Incoherent Source: Understanding Spatial Coherence and the van Cittert-Zernike Theorem

Light sources can be broadly classified into two categories: coherent and incoherent. A laser produces coherent light, where waves march in perfect unison, creating sharp interference patterns. In contrast, sources like a candle, a lightbulb, or a distant star emit incoherent light—a jumble of independent waves. This raises a fundamental question: if incoherent light is so chaotic, how can it produce any form of predictable structure in the world around us? This article explores the hidden order within seemingly random light, revealing how the properties of an incoherent source are imprinted onto the light field it generates.

The journey to understanding this phenomenon unfolds across two key sections. In the first part, ​​Principles and Mechanisms​​, we will dissect the concept of spatial coherence. We will explore why interference fringes from an extended source "wash out," define the critical concept of coherence length, and build up to the van Cittert-Zernike theorem—a powerful principle that connects a source's shape to its light's coherence via the Fourier transform. Following this theoretical foundation, the second part, ​​Applications and Interdisciplinary Connections​​, will showcase the profound impact of these ideas. We will see how astronomers use coherence to measure stars, how microscopists engineer light for clearer images, and how the nature of an incoherent source even influences interactions at the quantum level.

Have you ever wondered why you can't see the beautiful rainbow-like interference fringes, like those from a soap bubble, when you look at light from a large frosted lightbulb? Yet, if you shine a laser pointer through two tiny pinholes, crisp, beautiful fringes appear as if by magic. The secret lies not in the light itself, but in its source. A laser beam is what we call ​​coherent​​—its waves march in lockstep across space. The light from a bulb, a candle flame, or a distant star is ​​incoherent​​—a chaotic jumble of light waves emitted independently from countless points across the source.

Our journey is to understand how an incoherent source, a chorus of independent singers, can sometimes conspire to produce a surprisingly ordered song, and how the size and shape of that chorus determine the nature of the harmony.

Let's return to our hero, the Young's double-slit experiment. Imagine illuminating two narrow slits, separated by a distance $d$, with a single, tiny point source of light with wavelength $\lambda$. A beautiful, high-contrast interference pattern appears on a distant screen—a classic sine wave of light and dark bands.

Now, what happens if we replace our single point source with a slightly wider, incoherent line source? Think of this line source as a row of many independent point sources, all packed together. Each point source still creates its own perfect interference pattern on the screen.

A source point directly in the middle creates a standard pattern centered on the screen's axis. But what about a point slightly off-center, say at a position $s$? As seen from the slits, its light arrives at a slight angle. This angle introduces an extra path difference between the two slits, effectively shifting its entire interference pattern sideways. The whole pattern created by source point $s$ is shifted by an amount proportional to $s$.

Because the source is incoherent, we don't add the waves' amplitudes; we add their intensities. We are summing up a whole collection of identical interference patterns, each one slightly shifted from the next.

If the source is very narrow, all the patterns are nearly on top of each other. They add up constructively, and we still see clear fringes. But as the source gets wider, the shifts become more significant. Eventually, a point on one edge of the source will be producing a bright fringe right where a point on the other edge is producing a dark fringe. The patterns begin to "wash each other out."

There is a critical source width where this cancellation becomes perfect. The first time the fringes disappear completely occurs when the pattern from one edge of the source is shifted by exactly half a fringe relative to the pattern from the center. This washes out the entire modulation. This critical condition links the source width $w$, its distance to the slits $S$, the slit separation $d$, and the wavelength $\lambda$. For a line source perpendicular to the slit separation axis, this happens when the angular size of the source, $\frac{w}{S}$, is equal to the angular separation of the fringes, $\frac{\lambda}{d}$.

This simple and powerful equation tells us something profound: the ability to see interference depends on a competition between the angular size of the source and the angular scale of the pattern you're trying to see. A larger source washes out finer fringes. Conversely, to see fringes from a given source, your slits must be close enough together. This leads us to a new, crucial concept.

The "washing out" effect is a manifestation of ​​spatial coherence​​. Light from an extended incoherent source is not completely chaotic everywhere. If you pick two points very close together in the illuminated field, the light waves arriving at them will be highly correlated, or coherent. This is because from the perspective of these two nearby points, all the different parts of the source are in almost the same direction. As you pull these two observation points further apart, they start to "see" the different parts of the source at more distinct angles, and the correlation between the waves drops.

