Coherence and Visibility

The intricate patterns formed by interfering waves—from ripples on a pond to the colors in a soap bubble—are a hallmark of the physical world. Yet, not all waves produce such clear, stable patterns. The ability to do so is governed by a fundamental property known as coherence. This article addresses the core question: What is coherence, and how does it dictate what we can observe and measure through interference? It demystifies the link between the mathematical description of coherence and the practical visibility of interference fringes.

First, in the "Principles and Mechanisms" chapter, we will dissect the concepts of coherence and visibility, defining their mathematical relationship and exploring the two main flavors of coherence: temporal and spatial. We will see how these properties are profoundly linked to the characteristics of a wave's source through powerful principles like the Wiener-Khinchin and van Cittert-Zernike theorems. Following this, the "Applications and Interdisciplinary Connections" chapter will showcase how these theoretical ideas become powerful tools, enabling us to measure the spectra of light with Fourier Transform Spectroscopy, determine the size of distant stars, peer into atomic crystals, and probe the strange and beautiful rules of quantum mechanics.

Imagine you are standing by a calm lake. You throw in two pebbles, side by side. From each point, concentric circles of ripples spread outwards. Where the crest of one wave meets the crest of another, the water leaps up. Where a crest meets a trough, the water is calm. This beautiful, intricate pattern of high and low points is interference. It is the hallmark of all wave phenomena, from water and sound to light and even the matter waves of electrons.

But what if the "pebbles" were not simple, single splashes? What if each was a chaotic, sputtering fountain, shooting out waves at random? Would you still see a clear, stable interference pattern? Probably not. The pattern would flicker and shift, and if you averaged it out over time, the beautiful lacework of interference would likely wash away into a uniform mess. The ability of waves to produce a stable, high-contrast interference pattern is the very essence of ​​coherence​​. In this chapter, we will peel back the layers of this concept, revealing the simple and profound principles that govern it.

Let's get to the heart of the matter. Whenever two waves combine, the total intensity is not just the sum of the individual intensities. There is an extra piece, the interference term, that depends on the relationship between the two waves. If we denote the wave fields from two paths (say, two slits in an experiment) as $\psi_1$ and $\psi_2$, the total intensity $I$ we measure is the time-average of the squared magnitude of their sum:

$I = \langle |\psi_1 + \psi_2|^2 \rangle = \langle |\psi_1|^2 \rangle + \langle |\psi_2|^2 \rangle + 2\text{Re}\langle \psi_1^* \psi_2 \rangle$

The first two terms are just the intensities from each path alone, $I_1$ and $I_2$. They form a simple, flat background. All the interesting physics is locked inside that third term, $2\text{Re}\langle \psi_1^* \psi_2 \rangle$, the ​​interference term​​. This term tells us how the two waves "talk" to each other. To make things tidy, physicists define a normalized version of this cross-correlation called the ​​complex degree of coherence​​, $\gamma_{12}$:

$\gamma_{12} = \frac{\langle \psi_1^* \psi_2 \rangle}{\sqrt{I_1 I_2}}$

This quantity, $\gamma_{12}$, is our main character. It's a complex number that holds all the secrets. Its magnitude, $|\gamma_{12}|$, tells us how strong the correlation is, and its argument (its angle in the complex plane) tells us about the relative phase between the waves. With this, our intensity formula becomes wonderfully transparent:

$I = I_1 + I_2 + 2\sqrt{I_1 I_2} |\gamma_{12}| \cos(\Delta\phi + \arg(\gamma_{12}))$

where $\Delta\phi$ is any phase difference we control experimentally. Look at this equation! It tells us everything. The interference pattern is a cosine ripple. The argument of $\gamma_{12}$ simply shifts the position of the fringes, like sliding the whole pattern left or right. But the contrast of the fringes—how dramatic the difference is between the brightest brights and the darkest darks—is set by the magnitude $|\gamma_{12}|$.

This contrast is captured by a quantity called ​​fringe visibility​​, $V$, defined as $V = (I_{max} - I_{min}) / (I_{max} + I_{min})$. A quick calculation from our intensity formula reveals the crucial link:

$V = \frac{2\sqrt{I_1 I_2}}{I_1 + I_2} |\gamma_{12}|$

If the intensities from both paths are equal ($I_1 = I_2$), the fraction becomes 1, and the visibility is simply $V = |\gamma_{12}|$. If coherence is perfect ($|\gamma_{12}| = 1$), you get perfect visibility ($V=1$), with bright fringes twice the average intensity and dark fringes of pure black. If there is no coherence ($|\gamma_{12}| = 0$), the visibility is zero; the waves add like random noise, and the interference pattern vanishes completely. All the shades of grey in between correspond to partial coherence, $0 |\gamma_{12}| 1$.

