Imagine you have a complex drawing—a butterfly, a heart, or even your own signature. It looks like a messy, complicated shape. But what if I told you that this drawing is actually just a combination of simple, spinning circles?

This is the magic of the Fourier Transform. Discovered by the French mathematician Joseph Fourier in the early 19th century, this mathematical tool allows us to break down any complex signal or shape into a set of much simpler components. In the case of drawings, those components are rotating circles, often called epicycles.

The Mathematics of Shapes

To understand how this works, we need to think of a drawing not just as lines on a page, but as a path traveling through time. A closed curve in a 2D plane can be represented mathematically as a complex-valued function:

z(t) = x(t) + i y(t)

Here, t represents time as you trace the drawing from start to finish. If the shape is closed (meaning the end connects back to the start), we say it's periodic. According to Fourier's theory, any periodic function can be expressed as an infinite sum of simple sine and cosine waves. In the language of complex numbers, these waves are represented by rotating vectors (or circles).

Decomposing the Drawing

The Fourier Transform calculates exactly how large each circle needs to be, how fast it should spin, and where it should start (its initial phase). The formula looks like this:

$$ z(t) = \sum \, _{k \in \mathbb{Z}} ~ c_{k}\, e^{ i\, k\, t}, \quad c_{k} = \frac{1}{2\pi } \int_{0}^{2\pi} z(t)\, e^{-i\, k\, t} dt $$

Let's break this down into plain English:

• The term ck is a complex number that tells us the radius and the starting angle of one specific circle.
• The term eikt represents the circle spinning at a certain frequency, k.
• The equation says: if you take an infinite number of these circles, attach them tip-to-tail, and let them spin, the tip of the final circle will trace out the exact shape of your drawing!

Approximating Reality

While the math requires an infinite number of circles to be perfectly accurate, we don't actually need that many in practice. We can sort the circles by size and only use the largest ones.

The largest circles create the general shape (the "low frequencies"), while the tiny, fast-spinning circles add the sharp details and corners (the "high frequencies"). By throwing away the smallest circles, we get a slightly smoother, simplified version of the drawing. This concept of discarding information while measuring the remaining error is known as Parseval's identity.

Conclusion

The Fourier Transform isn't just for making cool animations. It's one of the most important tools in modern science and engineering. It's used in audio processing (to compress MP3 files or cancel out noise in your headphones), image compression (like JPEGs), quantum mechanics, and even discovering exoplanets!

You can see this math in action by drawing your own shapes and watching the epicycles recreate them in Experiment 003.