There is a characteristic distance over which the light remains correlated enough to produce interference. We call this the ​​transverse coherence length​​, often denoted $l_c$. Our previous result gives us a direct measure of it! For a slit separation $d$ much larger than $l_c$, the fringes are washed out. For $d \ll l_c$, the fringes are sharp. The coherence length is therefore roughly:

The coherence length is the "ruler" for interference. It tells you the size of the region within which the light waves "know" about each other. A Young's experiment is, in essence, a device for measuring the spatial coherence length of a light field. Fringes are visible as long as the slit separation $d$ is smaller than $l_c$. The fringes first vanish when $d$ becomes comparable to $l_c$.

This relationship is beautifully symmetric. Consider a setup where the source width $w$ and slit separation $d$ are fixed. We can make fringes appear and disappear simply by changing the distance $S$ between the source and the slits. As we move the source closer (decreasing $S$), the coherence length shrinks, and the fringes vanish. As we move it farther away, the coherence length grows, and the fringes reappear! This is not just a thought experiment; it's a real phenomenon.

So far, we have built an intuitive picture. Now, we are ready for the grand unifying principle, a result of astonishing elegance and power known as the ​​van Cittert-Zernike theorem​​.

In simple terms, the theorem states:

The complex degree of spatial coherence between any two points in the field radiated by a distant, incoherent source is given by the Fourier transform of the source's intensity distribution.

Let's pause and appreciate this. The theorem reveals a deep duality. We know that the far-field diffraction pattern of a coherent aperture (like a single slit) is given by the Fourier transform of the aperture's shape. The van Cittert-Zernike theorem tells us that the far-field coherence pattern of an incoherent source is given by the Fourier transform of the source's shape. The mathematics describing how light's amplitude spreads from a coherent object is the exact same mathematics that describes how light's coherence spreads from an incoherent object of the same shape. It's one of nature's most beautiful pieces of poetry.

This theorem is not just a mathematical curiosity; it's a predictive powerhouse. It means the shape of the source leaves a direct "fingerprint" on the coherence of the light field it produces.

• ​​Reciprocity of Size​​: A fundamental property of the Fourier transform is that a "wide" function in space transforms into a "narrow" function in frequency, and vice-versa. For us, this means a large incoherent source produces a small region of coherence (a short coherence length). A tiny, point-like source produces a very large region of coherence. This is why starlight, coming from a source that is physically enormous but angularly tiny, is highly spatially coherent when it reaches Earth, allowing us to see it twinkle. It's also why we use a tiny pinhole to make a lamp's light coherent enough for a classroom interference experiment. The coherence area, $A_c$, is inversely proportional to the source area, $A_s$.

• ​​Reciprocity of Shape​​: The reciprocity extends to shape. If an astronomer observes a distant, uniform elliptical nebula that is twice as tall as it is wide, the theorem predicts that the area of high coherence on Earth will also be an ellipse, but one that is twice as wide as it is tall! The axes are flipped. By measuring the shape of the coherence area, we can deduce the shape of the unresolvable source. Sometimes the relationship is even more direct. An incoherent source shaped like an 'X' produces a region of high coherence that is also shaped like an 'X'.

• ​​The Source's Structure​​: The theorem applies to any source shape. Imagine a source that is itself a double slit—two narrow, incoherent strips of light separated by a distance $d$. What is the coherence pattern it produces? The Fourier transform of two sharp peaks is a cosine wave. Thus, the coherence function $\mu(x)$ in the far field will be an oscillating cosine function. Light in the observation plane will be coherent at regularly spaced bands, mimicking the interference pattern that a coherent double slit would produce. The source's structure is imprinted directly onto the field's coherence.

This theorem provides the rigorous foundation for our initial intuitive picture. For a line source of width $w$, its Fourier transform is a $\text{sinc}$ function, $\frac{\sin(x)}{x}$. The visibility of the fringes in our double-slit experiment is precisely the absolute value of this $\text{sinc}$ function. For a circular source of diameter $D$, like a star, the Fourier transform is the famous Airy pattern function, $\frac{2J_1(x)}{x}$. The first zero of this function is at $x \approx 3.83$, which gives rise to the factor of $1.22$ used by astronomers to find a star's diameter by observing when interference fringes vanish.

This principle is the bedrock of many advanced optical techniques. In microscopy, for instance, in a technique called ​​Köhler illumination​​, an extended incoherent source is used to illuminate the sample. Why not just use a laser? Because illuminating with perfectly coherent light can introduce artifacts like speckle and ringing at sharp edges. Using a source with controlled partial coherence provides a smoother, more faithful image.

The van Cittert-Zernike theorem tells us exactly how this works. An optical system can be designed to project an image of the source onto the back focal plane of the imaging lens. The diffraction pattern of the sample also forms in this plane. What you see is each diffraction spot of the sample being "blurred" into a miniature image of the source itself. By changing the size of the illumination source (often with an adjustable diaphragm), a microscopist can directly control the size of these blurred spots, tuning the coherence of the illumination to get the best possible contrast and resolution for the specific sample being viewed.