So, our mission is clear: to understand interference, we must understand what determines the degree of coherence, $\gamma$. We'll find that coherence comes in two main flavors: temporal and spatial.

Temporal coherence asks a simple question: If you look at a wave at a certain point now, how well can you predict what it will be doing at that same point a short time $\tau$ later? It's a measure of a wave's correlation with its own past. To measure this, we use an ingenious device called a ​​Michelson interferometer​​. It splits a beam of light, sends the two copies down different paths, and then recombines them. By changing the length of one path, we introduce a time delay $\tau$ between the two copies before they interfere. The visibility of the resulting fringes as a function of this delay, $V(\tau)$, directly maps out the magnitude of the temporal coherence, $|\gamma(\tau)|$.

What determines this function $\gamma(\tau)$? A profound and powerful result, the ​​Wiener-Khinchin theorem​​, provides the answer: the temporal coherence function $\gamma(\tau)$ is the Fourier transform of the light source's power spectrum, $S(\nu)$.

Let's unpack this. The spectrum, $S(\nu)$, is like the musical score of the light—it tells us which frequencies (colors) are present and in what amounts. The coherence function, $\gamma(\tau)$, is like the resulting sound wave's texture over time. The theorem states that these two descriptions are mathematically linked through a Fourier transform. This means if you know the colors in your light source, you can predict its interference pattern, and vice versa. This is the bedrock of an entire field called ​​Fourier Transform Spectroscopy​​.

Let's explore this with a few examples:

• ​​A Perfect Note:​​ Imagine a hypothetical laser that emits a single, perfect frequency $\omega_0$. Its spectrum is a sharp spike, a Dirac delta function. The Fourier transform of a delta function is a pure, unending cosine wave. This means $\gamma(\tau)$ has a magnitude of 1 for all time delays $\tau$. The light is perfectly coherent with itself, no matter how far into its past you look.

• ​​A Two-Note Chord:​​ What if the source emits two distinct, sharp frequencies, $\omega_1$ and $\omega_2$? This is like striking two keys on a piano simultaneously. The spectrum consists of two delta functions. The Fourier transform of this pair of spikes gives a coherence function that looks like a high-frequency wave modulated by a lower-frequency envelope: $\gamma(\tau) = e^{-i\omega_0\tau} \cos(\Delta\omega\tau)$, where $\omega_0$ is the average frequency and $\Delta\omega$ is half the difference. The visibility, $|\cos(\Delta\omega\tau)|$, now oscillates! It starts at 1, drops to 0, revives back to 1, and so on. This phenomenon is known as "beats," familiar from sound waves. The interference fringes will disappear and reappear as you increase the path difference in your interferometer.

• ​​A Blurry Note:​​ Real-world sources, like atoms in a hot gas or LEDs, don't emit perfect frequencies. Their emission is "broadened," meaning their spectrum has a certain width, $\Delta\nu$. What happens now?

  • If the spectrum has a Lorentzian shape (common for atomic transitions), its Fourier transform is an exponential decay. The visibility of the fringes will drop off exponentially as $V(\tau) \propto \exp(-\pi \Delta\nu |\tau|)$.
  • If the spectrum has a Gaussian shape (common for Doppler-broadened gases or very short pulses), its Fourier transform is also a Gaussian. The visibility will fall off as a Gaussian function of delay $\tau$.

This leads us to two crucial concepts: the ​​coherence time​​, $\tau_c$, and the ​​coherence length​​, $L_c$. The coherence time is roughly the time delay $\tau$ over which the fringe visibility remains significant. It's inversely proportional to the spectral bandwidth $\Delta\nu$: $\tau_c \sim 1/\Delta\nu$. The broader the spectrum (more "colors" or "notes"), the shorter the coherence time. The coherence length is simply the distance light travels in this time, $L_c = c \cdot \tau_c$, where $c$ is the speed of light in the medium. White light, with its enormous bandwidth, has a coherence length of only a few microns. This is why you need a very precisely balanced interferometer to see its fringes.

Now we turn to a different, and perhaps more startling, question. How correlated is a wave field at one point in space, $P_1$, with another point, $P_2$, at the same instant? This is ​​spatial coherence​​.

