Thaumatrope Persistence of Vision: Build One Today

Draw a bird on one side of a card and an empty cage on the other. Spin the card fast and the bird is in the cage. That is a thaumatrope, an 1820s parlour toy, and it is the best five-minute demonstration of thaumatrope persistence of vision you can make from a cereal box. There is one detail nearly every tutorial online gets wrong, and it is the reason most home attempts produce an upside-down bird above a cage instead of a bird inside it: the image on the back has to be drawn the wrong way up. Get that right and it works first time. There is also a popular claim attached to this toy that is not true, and taking it apart is the most interesting part of the afternoon.

Key Takeaways

• You build a card disc with two matching images on opposite faces and spin it on twisted strings until the two fuse into one picture.
• The back image must be drawn upside down relative to the front, because the disc flips end over end. This single detail is why most attempts fail.
• You need roughly five turns per second, which gives ten image presentations per second. Slower than that and you see two pictures flickering.
• Best for ages 5 to 11. A six-year-old can draw and spin it; the fusion threshold measurement suits 10 and up.
• The claim that “persistence of vision explains how films work” is wrong, and the thaumatrope is the one device where the idea genuinely does apply.

What You’ll Need

Item	How much	Where to get it	If you don’t have it
Stiff white card	A piece 10 × 10 cm	Cereal box, index card, card stock	Two layers of printer paper glued together, which warps more
Cotton string or thin twine	2 lengths of 30 cm	n/a	Dental floss, or embroidery thread doubled
Hole punch or a pencil point	1	n/a	Push a pencil through onto a rubber eraser
Pencil and felt-tip pens	n/a	n/a	Dark lines work far better than pale ones
Scissors	1	n/a	Adult cuts for under-6s
Bright lamp and a dark room	n/a	n/a	Needed for the luminance experiment later
Phone with slow-motion video	1	n/a	Used to count the rotation rate

Time: prep 15 min, experiment 15 min, cleanup 2 min Cost: nothing Mess level: low

Do the Experiment

Step 1 - Cut a disc and punch two holes

Cut a circle about 7 cm across from the card. Punch a small hole on each side, level with the centre and about 3 mm in from the edge. The two holes and the centre must lie on one straight line, because that line is the axis the disc will spin about.

Step 2 - Draw the front, then trace the outline through

Draw the bird in the middle of one face. Then hold the card up to a window and pencil a faint outline of where the bird sits, on the back. This guarantees the two images will line up when they fuse, which is the difference between a bird in a cage and a bird beside a cage.

Step 3 - Draw the back image upside down

This is the step that matters. Turn the card over by flipping it top to bottom, about the line through your two holes, and draw the cage upside down relative to the front image. If you flip the card left to right instead, you must mirror the back image horizontally instead. Either way, the back drawing is not simply the same way up as the front.

Step 4 - Thread the strings and predict

Thread a 30 cm string through each hole and knot it into a loop. Hold one loop in each hand. Before spinning, ask: “What will you see when it goes fast? Two pictures taking turns, one picture, or a blur?” Record the answer.

Step 5 - Wind up and spin

Twist both strings in the same direction thirty or forty times, then pull the loops apart steadily so the disc spins up. Keep the disc at arm’s length and roughly at eye level.

What you should see: At slow speeds, two pictures alternating, visibly flickering. As the spin passes about five turns a second the flicker disappears and you see one steady composite image: the bird inside the cage. The transition is surprisingly sharp. If your bird appears upside down above the cage, go back to Step 3.

Step 6 - Find the fusion threshold and test the light level

Record a slow-motion video of the disc as it spins down, and count rotations per second at the moment the image stops fusing. Then repeat the whole thing in a dim room.

What you should see: In a dim room the images fuse at a lower spin rate than in bright light. This is a real and well-documented property of the visual system, and it is the most interesting measurement in the activity.

What’s Actually Going On: Thaumatrope Persistence of Vision

Your visual system does not take instantaneous snapshots. A photoreceptor’s response to a flash of light builds and decays over tens of milliseconds, and the neural stages after it integrate over a similar window. The practical consequence is that two images arriving within roughly 50 to 100 milliseconds of each other are summed into a single percept rather than seen as separate events. Spin the disc at five turns a second and each face appears every 100 ms, which puts the two images inside that integration window. Bird and cage add together.

