7 Free Math Visualizations That Show How Everyday Tech Actually Works
The Fourier Transform runs every time you stream music. Convolution processes every image your neural network touches. The double pendulum is why your weather app is sometimes catastrophically wrong.
Most of us learn these concepts from textbooks. But there's a better way: interactive visualizations that let you see the math running, not just read equations about it.
ElysiaTools hosts a growing collection of free, browser-based math and physics visualizations. No signup. No plugins. Just open and explore.
Here are 7 that will permanently change how you think about the tech around you.
1. Fourier Series Approximation — Why MP3 Files Exist
A square wave looks nothing like a smooth sine wave. Yet any periodic signal — your voice, a guitar chord, an EEG reading — can be broken down into nothing but sine waves added together.
This is Fourier's insight from 1807, and it's the entire basis of MP3 compression.
The Fourier Series Approximation tool lets you watch it happen in real-time. Pick a target waveform (square, sawtooth, or triangle), then slide the harmonic count from 1 to 50. Watch the jagged approximation smooth out as you add more sine wave components.
Why it matters: JPEG, MP3, AAC, and every audio codec in your phone works by decomposing signals into frequency components, discarding the ones you can't hear, and reconstructing the rest. The Fourier Series is the reason lossy compression exists at all.
Use it to: Understand why turning up the bass on your EQ doesn't add new audio — it just changes the amplitude coefficients of existing frequency components.
2. Fourier Transform Family — The Full Picture From Theory to FFT
The Fourier Series handles periodic signals. But what about a single pulse? Or a continuous stream of live microphone data?
The Fourier Transform Family visualization maps the entire landscape: Continuous FT → Fourier Series → DTFT → DFT → FFT. Each transform exists because the previous one was computationally impractical at scale.
The critical insight: The DFT formula works — but naive implementation is O(N²). The FFT (Cooley-Tukey, 1965) reduces this to O(N log N) by exploiting symmetry in the "twiddle factors" (roots of unity on the unit circle). This is why real-time audio processing became possible.
The visualization includes a butterfly diagram that shows exactly how the FFT recurses through stages, plus a live audio spectrum analyzer that uses your microphone or an oscillator to visualize frequency content in real-time.
Why it matters: 4G, 5G, WiFi, and every modern wireless protocol uses OFDM (Orthogonal Frequency Division Multiplexing), which is FFT applied to hundreds of tightly packed subcarriers. Without the FFT, none of these work.
Use it to: See why a 1024-point FFT runs ~100× faster than a naive DFT, and how the Cooley-Tukey butterfly diagram achieves it.
3. Convolution — The Operation Behind Every Neural Network
Every convolutional layer in ResNet, VGG, or YOLO does the same thing: it slides a small matrix (the kernel) across an image, multiplies overlapping values, and sums them up. That's convolution.
The Convolution visualization lets you pick an input signal (rectangular pulse, Gaussian, sinc) and a kernel (moving average, Gaussian blur, Sobel edge detector), then watch the flip-slide-multiply-sum process animate frame by frame.
The convolution theorem — shown interactively — states that convolving in the time domain equals multiplying in the frequency domain (and vice versa). This is why modern deep learning frameworks do most convolution operations as FFT multiplications for large kernels.
Why it matters: GitHub Copilot, Claude, Midjourney — all of them run on transformers, which internally rely on self-attention. Attention is a form of convolution. The layer you're reading this article through is, mathematically, a series of giant convolutions.
Use it to: Understand why a 3×3 Sobel kernel detects edges (it's computing a numerical gradient), and why larger kernels in neural networks capture more complex patterns.
4. Double Pendulum — Why Weather Forecasts Fail
A double pendulum is simple to describe: two weights on two rods, free to swing. That's it. But at high energy, it never repeats. The same starting position produces something that looks genuinely random — even though the equations are completely deterministic.
The Double Pendulum tool has a built-in butterfly effect demo: it launches three pendulums with initial angles differing by only 0.001 radians. Within seconds, the three trajectories have completely diverged.