Imagine illuminating a Young's double-slit apparatus not with a tiny point source, but with a large, extended source like a frosted lightbulb. The bulb is a chaotic swarm of countless independent atomic emitters. It is the very definition of an incoherent source. Yet, if we look at the light that arrives on a distant screen, we find that it isn't completely random. Over small regions, the field is correlated. There is a degree of spatial coherence. How can this be?

The answer lies in one of the most beautiful results in optics, the ​​van Cittert-Zernike theorem​​. It states that the spatial coherence function $\gamma_{12}$ on an observation plane is the Fourier transform of the normalized brightness distribution of the distant, incoherent source.

Let that sink in. The chaotic, jumbled light from the source organizes itself through the simple act of propagation. The geometry of the source is encoded, via a Fourier transform, into the spatial correlation pattern of the light far away.

The classic application is in astronomy. A star is a gigantic, incandescent ball of gas—a profoundly incoherent source. Yet, it is so far away that by the time its light reaches Earth, it has acquired a degree of spatial coherence. By building an interferometer with two telescopes separated by a baseline $d$, astronomers can sample the spatial coherence of the starlight. As they increase the baseline, the visibility of the interference fringes changes. For a circular star, the visibility follows a Bessel function pattern. The baseline $d_{min}$ at which the fringes first disappear tells them the star's angular diameter $\theta$ with incredible precision via the relation $d_{min} \approx 1.22 \lambda / \theta$. This is how we know the sizes of distant stars! The same physics applies in a laboratory setting: if you illuminate two pinholes with an extended circular source, the fringes between them will disappear when the source diameter reaches a specific size related to the pinhole separation.

To truly grasp the magic of the van Cittert-Zernike theorem, consider a source made of just two point emitters.

• ​​Case A: Incoherent Emitters.​​ The two points flash randomly and independently. The theorem applies. The coherence pattern on a distant screen is the Fourier transform of two points, which is a cosine function. The fringe visibility measured on the screen is not 1 everywhere; it varies sinusoidally with position, dropping from 1 to 0 and back again.
• ​​Case B: Coherent Emitters.​​ Now, imagine the two points are driven by the same oscillator, flashing in perfect phase-locked unison. The source itself is now coherent. The van Cittert-Zernike theorem does not apply because its starting assumption (an incoherent source) is violated. The field on the screen is now a simple, deterministic interference pattern. The entire light field is perfectly correlated with itself. If you place two detector pinholes anywhere in this field, the light they receive will be perfectly coherent, and the visibility will always be $V_B = 1$.

This comparison is the key. Spatial coherence can arise in two ways: it can be an intrinsic property of the source (Case B), or it can be generated through propagation from a spatially incoherent source (Case A).

So far, we have spoken of waves. But we know that light is made of photons, and electrons can behave as waves. Does this beautiful framework of coherence and visibility still hold? The answer is a resounding yes, and it reveals the deep unity of physics.

Consider a Michelson interferometer. If we send in a powerful beam of light, we are analyzing the collective behavior of countless photons. But what if we turn the intensity down so low that only one photon enters the interferometer at a time? There is nothing for it to interfere with, except... itself. The photon's wave packet is split, travels both paths, and is recombined to interfere with itself at the detector.

If we measure the fringe visibility by averaging over many such single-photon events, we find something remarkable: the result is exactly the same as the classical case. The visibility is still given by the Fourier transform of the spectral content of the single-photon wave packet. The coherence time is still inversely related to its spectral bandwidth. The framework holds perfectly. The same is true for electrons in a two-slit experiment. The coherence properties we attribute to classical waves are, in fact, inherent properties of the wave functions of individual quantum particles.

From the ripples in a pond to the light from a distant star, and down to the probabilistic waves of a single electron, the principles of coherence and interference provide a unified language to describe the dance of waves. It is a testament to the elegant and interconnected nature of our physical world.

Now that we have explored the fundamental connection between coherence and the visibility of interference, you might be wondering, "What is this all for?" It is a fair question. These concepts, born from simple observations of light and shadow, might seem like a niche corner of physics. But nothing could be further from the truth. The story of coherence is not a self-contained chapter in a textbook; it is a golden thread that weaves through the entire fabric of modern science and technology. It is our ruler for the unimaginably small, our telescope for the unimaginably distant, and our language for understanding the deepest quantum mysteries.

Let's embark on a journey to see where this thread leads us, from the workshop bench to the farthest stars, and into the very heart of matter.

