Fraunhofer Diffraction Regime

Send a laser through a narrow slit and catch the light on a wall across the room, and something quietly astonishing happens. The sharp-edged slit does not throw a sharp-edged bar of light; it paints a spread of bright and dark bands — brilliant in the middle, fading outward in a fixed, calculable rhythm. Move the wall twice as far away and the whole pattern simply doubles in size without changing its shape one bit. That scale-frozen, distance-independent pattern is the fingerprint of the far field, the Fraunhofer regime, and its explanation is one of the most powerful ideas in all of optics: the pattern on the wall is the Fourier transform of the hole in the screen. Why a Fourier transform? To a distant observer the wavelets leaving different parts of the aperture travel along essentially parallel rays, so their only relative phase is the linear path difference set by where each started. Summing those linear-phase contributions across the aperture is, letter for letter, the definition of a Fourier transform — aperture and far-field pattern form a transform pair, with observation angle playing the role of spatial frequency. A narrow aperture spreads the light widely, a wide one keeps it tight: this reciprocity, $\Delta\theta\sim\lambda/a$ , is the diffraction limit itself. The catch is distance. The clean transform emerges only once the wavefront curvature across the aperture becomes negligible — quantified by the Fresnel number $N_F=a^2/\lambda z$ falling well below one. In the lab that would demand impractical tens of metres, so a lens is used instead: it supplies the missing curvature and brings the Fraunhofer pattern into its focal plane. Every spectrometer, every telescope's diffraction limit, and all of Fourier optics rest on this one regime.

Key takeaways

Fraunhofer (far-field) diffraction is the limit in which the diffracted amplitude is the two-dimensional Fourier transform of the aperture, evaluated at spatial frequencies fixed by the observation angles ( $f_x=\sin\theta_x/\lambda$ ).
It holds when the Fresnel number $N_F=a^2/\lambda z\ll1$ , equivalently $z\gg a^2/\lambda$ — the distance at which the aperture's own wavefront curvature stops mattering.
The pattern depends only on the aperture shape: a slit gives $\mathrm{sinc}^2$ , a rectangle a product of two sincs, a circular hole the Airy disk. Its angular width scales as $\lambda/a$ .
A converging lens performs the transform physically, placing the far-field pattern in its back focal plane — which is why "far field" and "focal plane" are the same picture at finite distance.
Key idea: in the far field the diffracted amplitude is the Fourier transform of the aperture's transmission function; the aperture's own quadratic phase is negligible exactly when the Fresnel number $N_F=a^2/\lambda z\ll1$ .

The core insight

In the far field the diffracted amplitude is the two-dimensional Fourier transform of the aperture's complex transmission, sampled at spatial frequencies set by the diffraction angles. The aperture's quadratic (wavefront-curvature) phase — the term that keeps near-field patterns structured and distance-dependent — becomes negligible precisely when the Fresnel number $N_F=a^2/\lambda z$ drops well below one. A lens supplies that quadratic phase deliberately and relocates the whole Fraunhofer pattern into its focal plane.

Where this lives today

Fraunhofer diffraction is the operating principle of every dispersive spectrometer: a grating is simply a periodic array of apertures whose far field is the single-aperture pattern (the envelope) modulated by the interference of the array (the sharp orders), so the grating equation $d\sin\theta=m\lambda$ fans wavelengths apart in the focal plane. The same Fourier-transform relation is the whole foundation of Fourier optics — a lens delivers the transform of an input in its back focal plane, so a mask placed there filters spatial frequencies directly. This is the mechanism behind phase-contrast microscopy, optical low- and high-pass filtering, the $4f$ correlator, and the design of diffractive optical elements and computer-generated holograms. It also fixes the ceiling on every imaging instrument: because each aperture forms a diffraction pattern for every point source, the aperture diameter and wavelength set the finest resolvable detail through the $\lambda/a$ spread — the diffraction limit that the Rayleigh Resolution Criterion quantifies, and that drives astronomers toward ever-larger mirrors and lithographers toward ever-shorter wavelengths.

