In THE LAYOUT MANUAL section titled 3 DIMENSIONAL MODELS, there are descriptions for several different methods of shooting models, primarily planes, on either the 11F Crane or the Horizontal Multiplane camera rig. Each method employed was based on the desired effect that could be achieved. Shooting models frame-by-frame was not new to the Disney Studios, it was a technique used on Pinocchio (1940) for the gypsy wagon, bird cage and stagecoach. By filming these models, the animation department was then able to rotoscope or draw over the film images to get a more accurate process of creating the animation. This method saved both time and money during production while yielding better quality drawings than if it was solely done by hand with little or no reference.(1)
The text in the 3 DIMENSIONAL MODELS section states that, “certain 3 dimensional models, (airplanes mounted on sticks,) are usually shot on the 11 F crane. These models have limitations, in that they must always be shot from above and the movement is limited to rotation about their own axis.”(2) Most of these models varied in size but were all relatively small. These models were mounted on the “light-box attachment,” described in The Layout Manual, Part 3—The 11 Field Crane,(3) that allowed for horizontal moves under the camera.(4) The manual continues, “the rotation is done by hand and by eye and is surprisingly accurate, but a few frames leeway must be allowed for maneuvers, for the flick of a cameraman’s thumb is not always right to frame.” In essence, the timing is reliant on the cameraman’s experience and varies in the same way that two animators’ timing would vary when creating an animated action—neither is wrong, just different.
The manual makes the point of advising that a layout be created with the path of action for the directional move as well as notation for the planes degree of tilt. This layout would then be used for the pantograph crosshair cup to create the frame-by-frame movement under the camera. If a scene had more than one plane traveling in formation, than a path of action layout was created for each plane and shot as a separate pass and “doubled in.” This means that after each model plane is filmed against a black background, the film in the camera is rewound back to frame one and then the next plane is filmed. The black background preserves the film emulsion allowing for the multiple passes on the film. This process is repeated for each plane in the formation that is to be created. The manual does make the note that “exact synchronization is difficult resulting in probable jitter in the relationship between the planes.”(5)
The use of models in some scenes that had a viewpoint or action which made the model supports visible required some additional work. The manual notes that “it is possible to actuate the model by hand, then rotoscope it from the film and ink and paint the cels” in the traditional cel method of animation. Rotoscoping was still an efficient way to create accurate animation provided that the action was at a sufficient speed. If the action was extremely slow “some other method would have to be found” as the rotoscoping of a very slow moving model would be fraught with undesirable jitter. The mere fact that the drawings are being traced off live-action introduces the variances associated with hand drawn animation. It would be simply too time consuming to do slow action models rotoscoping efficiently for these training films.(6)
Additionally, bullets or other effects could also be doubled in using either the “sliding cel method” or using a “bottom-light” effect that is also doubled in to the film. Both of these methods will be discussed in more detail later.
The manual notes, in regards to filming models, that “with judicious cutting, this method produces excellently realistic results” and points to examples in FIXED GUNNERY AND FIGHTER TACTICS, also known under the code name Jacksonville Project, Prod. 2648, created for the U.S. Navy in 1943.(7) In this particular film, the models were filmed in color.(8)
There were two additional methods of shooting models that allowed for greater “maneuverability” and also larger size plane models. Both methods utilized and were planned for the HORIZONTAL MULTIPLANE crane setup. There were some ingenious devices created for use with this camera rig that allowed for very realistic animation to be shot frame-by-frame.
The first method was using models suspended by wires for shooting on the HORIZONTAL MULTIPLANE crane. The text notes that “the wires are painted to match the general tone of the background, the lower ends being attached to finned weights sunk in a drum of heavy oil to dampen out vibrations.” In other words, the cameraman had to devise a way of keeping the model stabilized on the wires, so aside from the models hanging from above by wires, there were also wires that attached to the “finned weights” that were in the heavy oil below. This kept the models from swaying or moving slightly as they would do if only hanging from above—even the slightest air current from a door opening could make a hanging model move and ruin the shot.