This is sensitive dependence on initial conditions — the mathematical core of chaos theory.
Why it matters: Lorenz discovered this in 1963 by accident. He was rerunning a weather simulation and rounded 0.506127 to 0.506. The result was completely different. He had stumbled onto deterministic chaos — and changed meteorology, physics, and our understanding of predictability forever.
Use it to: Understand why a 10-day weather forecast is inherently unreliable. The atmosphere is a multi-million-degree-of-freedom double pendulum. Tiny measurement errors amplify exponentially.
5. Lorenz Attractor — The Beautiful Shape of Unpredictability
The Lorenz system is three coupled differential equations that Lorenz derived from a simplified model of atmospheric convection. They produce a trajectory that spirals around two interlocking lobes — never intersecting, never escaping, always staying within a bounded region.
The Lorenz Attractor visualization renders this in 3D. You can rotate the view, adjust σ (Prandtl number), ρ (Rayleigh number), and β (geometric factor), and run the butterfly effect comparison to see two trajectories diverge in real-time.
The strange part: The Hausdorff dimension of the Lorenz attractor is approximately 2.06. It's not quite a surface, not quite a line — it's a fractal object with fractional dimension. This is what "strange attractor" means.
Why it matters: Strange attractors appear in laser physics, chemical reactions, water wheels, and superconducting microwave resonators. The Lorenz attractor is a template for understanding how order emerges from chaos in real physical systems.
Use it to: Appreciate why chaos isn't randomness — it's bounded, deterministic, geometry with no periodic orbits. The shape is beautiful because it's unpredictable.
6. Game of Life — Computation Without a Processor
Conway's Game of Life runs on four rules: a live cell dies if it has fewer than 2 or more than 3 live neighbors; a dead cell becomes alive if it has exactly 3 neighbors. That's it.
But this minimal rule set is Turing-complete. You can build a Turing machine inside Game of Life. Glider guns, oscillators, and logic gates have all been constructed.
The Game of Life visualization lets you set initial conditions, adjust the rule set, and run cellular automata at different scales. You can explore Rule 30 and Rule 110 — the latter is provably Turing-complete.
Why it matters: Cellular automata model emergent behavior in biology (seashell patterns, animal coats, fungal growth), traffic flow, and - more speculatively - the fundamental structure of physics at the Planck scale.
Use it to: Understand why complex, organized behavior can arise from simple, local rules with no central controller. This is the same mathematical structure behind swarm robotics and ant colony optimization.
7. Option Greeks Visualizer — Calculus That Moves Markets
Delta, Gamma, Theta, Vega — the "Greeks" measure how an option's price changes with respect to the underlying asset price, volatility, time, and interest rates. They're partial derivatives of the Black-Scholes pricing formula.
The Option Greeks Visualizer renders these surfaces in 3D, letting you rotate and explore how each Greek varies across different strike prices and time-to-expiration.
Why it matters: Every options desk at every hedge fund runs on these surfaces. Gamma scalping — the practice of delta-hedging an options position to harvest theta — is essentially a calculus arbitrage. The math that won Scholes and Merton the Nobel also helped blow up Long-Term Capital Management in 1998.
Use it to: See why a deep-in-the-money put option has gamma that spikes near expiration — small moves in the underlying cause massive swings in the option's delta, which forces delta-hedging that compounds losses.
The Unseen Math Layer
Every time you stream a song, detect an object in a photo, check the weather, or trade a stock option, you're interacting with mathematics that was discovered decades or centuries ago — Fourier (1807), Lorenz (1963), Black-Scholes (1973).
The visualizations above let you build genuine intuition for these concepts by watching them run. That's different from memorizing equations. It's the difference between reading about a fire and standing next to one.
Explore the full collection at elysiatools.com — free, no signup required.
The open question: what mathematical concept is hiding in the next technology you use, waiting for someone to visualize it?
Top comments (0)