Pepper's Ghost Puppet Theater

Research & Concept

The goal: to create a puppet theater that uses a built-in Pepper's Ghost illusion. This theater should be about the size of a photo booth, and must be easily transported and set up at different events. It also should be adaptable, so that multiple short pieces could be performed within the same space, with swappable scenery and simple tech. Importantly, of course, it can't be terribly expensive.

Concept/Initial Musings:

I am setting out to make a Pepper's Ghost illusion of my own. The only issue is, the majority of how-to's focus only on creating the illusion using a screen or projected image. In fact, if you looked through any of the above links, you'll find that there is a complete full stop between the old fashioned analog creations of the late Victorian, and the modern use of screens in the illusion. Is nobody using the analog form? Well someone is, of course, Disney's Haunted Mansion features a incredibly impressive ballroom scene! But that's... kinda it? Future research - weeding out these digital imposters - leaves just a handful of small old-school carnival tricks and educational models. Nothing near the size of the theatrical spectacle that it once was.

So there's my inspiration: to break new ground (or rather, very old ground). But also my issue: there was no one to follow. Concepts and plans and measurements yes but not many tutorials. So I'm venturing out into new territory here, but I hope this log will provide some good inspiration in case anyone plans to follow in my footsteps.

Once I have the theater finished, I'll need to have the same styles in mind when it comes to designing the puppets. They'll have to be made specifically for the Pepper's Ghost. What I've discovered through early experiments is that you can actually have too much light - that is, the reflection works too well and you can't see the background through the reflected figure. So a lower level of light is needed to sell the effect, but this means that the subjects need those bright colors and stark shading to be dicernable. I'm sure that that's pretty obvious, and most people would go with that style anyway to sell the ghostly mood, but the practical reason is important to keep in mind. I struggled with the visibility of characters with more muted tones.

My end goal for this theater is to take it to night markets, theater lobbies, places like Maker Faire, and anywhere else that people can enjoy it. Because of the limited sightlines needed to really sell the illusion - as well as the possible light leak leading towards the need for a curtain around/over the audience - it's likley that only 1-3 people will be able to view it at a time. So my goal is to create a space to host 5-8 minute pieces and frame it as a bit of a sideshow attraction, likley charging very little if not just asking for donations, in a low-key environment where people can casually wait their turn. Preferably with a barker/host to manage crowds, and a couple puppeteers hiding underneath.

That's another important note. For it to be reasonably comfortable for the puppeteers.

It surprises me just how little this illusion is made these days using the analog form. It's not expensive to make in a medium to small scale, and just takes some careful adjusting of sight lines. I think at this point, many people associate it entirely with Disney, and that makes me irrationally sad. Why should they have the monopoly on cool effects? Especially ones that were already a century old when Haunted Mansion was built? My point being, people associate it with these huge expensive projects, when in reality it's a very simple process. It deserves to be demistified.

First Test (2023)

This first test was created in October of 2023 after my initial interest in the concept.

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Integer sit amet ex quis nisi auctor mollis. Phasellus nec nunc sit amet purus mattis sodales.

Very early concept sketches.

My initial design used a mirror along with the pane of glass. I wanted to create a style that would be the most comfortable for a performer, and hopefully hide some of the tricky sightlines that are a common issue with mirror illusions. You can also see me here drawing out the dimensions for each side. As I made it however, these shifted around a bunch as I tested different layouts, so don't take them too seriously.

I used Brúni as a test here, and you can see that his more muted color palette doesn't show up well - his face has all but dissapeared. One benifit is that he is more three-dimensional than later tests I would try out, and shows off the unique 3D nature of an analog illusion.

Debrief:

What did I learn from this test? Well, first off, that I was overcomplicating things. I had based my design off of the earlier Dircks version of the illusion, which uses an additional mirror to reflect the picture twice. At such a small scale I didn't think much of the extra space, but soon realized there was another issue with this design: scaling. With an additional mirror and more space to travel, the effect looks much smaller through the reflection than the original object. If course, the image still has to travel the distance from the object, through both reflections, and into your eyes. I had thought that this layout would improve sightlines, but it didn't have much effect.

I also learned that I need to have more control over sightlines. I had originally wanted to avoid narrowing the field of view to a peephole since it would force only one audience member at a time to have the experience. But after this test, I believe something like that will be necessary. Sightlines will need to be carefully controlled.

The last issue that I noticed is with the glass I'm using. The edges are beveled, which ruin some of the illusion since you can clearly see the edges and it doesn't "dissapear" as well as it could. There is also a slight blur. I believe it's an issue of the small scale, but the double layer of glass provides a slight double image that blurrs the object. On a larger scale this 3mm shift likley won't be as noticeable.

Lastly, during building I realized that it isn't as simple as having a square mirror and basing all my measurements off of squares. Since when turned at a 45 degree angle, the two sides of the space a mirror takes up get shorter (pythangorean theorem and all that). It means that a rectangular mirror would be needed to make a fully square playing space, and that nect time I'm drawing out blueprints I should remember to use the width of the mirror while it's at an angle instead.

