing computer and translated into digital format so that computer artists could begin the time-consuming process of making the transitions appear seamless. "Basically, we had to pretend that the picture was on a sheet of rubber," explained digital compositor Les Dittert. "With morfing we were able to grab part of that picture and stretch it. For instance, if we had to line up the outside of the second head with the same frame number of the third head, we used the computer to shift around this imaginary sheet of rubber so that the edges lined up perfectly. By doing that, we were able to correct any misregistration on the photography as well as line up features within the faces. We could also control the speed of the dissolve to different areas of the frame. For instance, we could make the nose of the second head come through before the rest of the face. Something like that would be hard to do in a normal optical. In optical you can realign for one thing, like the tip of the nose, but then the side of the head would probably be off."

Making the morfing even more complicated was the fact that the second and third head imagery had to be placed directly over the first head. In the scene, the camera watches Donovan decompose from over Elsa's shoulder. As a result, the first head was shot on a torso with a double for Alison Doody in the foreground and a portion of the chamber set in the rear. The double was then sent home and the other two heads were filmed without her. "The background and hair of the first head were used through the entire shot," Dittert said, "and the other faces were mapped over the first one. That presented problems because the last head was practically a skeleton and it was a lot smaller than the first head. If we had just matted the last head in, audiences would have seen what we called the 'peek-a-boo effect' where the first head would have peeked out on one side or the other because it was bigger. So we had to do a little face-lift and pull the whole face over." Because Donovan's clothing also had to age, some miniature clothes were shot on a fourth torso puppet, also motion-controlled. "We had two clothes elements," said computer graphics animator Doug Smythe, "the second more deteriorated than the first. Both had to be mapped over Donovan's upper body and shifted into position so they would not expose the torso mechanism. Those were shot against blue screen and scanned into the computer. Then the blue screen was extracted and the clothes were morfed to make them fit where they had to go."

To complete the effect McAlister needed some information on how to rapidly deteriorate clothing without using dangerous acids. He called an association of dry cleaners, who referred him to a clothing damage expert in Utah. Utah sent them to Mary Baker, who was employed at the textile research facility at the Smithsonian Institute. McAlister explained his situation and asked for any advise. The reply that came over the phone sounded like a joke, "Have you tried ILM? They can do anything." After a chuckle, McAlister responded, "Ma'am, we are ILM." As it turned out, Mary Baker was involved in preserving Yoda for the Smithsonian's traveling exhibit on Star Wars. Eventually she came to ILM to train the crew in the handling and use of the acids involved.

Once the clothing and the faces were successfully merged in the computer, the whole shot was digitally composited and scanned back out in Vistavision format. Ultimately, this image was not the final composite because it had to be reduced in optical to four-perf projection format. At that time, an additional smoke element was incorporated to help it match other shots in the Grail chamber.

Donovan's destruction is complete when Indy pushes the skeletal remains away from a screaming Elsa and the bones fly against a chamber wall and shatter. To create this effect, a skeleton was assembled from bones made out of a brittle polymer. These hollow bones were then filled with fuller's earth and a ground glass product called Cab-I-sil and hung together by wires on a trapeze-like device. "Donovan's skeleton looked like a giant marionette," Dupuis recalled, "which we swung into a mockup of a cave wall. All the bones were supposed to release on impact, but we had to do several takes because there was always one bone that would not detach and would still be floating in midair after the rest had crashed into the wall. We finally had our pyro expert, Bob Finley, attach explosive squibs to the wires on the trapeze. When the skeleton hit the wall, he pressed a button and all the wires broke on cue. The shot was so quick and there was so much dust and smoke in the shot that the wires were easily concealed." Equal to Belloq's explosion in Raiders Donovan's Destruction was one of the film's most thrilled scenes and at the same time marked a technological breakthrough in effects work.

Armed with the authentic Grail, Indy races back through the temple and pours water from the sacred vessel onto his father's mortal wound. The Grail proves its healing powers when the elder Jones' wound mends right before Indy's eyes. The healing of Henry Jones was made in two cuts. The first was a make-up effect involving baking soda and vinegar. The makeup had a high content of baking soda so that when the vinegar was poured onto it, the baking soda fizzed up like hydrogen peroxide. In the second cut, the wound appears to be foaming up even more and when Indy pours more water on it the wound washes away revealing healed flesh. For this shot, the effects team simply sprinkled baking soda on the actor's skin, dumped vinegar on it until it got real foamy, turned the camera on and then washed the foam away with more vinegar. It was shot more or less live-action style with a four-perf camera and did a bunch of takes until they got one that looked right.

"The most common challenge in effects is not to come up with brand new techniques, but to find new ways of using old ones," comments Micheal McAlister, visual effects supervisor for "Indiana Jones and the Last Crusade." McAlister was one of the four filmmakers who won an Oscar for the special effects in "Indiana Jones and the Temple of Doom."

"All effects are geared to the story," McAlister says. "They enable it to be told the way Steven wants to tell it." He reveals that altogether 80 shots in the film involve an eclectic variety of visual effects, including blue screen, matte paintings, and creature puppet effects combined with computer graphics in a process he calls "morphing technique," in which there is a metamorphosis of one image changing and blending into another.

The creation of the visual effects began during principal photography with the filming of backgrounds and blue screen live- action photography. Similar to the building of the soundtracks and the editing of the film, the effects are accomplished during post production.

In the movie, Indy must interpret a series of riddles to avoid certain painful death, and the object of these riddles was usually carried out with special effects.

"Many of the effects in this film broaden the scope of danger in a scene," McAlister states, "and some, like the aerial chase sequence, couldn't have been done without visual effects."

'Indiana Jones and the Last Crusade' has a collection of all the highlights any sound editor could dream of: explosions, fire, gunfights, earthquakes, windstorms, and chases in all sorts of different vehicles," sound designer Ben Burtt says. "I could go to my sound effects library catalogs and no matter what page I turned to, there would be an appropriate sound for a scene in the film."

Although Burtt has built a large Indiana Jones sound library since working on "Raiders of the Lost Ark" (which brought him an Oscar in 1982), he states that he and his associates always look for original sound material for each film. For "Indiana Jones and the Last Crusade," special new recordings were made for such sounds as horses, crashes, aircraft, a tank, and artillery. Several hundred sound effects were collected on 41 tapes with an average of 20 effects on each tape - around 800 in all.

After Burtt began his work by categorizing the sounds he planned to use, he started to collect them, on one occasion riding in a biplane upside down while recording. Traveling to find the best sounds possible, Burtt attended the destruction of a wind turbine over 300 feet high to record sounds to be utilized for crash scenes. Gunshots and ricochets were created and recorded in isolated locations in Utah and Texas.

The recording of a styrofoam cup at a family picnic became the basis for a roaring inferno after Burtt multiplied the sound at different frequencies with a computer. "The digital process of manipulating sounds-stretching them, shrinking them, or changing the pitch-is one of the big steps forward in the last ten years. Most of the things we now do with a Synclavier computer we used to do with tape and cutting."

"Indiana Jones and the Last Crusade" is being released in selected theatres in a special sound format called 70mm Full-Field Sound. "With this format we can move sounds not only from front to back in the theatre but from side to side," Burtt says.

"A film like this is brought to life with sound, which dramatically enhances and brings a natural dimension to the special effects," Burtt declares.

The Last Crusade was being released in selected theatres in a special sound format called 70mm Full-Field Sound. "With this format we can move sounds not only from front to back in the theatre but from side to side," Burtt said. "A film like this is brought to life with sound, which dramatically enhances and brings a natural dimension to the special effects."