At its most practical level, interference is a ruler. If you want to measure something with breathtaking precision—a length, a displacement, a thickness—you use light waves. But as we've learned, you can only see interference fringes if the path difference between the interfering beams is not too large. There is a limit, set by the temporal coherence of your light source.

Imagine creating Fizeau fringes, those beautiful contour lines of light and dark, using an air wedge between two glass plates. If you use a laser with its long coherence length, you can see many, many fringes extending far from the point of contact. But if you use an ordinary light bulb, the fringes will be clear only very close to the contact point and will quickly fade into a uniform blur. Why? Because as the air gap gets thicker, the path difference for light reflecting from the top and bottom surfaces increases. Once this path difference exceeds the light's coherence length, the waves can no longer interfere constructively or destructively. The fringes vanish.

You might see this as a limitation. But in science, a limitation is often just a new kind of tool in disguise. If the visibility of fringes depends on the source's properties, then by measuring the visibility, we can deduce those properties! This is the principle behind a powerful technique called ​​Fourier Transform Spectroscopy​​.

Using an instrument like a Michelson interferometer, we can systematically vary the path difference $\Delta x$ and record how the fringe visibility $V$ changes. The resulting plot of $V(\Delta x)$ is, in essence, a map of the light's temporal coherence. The central insight is that this coherence map is mathematically linked, via a Fourier transform, to the light's power spectrum—its distribution of colors or frequencies. A sharp, narrow spectrum (like a laser) corresponds to a coherence function that stretches out for a long distance. A broad, rich spectrum (like a superluminescent diode or SLD) corresponds to a coherence function that dies out very quickly. By measuring how rapidly the "wiggles" of an interferogram fade away, we can precisely calculate the spectral bandwidth $\Delta\nu$ of the source, a crucial parameter for everything from fiber optic communications to medical imaging. What was once a nuisance—the fading of fringes—has become our most precise method for measuring the spectrum of light.

Let's now turn our gaze from the lab bench to the heavens. You look up at a star. It's a point of light, a pinprick in the darkness. How can we possibly know how big it is? Or that it's not one star, but two, locked in a cosmic dance? The light that reaches us is a jumble of waves emitted from countless independent atoms on the star's surface—it's a classic example of a spatially incoherent source.

And yet, something magical happens as this light travels across the vast emptiness of space. The van Cittert-Zernike theorem tells us that propagation itself imposes a degree of order. By the time the light reaches Earth, it's no longer completely incoherent. Across any small region on Earth, the light develops "patches" of spatial coherence. The size of these coherence patches is inversely related to the angular size of the source in the sky. A truly tiny, point-like source would create a large coherence area. A larger, "fluffier" source creates smaller coherence areas.

This is the key. In the 1920s, Albert A. Michelson realized he could exploit this. Using what is now called a ​​Michelson stellar interferometer​​, he used two mirrors on a long beam, separated by a variable distance $D$. Each mirror collected starlight and directed it to a common detector to produce interference fringes. For a small baseline $D$, inside the coherence patch, the fringes were sharp and clear. But as he increased the distance $D$, moving one mirror outside the patch, the fringe visibility dropped. For a binary star system, the fringes would vanish completely when the baseline was just right, then reappear, then vanish again in a periodic pattern. The baseline $D$ at which the fringes first disappear gives a direct measure of the angular separation $\theta$ of the two stars, through the simple relation $D \approx \lambda / (2\theta)$. It is a stunning achievement: by observing the loss of interference on Earth, we can measure the structure of objects light-years away.

This principle, reconstructing a source's spatial profile from its far-field coherence function, is the foundation of modern radio astronomy. Giant arrays of radio telescopes, spread across continents, act as a single, enormous interferometer. They don't measure the image directly; they measure the coherence of the radio waves at different baselines. Then, a computer performs the inverse Fourier transform to reconstruct the image of the source. This is precisely how the Event Horizon Telescope gave humanity its first-ever picture of a black hole—not by building a lens the size of the Earth, but by masterfully sampling the coherence of light from the black hole's edge.

From the cosmic scale, let's dive down into the microscopic. The same principles of coherence are just as crucial for seeing the world of atoms and cells.

