Shadowgraph of a Mouse

by John Holbo on November 19, 2017

I’m reading a fun book, Shadow of a Mouse: Performance, Belief and World-Making In Animation, by Donald Crafton. The author is an animation historian/film studies scholar. I’m interested in the history but also – as is the author – the theory of animation ‘performance’. I’ll snip a nifty bit from Chapter 4, about the evolution of devices, conventions and styles for handling space. The author uses the evolution of the treatment of shadows as a nice hook, per his book title.

Convincing shadows could be rendered intuitively with relative ease when they fell on the floor or ground of the background. If the shadow fell on a curved or oblique surface, however, the task became enormously more complicated. To facilitate rendering these sorts of shadow, the Disney studio developed a device they called the Shadowgraph, and in September 1936 they applied for a patent for it. To use the Shadowgraph, first the background art would be photographed. It would then be used to construct a miniature stage with three-dimensional cardboard objects, such as tables or rocks, placed in their proper location on the floor plan. Next, the transparent cel with the character that was supposed to cast the shadow would be set up in front of the set and a strong light would be shone through it. The character’s opaque outline would then cast a real shadow upon the set and the props. As Bob Martsch of the special effects department described it, “The shadow necessarily follows the animated cells, is made by them, and distorts and changes over the three dimensional background which has been built to register with the layout of the scene. The shadow may be made long or short, to move over curved, flat or angular surfaces in true relationship to the character casting the shadow.” A camera above the set photographed each cel’s shadow. This operation was repeated for every cel in the sequence. Using these photographs, specialized layout artists would ink the shadows on top cels in their proper place to be included in the scene setup. As an example, Martsch projected a scene from Snow White. Throughout that film, not only does the transparency of the shadows vary according to the strength of the ambient lighting, but the shadows are also true to their sources. When Snow White ascends the cottage staircase, for example, the candlelight throws shadows onto the stairs and floor in perfect perspective. Furthermore, when the candle flame flickers, so do the woodland animals’ shadows (Plate 8). Accurate as a sundial, these custom shadows have replaced the generic noonday kind. So when our heroine sees the dwarfs off to work we see the slanted violet-tinged shadows of dawn. When the witch comes calling and her engulfing shadow heralds her, it’s an effect borrowed from German expressionist cinema, as when the vampire’s shadow moves across the heroine in Nosferatu (F.W. Murnau, 19 22 ).

These shadows were a very important component of the studio’s performance project because they break down the division between foreground and background …

In these Disney films from 1937 and in the features that followed, it is no longer meaningful to speak of the shot’s foreground and background, since now the entire scene is camera space, a common, modulated performance environment that incorporates the actor. Rather than calling attention to the biplanar nature of animation, this staging exists to facilitate an embodied performance. With his usual poetic flourish, Feild [Robert Feild, author of The Art of Walt Disney (1942)] enthused about the organic symbolism of this phase in the development of performance space for Disney’s cartoon actors: “From the moment they were conceived as embryonic beings drifting around in undefined space, they have been inseparable from their environment. Yet when the Studio comes specifically to define this transcendental realm, it has no better name for it than ‘background’!”

Reading this, I couldn’t quite figure out how the hell this Shadowgraph wossname is supposed to work, but then I found that someone had actually posted Disney’s patent submission for it. Now I get it. It’s like a little shadowbox space that sort of goes under the Acme animation stand. How clever!

You can get an idea about the sorts of results this device produced by looking at this clip, for example.

What’s that, you don’t know what an Acme animation stand is? Google it. It’s pretty cool.

OK, just a few notes about how I got here. I started with research into the history of caricature, and the huge r&d breakthroughs in caricature tech Disney pioneered. (No joke!) But Don Graham – whose book is great! – played an interesting and important role in all that.

Although Disney has been credited with the innovative exploration of performance space, it seems clear that the idea originated in discussions instigated by Graham, the teachers from Chouinard, and, of course, the art student-animators themselves. It is likely that Disney’s awareness and excitement about dimensionality flowed upward from his employees. Significantly, in the eight pages of the 1935 memo to Graham in which Disney elaborates his “scientific” approach to caricature, rhythm, gesture, and character, he never mentions the ability to add depth to scenes as desirable. His concern is with creating movement and “caricaturing life” to create performances. However, Graham and the animators knew that this could not happen in a spatial vacuum; the same principles that drove the push to embody performances were behind the need to create encompassing acting spaces.

It isn’t just that they realized there’s this thing: the illusion of space. It’s more than that, trust me.

Enough nerding out for me for tonight. I love this stuff.



mjfgates 11.19.17 at 7:01 pm

This is how innovation *always* works; the guy at the top says “make it work the way I want!,” the people at the bottom figure out SOMETHING, and off we go. I’d bet money that it was several of Rembrandt’s apprentices, working in concert, who figured out how to make a Dutch merchant look good in oil paint…


Michael 11.20.17 at 12:11 am

This post goes nicely with an article about Second Life by Leslie Jamison I just read in the December 2017 issue of Atlantic. The cartoon world and Second Life both adumbrate the great depth of worlds we imagine / make others imagine, alongside the collision of those imagined worlds with the unbridled unpredictable detail-richness of the not-imagined.


John Holbo 11.20.17 at 1:44 am

I’m glad this post got more than zero comments! I was worried it might be a goose egg.


Michael 11.20.17 at 11:35 am

Better than a goose egg. You’ve made me buy Shadow of a Mouse. I’ll display that prominently beside How to Read Nancy in my office. Should lull them into a false sense of security.


John Holbo 11.20.17 at 1:12 pm

“I’ll display that prominently beside How to Read Nancy in my office.”



Glen Tomkins 11.20.17 at 2:17 pm

Who knew that Disney was a Plato fan?

We’ve found skiagraphia in the Disney oeuvre. What next, the ring of Gyges?


Kenny Easwaran 11.21.17 at 4:16 pm

Unfortunately the clip you told us to watch to see the effect has been removed from YouTube by its owner!


Evan 11.21.17 at 10:05 pm

Could you tell us what the clip was, since it no longer exists? We might be able to find it somewhere else.


John Holbo 11.21.17 at 11:14 pm

The clip is of the scene when the dwarfs return home and the house has been cleaned and Snow is in the bed. There is a lot of candle-play shadowplay over the walls as they go up the stairs and such.


Michael Cain 11.22.17 at 4:41 pm

There is a lot of attention-catching shadow work in the Sorcerer’s Apprentice sequence in Fantasia.


Collin Street 11.23.17 at 10:33 am

There’s been a bit of a discussion in anime technical circles about basically the opposite problem: how or if to maintain the visual distinction between foreground and background when it’s all CG and the original technical distinction has vanished.

The most straightforward approach is to use different rendering styles for things coded “foreground” and “background”, but we’re seeing some more interesting choices; this season’s Land of the Lustrous is [I haven’t seen more than clips] putting a lot of the load on the colour palette, with the “foreground” bits in colours reminiscent of strongly-tinted cel paints and the “background” bits in more subdued gauche-esque tints.

All sorts of skeuomorphic choices.

Comments on this entry are closed.