The threshold depends on brightness, which is why Step 6 works. The critical flicker fusion frequency, the rate at which a flickering light stops looking like flicker and starts looking steady, rises with luminance: it sits somewhere around 20 Hz for dim stimuli and can exceed 50 Hz for bright ones. D. H. Kelly’s measurements of temporal contrast sensitivity, published in the Journal of the Optical Society of America in 1961, mapped out that dependence in detail. In a dark room your visual system integrates over a longer effective window, so a slower spin suffices.

Now the correction, and it matters because this toy is the standard exhibit for a claim that is not supported. You will read everywhere that “persistence of vision is how films and cartoons work.” Persistence of vision is real: it explains why you do not see the dark gap between film frames, and it explains the thaumatrope exactly. What it does not explain is the perception of smooth motion. Summing successive images would produce a smeared superposition, which is what your thaumatrope does, not the appearance of a bird flying. Motion perception depends on dedicated motion-detecting mechanisms, which respond to the displacement of a pattern over time. Oliver Braddick’s 1980 paper in the Philosophical Transactions of the Royal Society B distinguished a short-range process, which handles small displacements between closely spaced frames, from a longer-range one, and it is the short-range process that makes animation read as movement rather than as a sequence of stills. Paul St. George argued in Animation Studies Journal in 2009 that the persistence-of-vision explanation for cinema should be retired in favour of an account based on chronophotography and apparent motion.

So the honest summary is a two-part answer. Your thaumatrope works by visual persistence, and it is one of the few devices where that explanation is right. A zoetrope, a flip-book and a film work by apparent motion, and persistence only explains why they do not flicker. A child who can state that distinction knows something most adults do not.

One more real detail. Eric Castet’s 1994 work in Vision Research on how visible persistence changes with the interval between frames in apparent-motion displays shows that the two phenomena interact rather than being wholly separate, so the story is not “persistence is a myth” but “persistence is the wrong mechanism for the specific job of motion.”

How to Explain It to Your Kid

Ages 5–8: Your eyes keep the picture for a moment after it’s gone

Have them stare at a bright window for five seconds, then close their eyes and describe the glowing shape that lingers. That lingering is the same effect, just slower. Then tell them the spinning card shows each picture so fast that the first one has not finished fading when the second arrives, so both are there at once. Let them design their own pairs: a fish and a bowl, a face and a hat, a rocket and a planet.

Ages 9–12: About a tenth of a second

Give them the number. Each image needs to arrive within about a tenth of a second of the last one, which means five spins a second. Then run the bright-versus-dim comparison and let them discover that the threshold moves. Ask what that implies about watching a screen in a dark room. The same integration time explains why a phone camera’s video and your eye disagree about a spinning fan, which connects to how a camera sensor samples light over time.

Ages 13+: Separate persistence from apparent motion

Build a flip-book alongside the thaumatrope. The thaumatrope superimposes; the flip-book moves. Ask them to explain why one mechanism cannot account for both, then introduce short-range apparent motion. The payoff question: if persistence alone explained film, what would a film actually look like? The answer, an ever-thickening smear of all previous frames, is a strong argument from a prediction that fails.

The question to ask: “If I made a thaumatrope with three pictures on three faces of a triangular prism instead of two, would it still fuse, and would the picture be brighter or dimmer than with two?”

Why Teaching This Early Pays Off

This is one of the earliest opportunities a child gets to understand that perception is constructed rather than received. The bird is not in the cage. The card never shows both at once. The composite exists only in the visual system, as a consequence of how long neurons take to respond. That is a genuinely different kind of fact from “water boils at 100 °C”, and meeting it at age eight makes later topics in biology and psychology much less strange.

It also builds an unusually transferable habit: distinguishing a phenomenon from its explanation. Persistence of vision is a real measured phenomenon. “Persistence of vision explains cinema” is a separate claim, widely repeated, and wrong about the part that matters. Taking those apart is exactly what Chi and Wylie’s ICAP framework calls constructive engagement, where the learner generates something beyond what was presented, and their 2014 analysis places that above simply watching a demonstration.

Keep the claim modest. One thaumatrope does not teach vision science. What it does is give a child a physical, repeatable object attached to a correct explanation, and a reason to be sceptical the next time they hear a tidy story about how animation works.