The physical picture

Picture the illuminated aperture as a dense array of coherent point emitters — the Huygens wavelets. Very far downstream the rays heading toward one screen point are essentially parallel, tilted at a single angle $\theta$ , so a wavelet launched from position $\xi$ is retarded only by the projected path $\xi\sin\theta$ — a phase linear in $\xi$ . Add up $e^{-ik\xi\sin\theta}$ over the opening and you have computed the aperture's spatial-frequency content at $\sin\theta/\lambda$ : the far-field amplitude in a given direction is one Fourier component of the hole.

Figure

Far-field geometry: a plane wave fills the aperture, parallel diffracted rays leave at angle θ, and a lens focuses each angle to a point at height f·sinθ in its back focal plane — the Fourier plane

At small angles none of the aperture's own curvature survives and the sum is a pure transform. The archetype is the single slit: a top-hat aperture whose Fourier transform is a sinc, so the screen intensity is a $\mathrm{sinc}^2$ . The next figure is that transform pair with real numbers on it — the face of Fraunhofer diffraction.

Figure

The archetypal Fraunhofer transform pair: a top-hat slit (width a = 0.10 mm) transforms into a sinc² far field. For λ = 633 nm the first minima sit at sinθ = ±λ/a = ±6.33 mrad (±0.363°); the first side lobe carries just 4.72% of the peak

This reciprocity — narrow aperture, broad pattern — is the same Fourier uncertainty that ties a short pulse to a broad spectrum, and it is why aperture diameter sets the ultimate resolving power of any imaging or spectroscopic instrument.

The core equations

The regime is defined by one statement: in the far field the diffracted amplitude is the Fourier transform of the aperture. Writing $A(\xi,\eta)$ for the complex transmission across the opening and keeping only the linear (tilt) phase that survives at large distance, the amplitude in the direction $(\theta_x,\theta_y)$ is the aperture's Fourier component at spatial frequencies $(\sin\theta_x/\lambda,\ \sin\theta_y/\lambda)$ .

U(\theta_x,\theta_y) \;\propto\; \iint_{\text{aperture}} A(\xi,\eta)\, e^{-ik(\xi\sin\theta_x + \eta\sin\theta_y)}\,\mathrm{d}\xi\,\mathrm{d}\eta \qquad (1)

where: $U$ — complex far-field amplitude in the observation direction; $A(\xi,\eta)$ — complex amplitude transmission of the aperture; $(\xi,\eta)$ — coordinates across the aperture; $k=2\pi/\lambda$ — wavenumber; $\theta_x,\theta_y$ — diffraction angles from the axis. Reading it physically: each direction on the screen samples exactly one spatial-frequency component of the hole, so the far field is the aperture's angular spectrum. That transform only emerges once the aperture's own wavefront curvature — the quadratic phase across it — can be dropped. The dimensionless dial that measures this is the Fresnel number, and the Fraunhofer regime is its small-number limit.

N_F = \frac{a^2}{\lambda z}, \qquad \text{Fraunhofer} \;\Longleftrightarrow\; N_F \ll 1 \;\Longleftrightarrow\; z \gg \frac{a^2}{\lambda} \qquad (2)

where: $a$ — aperture half-width (or radius); $z$ — aperture-to-screen distance; $\lambda$ — wavelength. The largest quadratic phase across the aperture edge is $\pi N_F$ ; requiring it to be $\ll\pi$ gives $N_F\ll1$ . When $N_F\gtrsim1$ that curvature must be kept and one is in the Fresnel (near-field) regime instead — the structured, distance-dependent patterns owned by Fresnel Diffraction Regime and dialed by Fresnel Number. Evaluating $(1)$ for a uniform slit of width $a$ — a top-hat aperture — produces the canonical result, the $\mathrm{sinc}^2$ pattern. It is the intensity from which nearly every diffraction estimate in this track descends.