Second Test (2024/5)

This second test model was created in October of 2024, when my interest in the illusion suddenly came crashing back. I decided this time to base my design off of Pepper's simpler, more sucessful edits to the concept.

With my previous experiment learned from, I also wanted to test out the biggest hurdle towards me making this in real life : size. I simply don't have much space, and if I do want to take it to events and such, it's going to need to fold up in someway for transport. And be a bit lighter.

These hinges are taken from Small Stage Sets on Tour by James Hull Miller. He calls them "oriental" hinges since they're based on Korean and Japenese paper screens, only with thick fabric instead of paper. While I sure hope someone out there proposes a different name (maybe I'll have to think of one) he's right about this being an incredibly usefull two-way hinge which completely covers the gap. I used this style of hinge in my new prototype, and it would scale well to a full size theater.

The shape is based on these square wicker baskets I had as a kid - meant to slot into a huge IKEA shelf I believe - which were made of two parts: a foldable set of four walls joined in a loop, and a single flat bottom. These perplexed me as a toddler, I could never seem to wrap my head around how to open them back up and get them set straight again. I'm a lot bigger and smarter now so hopefully I can figure them out well enough to make my own. In this concept, the four walls of the theater are joined using the above hinge style, and are able to lie nearly flat when there is nothing between them. A small shelf along the bottom sticks out, and catches a solid floor that slots in perfectly, keeping the thing square and stable. More shelves along the top do the same.

In my sketches above, you can also see me considering what it would look like to add legs to this design. It would mean fewer pieces, but doubles the height so it may be better for transport if the legs were fully removeable.

Here is a sideways view of the four main walls - and you can see how well they flatten. Aside - since I'm working on a much smaller scale - the weight of the carboard isn't enough to overcome the tension of the fabric, and so it defaults to the upright position and springs up when not being held down. With heavier materials of course, this would be reversed.

Up

Flattned!

With the front view, you can see the viewing hole for the audience. In order to control sightlines, this needed a tight frame to limit the angles the viewer can see. I decided to embelish this a bit, which in the final version could hide the fact that it's a required guard.

My top and bottom pieces on their own. Each piece has a diagonal slot which the pane of glass slides securely into. On a final piece, this would be better hidden with a proportionally shallower slot and a false floor, which would allow the glass to dissapear directly into the floor. A diagonal pattern would even further hide the seam.

The bottom has a large hole cut for the playing space - we're talking puppetry here so the simplest form is just to provide a large hole for bottom-up puppet styles. Though adjustments might allow for marionettes in the future.

The top has a small hole for the light. Managing the ammount and direction of light is incredibly important for this illusion, and this is just a stand-in for me to stick a reading lamp through. On a larger scale, there may not even be a hole here, and the lights could be contained entirely within the box itself. Or, possibly, a small hole would allow a set of lights to be lowered in, which would make for easier removal and repair.

Who knows, the first full-size test might just be a bunch of clip-on worklights plugged into an external dimmer.

Assembling the frame is simple - just align the bottom while holding the walls in the right position, and let it slide down into place. Next add the glass pane, and then set the top in place.

After pleying around with this version, I realized I should test out what it would mean to control the sightlines even more. Making the front frome smaller and having the viewer get even closer would allow them to look around the corner, and ruin the illusion, so the most important step was to set the audience a good distance away. Next, I covered up the majority of the frame except for a small hole.

While this was a great proof of concept for the importance of sightline control, it's clearly an awkward shape. Difficult to translate into measurements and even more difficult to create in a way that can be folded. I discuss this more in the debrief at the bottom of the page.

With the theater itself finished, it was time for a test!

Unlike many other versions of the Pepper's Ghost, my initial concept doesn't have an equally-sized room on either side of the reflection. While the playing space where the ghost is performed is a three-dimensional corridor, the corresponding space seen through the glass is a flat wall. I did this to limit the footprint of the theater. When working on a large scale it's impressive to see the ways that objects in the ghost corridor can perfectly line up in space with objects seen directly through the glass, because they will always line up perfectly if the illusion is properly created. That's how the haunted mansion ghosts are able to interact so well with the furniture around them. But on a small scale like this, I plan to think of the puppet theater as an old-fashioned Punch n Judy style stage. Furniture is rarely used since it doesn't have anywhere to be set on. The illusions of superimposing the ghost on top of another object is an additional affect, which I don't consider necessary.

But I still want it to look like the room allows for more space, in case the ghost may look squashed. So what's my plan? Forced perspective!

It's difficult to tell, but this page is lying perfectly flat!

The set easily slides in behind the glass

I drew up a drawing room style set, a little spooky and fun. The many picture frames and wainscotting sell the forced perspective. It slots easily into the back of the theater, which is a feature I'd want to keep in full-scale testing. I don't think backdrops are necessary for every performance, but a busy backdrop seen through the transparent figure emphasizes the ghost effect.