The other method of shooting models on HORIZONTAL MULTIPLANE crane was to attached the nose of a plane model to a large piece of plate glass “with the propeller on the other side, joined to the plane thru a hole in the glass.” The glass plate was mounted in a metal frame “which is calibrated to rock forwards and backwards as well as to rotate on a vertical axis.” The model plane could also rotate on its own axis and the prop was spun using an air hose. This method was able achieve “most aerial acrobatics” needed to for any of the training films being made by the Studios at that time that required the use of models.
Finally, under the 3-DIMENSIONAL MODEL section of the manual is an attachment for the Horizontal Multiplane that incorporates a three dimensional revolving globe. The Earth is mounted in a solid rectangular holder that is painted flat black, “to avoid reflections.” The Earth is rotated, using a calibration control from behind, as indicated in the accompanying diagram.(9) The principles employed in the globe attachment device “can be adapted to many uses” such as “machinery mockups particularly.” In other words, this device could hold say a gyroscope instead of the earth or some other spherical object that could be animated frame by frame.
The last page of this section notes that “better results have been achieved more economically by using models that could have been done with animation.” Certainly, much more film footage could be shot with models on a camera rig in a day than could be drawn by hand and the models would offer a more realistic look especially for these military training films.(10)
The Disney filmmakers had choices that they could make as to how best to utilize the animation methods at their disposal to create the WWII training films quickly while always maintaining the quality that the Disney Studios prided itself on. Every training film that was contracted to Disney during those war years gave the artists challenges that they met with a robust toolbox of animation techniques and equipment. It was the ingenuity of these artists and technicians that allowed for some of the devises and rigs discussed here to be invented and used.
At the very end of the 3-DIMENSIONAL MODEL section there is a note that is underlined and in all capital letters that says “ALWAYS CHECK WITH UB IWERKS WHEN PLANNING SCENES WHICH YOU THINK MIGHT USE 3 DIMENSIONAL MATERIAL.” Of course, Iwerks was one of those geniuses at the Disney Studios who was at the forefront of inventing some of the devices and rigs that were needed to create the animation for these WWII training films. The LAYOUT MANUAL is one of the few documents available that speaks to some of Iwerks’ inventions during the WWII era. The recently published Walt Disney’s Ultimate Inventor: The Genius of Ub Iwerks only briefly mentions the pantograph/lightbox setup for the 11 Field Crane.(11) But, there is no mention of the Circular Glass Disc attachment for the vertical Multiplane Camera, or even the Horizontal Multiplane Camera rig that was used extensively for shooting 3-Dimensional models. There is a brief mention of the original 1933 Horizontal Multiplane Camera that Iwerks invented for traditional animation and used at his own studio.(12) Although the book is technical in nature, it does cover many of Iwerks’ achievements in the live action and theme park arenas after the war.
Next month, I will be discussing the extensive and varied techniques of SLIDING CELS to create animation including rotation cels, rotation cel construction, Double rotation cels for slot-gags and straight off sliding cels all of which were used to create inexpensive yet highly effective animation for many of the WWII training films at the Walt Disney Studios.
©David Bossert 2020
FOOTNOTES
1 – Models, Miniatures, and Movie Magic, by Kevin Kidney, The Walt Disney Family Museum blog, July 8, 2011
2 – The 3 DIMENSIONAL MODELS pages, Layout Manual, Walt Disney Productions, 1943; authors copy.
3 – Bossert, David A., Walt Disney Classified: The Layout Manual, Part 3—The 11 Field Crane, Cartoon Research,
February 2020
4 – The 3 DIMENSIONAL MODELS pages, Layout Manual, Walt Disney Productions, 1943; authors copy.
5 – The 3 DIMENSIONAL MODELS pages, Layout Manual, Walt Disney Productions, 1943; authors copy.
6 – The 3 DIMENSIONAL MODELS pages, Layout Manual, Walt Disney Productions, 1943; authors copy.
7 – Shale, Richard Allen, Donald Duck Joins Up: The Walt Disney Studios During World War II, The University of
Michigan, Ph.D., 1976, Appendix C, page 292.
8 -The 3 DIMENSIONAL MODELS pages, Layout Manual, Walt Disney Productions, 1943; authors copy.
9 – The 3 DIMENSIONAL MODELS pages, Layout Manual, Walt Disney Productions, 1943; authors copy.
