Beta

How (and why) to take a logarithm of an image

Below is a short summary and detailed review of this video written by FutureFactual:

Escher and the Magic of Conformal Maps: Recreating Print Gallery with Complex Analysis

Overview

This video pulls MC Escher's Print Gallery into the language of complex analysis, showing how a self-similar Droste image and a conformal grid can be used to create a self-contained loop. The presenter guides you from an intuitive, artistic description of Escher's piece in the art room to the mathematical machinery of logarithms, exponentials, and conformal maps, and finally to a constructive function that turns a line of imagery into a closed loop as you walk the grid.

Prelude: Art, Math, and a Puzzle

The long journey begins by reframing MC Escher's Print Gallery as more than a clever optical trick. The piece depicts a boat in a harbor with the world subtly warping as one’s gaze traverses to the right and then downward into a hallway of artwork, where a picture of the harbor reappears. Escher himself described ideas like this as some of the most peculiar things he did, and the piece has long fascinated mathematicians for the wealth of geometric and topological ideas it contains. In this video, the host opts to visually unpack the analysis of two mathematicians, De Smit and Lenstra, who studied the “mind-bending, self-contained loop” Escher built into Print Gallery. The aim is not only to explain what is going on but to provide an intuitive path to rediscovering the logic himself.

Three Intuitive Steps to Escher’s Loop

The host begins with a purely intuitive description of Escher’s construction, which can be broken down into three conceptual steps. Step one mirrors a Droste effect, a self-similar image where the same idea repeats at smaller scales. Escher’s version involves a harbor, a town, a gallery, and the same man in a recursive nesting: a picture inside a picture inside a picture inside a picture, ad infinitum. This is called a Drosta or Droste effect, with a twist in scale: the self-similar copy Escher uses is 256 times smaller than the original in the canonical Print Gallery. The self-similarity is so deep that the zoom occurs implicitly as the viewer’s gaze travels along a circle, rather than through a sequence of explicit zooms.

Step two introduces a warped grid, or mesh warp, that encodes this zooming across four corners of a square. The original 256-fold zoom gets distributed so that moving from corner to corner effectively scales by a factor of two at each transition. In the film’s illustration, you see a grid where a square near the lower left expands into higher-order squares along the top and left edges, creating a consistent four-corner warp. The grid ensures a smooth transition from one corner to the next, so that the viewer experiences a seamless loop around the circle as they wander the image’s path.

Step three, the practical step, describes how to fill the gaps between the four warped corners. The artist would copy contents of each tiny square to the corresponding warped square, since at small scales a region is distortion-free. The process is iterative, with the local zones maintaining their square shape, which makes the entire piece coherent when the process runs around the full loop and returns to the starting point. The mesh warp is a technique used in graphic design beyond Escher, known as mesh warping, and the Print Gallery’s grid is notable for its approximate perfect right-angle intersections, particularly at small scales. This local conformality is the seed of the mathematical power the presenter then unfolds in the math department.

Conformal Maps and the Two-Dimensional Heart of the Idea

At the mathematical core, the idea is to model the transformation as a function from two-dimensional space to itself that preserves shapes locally. In mathematics, a function that does this in the complex plane is called a conformal map, a concept especially central to complex analysis. The video explains that a conformal map, by preserving squares at small scales, makes objects recognizable even when the global image is wildly warped. In this sense, Escher’s grid is not only aesthetically striking but also an almost perfect example of a conformal transformation in action. The video emphasizes that while many mesh warps produce parallelogram-like regions, Escher’s grid has nearly perfect squares in local tiles, a property that helps the final composition feel natural and consistent to the eye.

This notion of conformality in the 2D plane is a gateway to the broader mathematics of complex functions, notably functions with complex inputs and outputs. The presenter hints that the right framework to understand this is complex analysis, including a set of powerful, classic functions that preserve local geometry in a manner that echoes the art’s intuitive preservation of shape under distortion.

Complex Analysis: Numbers, Grids, and Derivatives

The host then pivots to a compact primer on complex numbers and complex functions. Complex numbers live in a two-dimensional plane with the imaginary unit i defined by i^2 = -1. A general complex number z can be written as x + iy, and a function F(z) maps the complex plane to itself. The simple operation of multiplication by a constant C is a geometric transformation: it scales and rotates the entire plane about the origin. Such transformations preserve shapes exactly because they are linear maps in the complex plane. The video gives a concrete example with F(z) = 2z or F(z) = iz, illustrating pure scaling and rotation.

More interesting are nontrivial complex functions, like z → z^2. Under such a map, grid lines bend and warp, and shapes are distorted. Crucially, at small enough scales, tiny squares remain approximately square, which is the heart of conformality. This local preservation of shape is what makes conformal maps so powerful for connecting geometry with the analysis of functions. The video emphasizes that not only polynomials but a broad family of complex functions have this conformal property in the limit as you zoom into a point, which is why complex analysis provides a natural language for the puzzle Escher created.