Before You Start: What Parents Should Know

Best age	5 to 11. Younger children need an adult to cut and thread; the measurement extension suits 10 and up
Supervision	Light. Scissors for under-6s
Real hazards	Loops of string are a strangulation hazard for children under three; keep the loops short, supervise, and put the toy away afterwards. The disc spins near the face, so hold it at arm’s length and keep it away from eyes
Don’t	Don’t substitute wire or stiff plastic for the card. A stiff spinning disc at arm’s length is a genuine eye risk, and card is safer and spins better

When It Doesn’t Work

• The two images appear separately or offset. The back image is not upside down. Flip the card about the line through the two holes and check: whichever way you flip it, the back drawing must be oriented so that it comes out the right way up during the spin.
• You see flicker instead of one picture. The spin is too slow. Wind the strings more times, pull harder and more steadily, and try again in a slightly dimmer room where the fusion threshold is lower.
• The disc wobbles and the picture smears sideways. The two holes are not level with each other or not equidistant from the centre. Re-punch them carefully on a line through the middle.
• The composite is faint and washed out. Your drawings are too pale, or the card is translucent and light shines through. Use heavy felt-tip lines and card you cannot see through.

Take It Further

• Measure the fusion threshold properly. Film the disc spinning down in bright light and in dim light, count the rotation rate from the video at the instant fusion is lost in each case, and repeat five times. You will get two numbers with real error bars, and the difference is the luminance dependence of flicker fusion.
• Build a zoetrope and compare mechanisms. A strip of twelve frames inside a slotted spinning drum produces apparent motion rather than superposition. Running both side by side is the clearest way to show that these are two different phenomena, and it pairs with the straight-line optics in shadow puppets as physics.
• Try a colour-mixing disc. Replace the two pictures with a red half-disc on one face and a green one on the other and spin it. The colours add rather than mixing like paint, for the reasons set out in why paint mixing and light mixing disagree.

What to Watch For Over the Next 3 Months

• Week 4: They notice afterimages on their own, staring at a lamp or a bright screen and describing what lingers.
• Month 2: They start asking about frame rates, in games or video, and whether a higher number is actually visible. That is the same threshold question with different units.
• Month 3: If they object when someone says “films work because your eyes keep the picture”, the distinction stuck. The follow-up worth encouraging is why a spinning wheel sometimes appears to turn backwards on video, which is aliasing rather than persistence.

Frequently Asked Questions

Why is my bird upside down?

Because the back image was drawn the same way up as the front. A disc on twisted strings flips end over end about the line through its two holes, so the rear drawing appears inverted during the spin. Draw it upside down and it will come out right.

How fast does it have to spin?

About five rotations per second, which presents each face ten times a second. Below about three rotations a second you see flicker. The exact threshold depends on how bright the room is, which is itself worth measuring.

Is persistence of vision a myth?

No, but one famous application of it is. Visible persistence is a measured property of the visual system and it is the correct explanation for a thaumatrope. The claim that it explains the smooth motion in films is wrong; that depends on apparent-motion mechanisms instead.

Can I use a pencil instead of strings?

Yes. Tape the card to a pencil and roll the pencil between your palms. It is easier for small hands but usually spins more slowly, so fusion is harder to reach. With the pencil version the disc rotates about a vertical axis, so the back image must be mirrored left to right rather than turned upside down.

Does it work on a phone screen?

Not as a thaumatrope. An animation of two alternating images on a screen is apparent motion plus flicker fusion, and the physical object is the point: the card genuinely only ever shows one image at a time, which is what makes the result convincing.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

---

Sources

1. Braddick, O. J. (1980). “Low-level and high-level processes in apparent motion.” Philosophical Transactions of the Royal Society of London. B, Biological Sciences, 290(1038), 137–151. https://doi.org/10.1098/rstb.1980.0087
2. St. George, P. (2009). “Using chronophotography to replace Persistence of Vision as a theory for explaining how animation and cinema produce the illusion of continuous motion.” Animation Studies Journal, 4. https://doi.org/10.64864/a9b67a5c-1fa4-4a21-ba1c-1aab5e6e4dfc
3. Kelly, D. H. (1961). “Visual Responses to Time-Dependent Stimuli. I. Amplitude Sensitivity Measurements.” Journal of the Optical Society of America, 51(4), 422. https://doi.org/10.1364/JOSA.51.000422
4. Castet, E. (1994). “Effect of the ISI on the visible persistence of a stimulus in apparent motion.” Vision Research, 34(16), 2103–2114. https://doi.org/10.1016/0042-6989(94)90320-4
5. Chi, M. T. H., & Wylie, R. (2014). “The ICAP Framework: Linking Cognitive Engagement to Active Learning Outcomes.” Educational Psychologist, 49(4), 219–243. https://doi.org/10.1080/00461520.2014.965823
6. National Research Council. (2012). A Framework for K-12 Science Education. National Academies Press. https://doi.org/10.17226/13165