10 – The 3 DIMENSIONAL MODELS pages, Layout Manual, Walt Disney Productions, 1943; authors copy.
11 – Iwerks, Don, Walt Disney’s Ultimate Inventor: The Genius of Ub Iwerks; Disney Editions, pg. 57
12 – Iwerks, Don, Walt Disney’s Ultimate Inventor: The Genius of Ub Iwerks; Disney Editions, pg. 48
David A. Bossert is an award-winning artist, filmmaker, and author. He received his B.A. from CalArts School of Film and Video with a major in Character Animation. As a 32-year veteran of The Walt Disney Company, he contributed his talents to The Black Cauldron (1985), Who Framed Roger Rabbit (1988), The Little Mermaid (1989), Beauty and the Beast (1991), Aladdin (1992), Tim Burton’s The Nightmare Before Christmas (1993), The Lion King (1995), Fantasia/2000 (1999), and the Academy Award-nominated shorts Runaway Brain (1995), Dali/Disney Destino (2003), and Lorenzo (2004), among many others. Bossert is now an independent producer, creative director, and writer.
I’m intrigued that scenes involving more than one plane were made using only one model, rewinding the film again and again to photograph each flight path. One might think that it would be more efficient to use multiple models and shoot the scene only once, as there would be fewer steps in the process; but because each individual step would be much more complex, the likelihood of error would be much higher. The aim seems to be to plan the scene so that each step in the process is as simple as possible — a basic tenet of assembly-line production.
In practice, was there any limit to the number of passes that could be made on a given length of film?
I’ve actually had to do a shot like this.
What usually happens is that the production wants the planes way too close to one another than what happens in reality. When a cameraman tries to put their hand into the scene to pivot one plane, you inevitably brush against another plane—causing an unwanted movement.
It was best to film each plane separately except when they visually overlap one another (otherwise you’d have to optically print them or use a bi-pack matte as I had to do several times).
Scenes with several clay or felt puppet models can have the same problem. Wood and metal models don’t have this problem unless they have cloth clothing.
And I thought I could answer your other question separately.
Yes, there were limits as to how many passes one could do. The more passes you made, the more likely that the blacks would turn milky and no longer be a solid black.
This problem ran rampant in photographic special effects (up until the 70’s)? FANTASIA is filled with this phenomenon. You can easily see it in 35mm prints. I’ll have to check some Blu-ray copies some day.
All cameramen & camerawomen seem to be able to remember the most passes they ever had to do. Mine was pretty low. I don’t think I ever went above 12 nor directed anyone to go above that. I tried to avoid it. I did not work on TRON but I saw several scenes being shot that went way beyond 12 passes. Each color or glow seemed to require its own f-stop and therefore its own pass.
Looks like the spinning globe rig is what they used for the True-Life Adventures titles. Similar rigs were also used in Fantasia for the snowflakes in “Nutcracker Suite” and the shot of early Earth in “Rite of Spring”, according to the Hermann Schultheis notebook (as seen in John Canemaker’s “The Lost Notebook”).
Dave, thanks for sharing this interesting information. Just to clarify, though: the 3-dimensional models in “Pinocchio” were not filmed frame by frame, but shot in continuous motion. The 35mm frames of the birdcage were enlarged as photostats, then traced — more or less the same thing as rotoscoping, as you suggest. The frames of Stromboli’s wagon were blown up and printed on washoff relief cels, then painted and used in the finished scene like ordinary cels. I wrote about this at some length in “Pinocchio: The Making of the Disney Epic.” You can see this passage on pp. 142-144, along with illustrations of the filming setup for the wagon.
Wow. Fascinating article. Thanks!
Hello David!
Another fascinating article! I’m always shocked when other people read this stuff and actually comment!
One aspect you might cover is how WDP shot the book openings of the classic animated features (SNOW WHITE, PINOCCHIO, CINDERELLA, SLEEPING BEAUTY, etc).
These shots all have some elements in common (all 4 corners of the book in frame during the opening of the book?) and some elements not in common (like candles burning)!
I’ve often wondered if most of them were shot on the horizontal multiplane? And, if the animation of the book opening and pages turning were done with black wire/thread. Perhaps you could cover this in a future article?