My test ghost

Almost everything comes apart to lie flat

Time for testing! With everything in place, I used a bendable reading light to illuminate the inside of the playing space, and turned off the lights for the best experience.

And here are some of those tests:

As cute as this little ghost is, his lack of a third dimension makes it difficult to highlight the really cool 3D capabilities of the illusion. So I decided on another test with a much more engaging subject: her name is Dehlia Doornail and she has her own page too.

Debrief:

My tests with making the frame foldable and lighter went well, but I'll still need to consider the cost and weight when using heavier materials. While testing I used solid pieces of cardboard, but scaling directly up to solid pieces of plywood might be too dramatic. A much lighter design would be to make each side as a frame of 2x4s, and either use fabric or a much thinner plywood to fill in the frames. I will have to use fabric on the inside regardless, since it's much better at absorbing light (some secondhand black velour curtains perhapse? I only need small chunks, so holes wouldn't be a problem). I may be able to simplify things then by making the main walls simply frames for the fabric curtains or stretched fabric flats.

On a similar note, my viewing port is an awkward shape to fold down. My prototype was cut without measurement to be an angled tunnel made out of trapazoid shapes, which wouldn't be incredibly difficult to make, but takes more math than I really want to do on this project. I'd like to figure out other methods and shapes that provide the two most important things : distance and a limited viewing-hole. While making the construction much simpler and easier to collapse down. It could possibly just be a pyramid- ending at a cut-off point- with four fabric stretched pannels. It's also important to remeber the ouside looks of the thing.

Lighting will be a big focus of later experiemnts. I only discovered a couple uses that incporporated black lights and black light paints - I think those would be a good path to try.

With Dehlia, I discovered a lot of lovely subtleties that weren't in my original paper ghost test. Most interestingly to me, her textured and complex form creates many shadows, and those spaces where the light can't reach end up not reflecting and showing us the backdrop behind the glass. The way these shadows naturally shift gives a lot more opportunity to show off the real three dimensions, and the transparent effect. Going with the classic "haunted mansion" look for the face worked well, and I'd like to try out more vivid colors.

Definitely an improvement over my first experiment - which is awesome! I am slowly getting somehwere.

Things I haven't yet fixed: the doubling. I still believe the only way this will be sorted out is by working at a larger scale. I also have a tendency to move too far forward in the space and the puppet becomes visible to the audience, but my puppets may just be too big for the theater. One possibility still on the table is a monitor. It's very difficult to get an idea of what the audience sees.

Third Test (2025)

Through my research, a couple more thoughts came up. These are leading the way for a third prototype for me to build in the future.

In previous attempts I had thought that I didn't have a need for an equally sized playing space in the viewer's direct line of sight. A "non-ghost" space. In my next prototype, I want to experiment with adding this, specifically with the addition of a second puppeteering hole. This would allow me to contrast the ghost character with a solid one, and experiment with the ghost holding and moving solid objects. I may not keep it, but I want to try. However, the new shape might create some issues with the previous easily foldable design, hopefully the two-way fabric hinges will help with that.

I'm also playing around with making the playing space rectangular, not square. On a larger scale, I'm not sure if the puppets will need the full dimension and ability to move upstage and downstage to the same extent that they move side to side. Though this may not be clearly demonstrated with a small scale model.

I've seen references to glass that has a reflective surface on one side, which may help with the slight double vision? I can't solve this with this prototype right now, it's just something I want to do further resaerch into. It may also be beyond my reach regardless, I don't want to buy specialized glass and was planning on using glass from a large picture frame or window.

I would really like to build a model which is a little more sturdy than a prototype, and properly to scale. Something with balsa wood and curtains and the proper minature lighting setup. This would take more tools than I currently have, but after 2-3 cardboard mockups it may be time to make something that can actually be transported and demonstrated.

My new concept sketch

I also read this really interesting research paper by Tai Cossich exploring the Girl-to-Gorilla illusion and using it as a metaphor and visual model for concepts of Amerindian Perspectivism. The idea of using the structure not as an illusion intended to dissapear, but forming a metaphor around it based on the way the light shifts the perspective of the viewer, is intriguing to me. In the future once I'm more familiar with the physical setup itself, I would love to see how it could be used in a more overt way to experiment with metaphor and more complex themes. Until then, I'm satisfied with exploring the physical construction. But I'm going to be thinking about it.

You know how when people say they have an idea, and they say that a muse has spoken to them? This muse has grabbed me by the shirt and refuses to let go until I finish the job. This muse is floating behind me like Ryuk waiting for me to feed it apples and do something entertaining.

Lobby

Library

Pepper's Ghost Puppet Theater

Personal Expreriment - 2023 - 2025

Research & Concept

Concept/Initial Musings:

More Reference Links:

First Test (2023)

Debrief:

Second Test (2024/5)

Debrief:

Third Test (2025)