The journey that began with a simple question about a lightbulb has led us to a deep and beautiful principle connecting diffraction, coherence, and the very shape of things. The chaotic, incoherent light from a distant star or a simple filament carries within it a hidden order, a spatial structure that, if we know how to look, tells us the story of its origin.

We have spent some time wrestling with the theoretical machinery of coherence, particularly the beautiful van Cittert-Zernike theorem, which tells us how a seemingly chaotic, jumbled-up incoherent source can produce a light field with a hidden, predictable structure. This is all very elegant, but the real fun begins when we ask: so what? What good is it? Where in the real world, from the grandest scales of the cosmos to the most delicate dance of atoms, do these ideas actually show up?

The answer, it turns out, is everywhere. The principles governing incoherent sources are not just abstract curiosities for the optical physicist; they are the bedrock of technologies that let us measure distant stars, peer inside living cells, and even build molecules one photon at a time. The journey from understanding the principle to applying it is where science truly comes alive. Let's embark on that journey.

Perhaps the most breathtaking application of these ideas is in astronomy. Most stars are, for all practical purposes, perfect examples of extended, incoherent sources. They are so far away that even our most powerful telescopes cannot resolve their surfaces into a disk; they remain pinpricks of light. So how could we ever hope to measure the size of a star?

The secret lies not in forming an image, but in measuring the coherence of its light. The van Cittert-Zernike theorem gives us a stunningly powerful tool: it says that the spatial coherence function of the light arriving here on Earth is directly related to the Fourier transform of the star's shape and size as it appears in the sky. Imagine two telescopes separated by a certain distance, both pointed at the same star. If we combine the light from these two telescopes and look for interference fringes, the visibility of those fringes tells us the degree of coherence between those two points. By varying the distance between the telescopes and measuring how the fringe visibility changes, we can trace out the coherence function. From there, it's a simple mathematical step—an inverse Fourier transform—to reconstruct the angular size and even the shape of the star! We are, in a very real sense, using the star's own light as a ruler to measure it across unimaginable distances. This is the foundational principle of stellar interferometry, a technique that has allowed us to measure the diameters of stars that would otherwise be forever unresolved specks.

This cosmic-scale measurement has a direct analogy in the laboratory. If we set up a classic interference experiment, like one using Fresnel's double mirrors, and illuminate it not with a perfect point source but with an extended, incoherent line source, we see this principle in action. As the source gets wider, the beautiful, crisp interference fringes begin to wash out. Each little piece of the source produces its own interference pattern, slightly shifted from the others. When they all add together, the peaks and troughs no longer align perfectly, and the overall contrast, or visibility, diminishes. There is a precise source width at which the fringes will disappear entirely. By measuring this visibility, we are doing exactly what the astronomers do: we are probing the spatial properties of the source through the coherence of the field it creates.

The same principles that let us measure stars also dictate how we build instruments to see the very small. In a microscope, the object we are looking at is not self-luminous; we have to illuminate it. The character of that illumination is critically important. If we use a completely spatially incoherent source, like a glowing filament, each point on the source effectively illuminates the sample from a slightly different angle.

Now, suppose we want to perform a clever trick called "spatial filtering." In this technique, we manipulate the light in the back focal plane of the objective lens—a special place where the different diffraction orders from the object are separated. For example, we could block the central, undiffracted light (the zeroth order) to create a "dark-field" image where only the edges of features are highlighted. If our illumination is a coherent laser, this works beautifully. But what if we use our incoherent source? Blocking the central spot in the focal plane only blocks the undiffracted light corresponding to one of the many illumination angles. The undiffracted light from all the other points on the source, arriving at slight angles, passes right by the central stop. The net result is that our filter has almost no effect! The image looks just like it would without the filter. This teaches us a crucial lesson: many advanced imaging techniques fundamentally rely on the light having at least some degree of spatial coherence.

The coherence of light within an optical system is a subtle thing. When you look at an incoherent source like a frosted lightbulb through a simple magnifying glass, the light that enters your pupil has a certain transverse coherence length. This coherence length is determined, just as the van Cittert-Zernike theorem predicts, by the wavelength of light, the focal length of the magnifier, and the physical diameter of the source you are looking at.