In ​​X-ray crystallography​​, scientists bombard a crystal with X-rays and study the resulting diffraction pattern to deduce the arrangement of atoms. The sharp spots in the pattern arise from the coherent superposition of waves scattered from billions of atoms in an ordered lattice. But this coherence has its limits. The X-ray beam itself has a finite temporal coherence length, $L_c$. This means that waves scattered from atoms that are too far apart within the crystal will not interfere effectively. The maximum number of lattice planes $N_{\max}$ that can contribute coherently to a Bragg peak is simply given by the ratio of the coherence length to the wavelength, $N_{\max} \approx L_c / \lambda_0$. For a typical synchrotron source, this might be around $10^4$ planes. This tells us that our ability to "see" the perfect order of a crystal is fundamentally limited by the quality of our light source. Furthermore, coherence is not just one number; we must distinguish between longitudinal (temporal) coherence, which affects scattering from atoms separated along the beam direction, and transverse (spatial) coherence, which governs scattering from atoms separated across the beam.

Moving to ​​optical microscopy​​, one might think that the most coherent light source, a laser, would be best for imaging. Often, the opposite is true. Highly coherent illumination can produce annoying speckle patterns and edge-ringing artifacts that obscure the very details we want to see. The goal is not maximum coherence, but optimal coherence. Microscopists carefully control the spatial coherence of their illumination by adjusting the condenser aperture. The degree of partial coherence, often denoted by the parameter $\sigma$, is a tunable knob that allows one to balance resolution and contrast. By changing the shape of the light source that illuminates the sample—for instance, using a ring-shaped (annular) aperture—one can precisely tailor the coherence function at the sample plane to highlight specific features. Understanding and engineering coherence is at the heart of advanced imaging techniques that allow biologists to see the delicate machinery inside a living cell.

Finally, we arrive at the deepest and most profound implications of coherence, in the realm of quantum mechanics. Here, coherence is not just a property of light, but of reality itself.

Louis de Broglie's revolutionary hypothesis was that particles like electrons and atoms also have a wave-like nature. The ultimate proof? Interference. Today, physicists can cool clouds of atoms to near absolute zero, creating a bizarre state of matter called a ​​Bose-Einstein Condensate (BEC)​​, where millions of atoms lose their individual identities and behave as a single, macroscopic quantum wave. If you take two such BECs and let them expand and overlap, they produce a stunning interference pattern of alternating high and low atomic density—a direct visualization of matter-wave interference. The visibility of these fringes is a direct measure of the phase coherence between the two separate quantum objects. Tiny fluctuations in the relative phase between the two condensates from one experiment to the next will wash out the fringes, reducing the visibility. By measuring this reduction, physicists can quantify the quantum noise in their system.

This leads us to one of the central tenets of quantum theory: wave-particle duality and the role of information. Consider a classic two-slit experiment, but now with a clever "which-path" detector. Imagine an atom going through the slits, and we place a special detector behind one slit—say, an optical cavity. If the atom passes through this slit, it imparts a tiny phase shift $\theta$ to the light field inside the cavity. If it goes through the other slit, the cavity field is unchanged. The atom and the cavity field are now entangled. The state of the cavity contains information about the atom's path.

What happens to the atom's interference pattern? Its visibility plummets. Why? Because to find the probability of finding the atom at some point on the screen, we must sum the amplitudes for the two paths. But now each path is "tagged" by a different state of the cavity. The interference term is proportional to the overlap between these two detector states, $\langle \alpha | \alpha e^{i\theta} \rangle$. The more distinguishable these two states are, the smaller this overlap becomes, and the lower the fringe visibility. The visibility $\mathcal{V}$ is found to be $\exp(-\bar{n}(1-\cos\theta))$, where $\bar{n}$ is the average number of photons in the cavity. If the states are perfectly distinguishable ($\theta = \pi$ and $\bar{n}$ is large), the overlap is zero, and the interference pattern vanishes completely. You have gained complete "which-path" information, and in doing so, you have completely destroyed the coherence. Coherence and information are two sides of the same quantum coin; you cannot have both at once.

This idea of coherence as a controllable quantum resource is the future. Imagine two quantum emitters (like artificial atoms) embedded in a tiny ring. Their ability to coherently exchange energy can be controlled by threading a magnetic flux $\Phi_B$ through the ring—a beautiful manifestation of the Aharonov-Bohm effect. By turning a knob that controls the magnetic field, an experimenter can literally switch the quantum coherence between the emitters on and off, directly manipulating the visibility of the light they emit.

From a simple tool for measuring length, to a cosmic yardstick, to a window into the quantum world, the concept of coherence reveals its power and beauty. It is a measure of order, of wavelike-ness, of potential for interference. And by learning to measure it, control it, and sometimes, to destroy it, we gain an ever-deeper understanding of the universe and our ability to shape it.