Logarithms and Exponentials: Visualizing the Core Tools

The narrative then introduces two indispensable toolbox functions for complex analysis: the exponential e^z and the logarithm. The exponential maps vertical lines in the z-plane to circles in the w-plane, with a fixed rate: increasing the imaginary part by 2π completes one rotation. This is a vivid way to visualize how exponential functions wrap the plane into circles, a central feature for turning a line into a loop in the final image. Conversely, the natural logarithm unwraps circles back into vertical lines, revealing how the multi-valued nature of log in the complex plane corresponds to the repetition of copies in the Droste-like pattern.

The video emphasizes that while the exponential is many-to-one in the complex plane (different imaginary parts can land on the same point due to 2π periodicity), the logarithm can be treated as multi-valued or with a chosen branch cut, depending on the application. This interplay between log and exp is the backbone of constructing a function that turns a diagonal in the log-space into a vertical segment of a 2π height, which, after exponentiation, yields a closed loop in the image space.

The Doubly Periodic Tiling and Elliptic Functions

A key pivot occurs when the video connects the construction to a doubly periodic pattern, one that repeats in two independent directions. This is where elliptic functions enter. Elliptic functions are complex functions that are periodic in two directions and thus are natural candidates to encode the kind of repeating tilings used in the Drosta-like constructions. The mathematicians De Smitt and Lenstra are cited as having linked this idea to deeper number theoretic structures, providing a bridge from a visually appealing puzzle to modern mathematical research. The presenter stresses that this is not mere decoration; elliptic functions are a powerful class of conformal maps with rich structure that resonate with the self-similar, tiled nature Escher exploited.

Putting It All Together: A Constructive Recipe

After laying out the components, the host introduces the outline of a single, compact formula that encapsulates Escher’s transformation in this modern language. The recipe has four stages: take a logarithm to produce a doubly periodic tiling, rotate and scale the tiling in just the right way to align the distant copies, and then apply the exponential to unwarp back to image space with a twist. The twist is to choose the rotation and scaling carefully so that a line in the log-space, connecting the big and the small copy, ends up vertical with height 2π. The fixed point around which this rotation is performed is denoted Z0, which ensures a chosen point remains stationary as the transformation is applied. The resulting final image is a Drasta-like tiling that spirals down toward infinity or fills the plane, depending on how far one extends the domain, but always preserves the essential local geometry due to the conformal nature of the map.

A Clean Reduction: A Simple Power Map

In a note of elegance, the presenter observes that the composite log-rotate-exponentiate process can be simplified to a simple power transform z → z^k, possibly with a shift. While this is a flattering simplification, the video cautions that the simplification should not obscure the underlying geometry; the interplay of log and exp, and the rotational alignment around a fixed point, is what gives Escher’s Print Gallery its particular character. The exploration also demonstrates what happens if one uses a horizontal line instead of the diagonal path in the log space, concluding that such a choice does not yield the same loop and misses Escher’s intended structure. The final perspective underscores a fundamental duality: in log space you rotate and scale to align images, but in image space you gain a looping, self-contained symmetry when you exponentiate.

From Art to Math: A Shared Sense of Fit

The closing arc emphasizes the kinship between Escher’s puzzle-solving instinct and mathematical problem solving. Escher’s own aesthetic often sought images that fit together in precise and aesthetically satisfying ways, which mirrors the mathematician’s search for elegant, conformal mappings and the right complex functions. The video suggests that these are not just analogies: the deep mathematics behind Escher’s puzzles, including conformal maps and elliptic functions, reveal a structure that is both visually compelling and mathematically rich. This convergence hints at a universality in human thought, where puzzle-solvers in art and math alike are drawn to configurations that show unity between local structure and global form. The host ends with a personal reflection on how Escher’s works and mathematical exploration share a common impulse—to discover a puzzle where a solution seems almost inevitable once the right lens is applied.

Implications and Beyond

The discussion hints at ongoing research where elliptic functions and complex analysis play central roles, touching on how such ideas connect to modern number theory, and how they surface in artistic representations of infinity and finitude. The broader message is that mathematical ideas can illuminate not only abstract theories but also the intuitive, visual puzzles that artists and scientists alike find captivating. The Print Gallery thus stands as a bridge between two modes of inquiry, each sharpening the other: the aesthetics of a self-contained loop in Escher’s composition and the rigorous, revealing structure of conformal maps in complex analysis.

To find out more about the video and 3Blue1Brown go to: How (and why) to take a logarithm of an image.