However, if you use an optical system like a spherical mirror to form a sharp image of that incoherent source, something different happens. Right at the image plane, the coherence length is no longer determined by the size of the original source. Instead, it's determined by the size of the mirror's aperture—its effective "pupil." Each point in the image is formed by a cone of light rays converging from the entire aperture. The diffraction from this aperture is what sets the ultimate limit on how "point-like" an image point can be, and it is this same diffraction that dictates the spatial coherence in the immediate vicinity of that image point. This distinction is fundamental to optical design, showing that coherence properties are transformed as light propagates through a system.

We can even turn this into a powerful tool. Consider the challenge of imaging a single cell deep inside a block of light-scattering biological tissue. The light that travels directly to the cell and reflects back (the "good" light) is mixed with a huge amount of light that has been scattered multiple times by the surrounding tissue (the "bad" light). This scattered light has lost its original coherence. What if we could tune our illumination so that its spatial coherence length at the depth of the cell is exactly matched to the size of the cell itself? By doing so, we can design systems that are preferentially sensitive to the coherent "good" light, effectively rejecting the incoherent background noise. This can be achieved simply by adjusting the diameter of the incoherent source used for illumination. It's a remarkable idea: using the controlled lack of coherence from a simple source to see clearly through a cloudy medium.

The way light interacts with a surface also transforms its coherence. When light from a distant star reflects off a perfect mirror, its coherence properties are preserved. The reflected field is just as coherent as the incident field. But if that same starlight hits a piece of white paper—a diffuse, Lambertian surface—the situation changes completely. The paper itself becomes a new, extended incoherent source. The coherence of the light observed after scattering from the paper is now determined not by the angular size of the distant star, but by the physical size of the paper and its distance from the observer. This is why a spotlight creates harsh shadows and a large softbox (a diffuser) creates gentle, wrapping light; the latter is a large, nearby incoherent source that produces a field with a very short coherence length.

The influence of an incoherent source extends down to the most fundamental level: the interaction of light and matter. Imagine a collection of atoms, each with a ground state and an excited state. To pump an atom into the excited state, we need to supply it with a photon of a specific energy, corresponding to the atom's resonant frequency.

If we illuminate these atoms with a broadband, incoherent light source—think of it as a jumble of all colors—the only part of the light that matters is the energy density right at the resonant frequency. The total power of the lamp is not the whole story; what counts is the spectral power density. The rate at which atoms are excited and de-excited, and thus the steady-state population in the excited level, is determined by a balance between absorption, stimulated emission (both driven by the light source), and spontaneous emission (the atom's natural decay). This provides a direct link between the spectral properties of an incoherent source and the quantum state of the matter it illuminates.

This becomes even clearer when we compare the efficiency of different light sources for driving a quantum process. Consider photoassociation, a technique used in cold atom physics to create a molecule by joining two colliding atoms with a photon. The transition from two free atoms to an excited molecule is resonant at a very specific frequency. If we use a perfectly coherent laser tuned to this frequency, all of its power is concentrated exactly where it's needed. It's like using the perfect key for a lock. The laser is exceptionally efficient at creating molecules.

Now, what if we use an incoherent thermal light source with the same total power, also centered on the resonant frequency? This source's power is spread out over a range of frequencies. Only the small fraction of the power that falls within the molecular transition's natural linewidth can actually drive the process. The rest is wasted. The incoherent source is like trying a big ring of similar, but not perfect, keys. The coherent laser will always be more efficient, and the degree to which it is better depends precisely on how the bandwidth of the incoherent source compares to the narrow absorption line of the molecule.

The interplay can become even more intricate. In the field of nonlinear optics, we can perform a kind of "coherence alchemy." We can take an incoherent ultraviolet (UV) beam, which has a certain spatial coherence area. By focusing this beam with a lens, we transform its coherence properties into an intensity pattern at the focal plane. If we place a special nonlinear crystal there, this intense spot of UV light can generate new, visible light through a process like parametric down-conversion. This illuminated spot on the crystal acts as a new incoherent source, now emitting at a visible wavelength. The light propagating from this new source will, according to the van Cittert-Zernike theorem, develop its own spatial coherence in the far field. We have a cascade: the coherence of the initial UV beam dictates the size of the new source in the crystal, which in turn dictates the coherence of the final visible beam. We can engineer the final coherence by manipulating the initial beam, a beautiful demonstration of how these principles can be chained together in advanced photonic systems.

From the vastness of space to the intimacy of a single quantum jump, the character of an incoherent source leaves its fingerprint on the world. What first appears as randomness and disorder turns out to be a rich tapestry of structure and information, waiting for the curious mind to unravel it. The light from a simple candle, a distant star, or a humble filament lamp contains hidden correlations that allow us to measure, image, and manipulate the universe on every scale. The principles are unified and simple, but their applications are as boundless as our imagination.