I(\theta) = I_0\,\mathrm{sinc}^2\beta = I_0\left(\frac{\sin\beta}{\beta}\right)^2,\qquad \beta=\frac{\pi a}{\lambda}\sin\theta,\qquad \text{minima at } a\sin\theta = m\lambda \qquad (3)

where: $I_0$ — on-axis peak intensity; $\beta$ — half the edge-to-edge phase difference across the slit; $\mathrm{sinc}\,\beta\equiv\sin\beta/\beta$ (the unnormalised convention, $=1$ at $\beta=0$ ); $m=\pm1,\pm2,\dots$ — nonzero integer order of the dark fringe. The central lobe runs between the $m=\pm1$ zeros, giving an angular full width $2\lambda/a$ ; the side lobes fall off as $1/\beta^2$ . Because a genuine far field would demand impractical distances, a converging lens is almost always used. It stamps a quadratic phase $e^{-ik(x^2+y^2)/2f}$ onto the beam that exactly cancels the aperture's residual curvature, so the field in the back focal plane is the pure Fourier transform of the aperture — regardless of $N_F$ .

U_f(x',y') \;\propto\; \tilde{A}\!\left(\frac{k x'}{f},\,\frac{k y'}{f}\right) \qquad (4)

where: $U_f$ — field in the back focal plane; $\tilde A$ — Fourier transform of the aperture transmission; $f$ — focal length (playing the role of $z$ ); $(x',y')$ — focal-plane coordinates, with $x'\approx f\sin\theta_x$ . A lens does not "remove diffraction" — it relocates the Fraunhofer pattern from infinity to the focal plane.

Derivation AFrom the Fresnel integral to the far-field Fourier transform

The goal here is to show that the Fourier-transform relation is not a postulate but the terminal step of a controlled approximation: start from the near-field (Fresnel) propagator handed down by the parent articles, expand the path length, and find the one phase term whose neglect turns the diffraction integral into a Fourier transform — while reading off exactly how far away that neglect becomes legitimate. Everything hinges on one quadratic term and the Fresnel number that measures it.

Step 1. Begin with the paraxial Fresnel diffraction integral for a monochromatic field, the near-field workhorse recalled from Fresnel Diffraction Regime and, above it, Rayleigh–Sommerfeld Diffraction: $U(x',y')=\dfrac{e^{ikz}}{i\lambda z}\iint A(\xi,\eta)\,\exp\!\big[\tfrac{ik}{2z}((x'-\xi)^2+(y'-\eta)^2)\big]\,\mathrm{d}\xi\,\mathrm{d}\eta$ . Why: at distances large compared with the aperture the exact spherical-wave phase is well approximated by its quadratic (paraxial) expansion, which is all that bends the light.

Step 2. Expand the square in the exponent: $(x'-\xi)^2=x'^2-2x'\xi+\xi^2$ . This splits the phase into three pieces — an outer screen-curvature term $\tfrac{k}{2z}(x'^2+y'^2)$ that factors outside the integral, a cross term $-\tfrac{k}{z}(x'\xi+y'\eta)$ that is the Fourier kernel, and the aperture's own quadratic phase $\tfrac{k}{2z}(\xi^2+\eta^2)$ . Why: only the cross term carries the observation coordinates into the integrand as a transform; the outer term is a pure phase that dies in the intensity, and the aperture term is the one we must judge.

Step 3. Bound the aperture term across an opening of half-width $a$ : its maximum is $\tfrac{k}{2z}a^2=\pi\,\dfrac{a^2}{\lambda z}=\pi N_F$ . When $\pi N_F\ll1$ — that is, $z\gg a^2/\lambda$ — this phase never swings through even a radian across the aperture and may be dropped. Why: the Fresnel number is precisely the aperture's self-curvature measured in half-wave units; once it is small, the aperture looks optically flat to the far field.

Step 4. With the aperture quadratic phase gone and the outer phase pulled out, the integral is $U\propto\iint A(\xi,\eta)\,e^{-ik(\xi\sin\theta_x+\eta\sin\theta_y)}\,\mathrm{d}\xi\,\mathrm{d}\eta$ , using $x'/z=\sin\theta_x$ — equation $(1)$ , a two-dimensional Fourier transform of the aperture at spatial frequencies $\sin\theta/\lambda$ . Why: this is the whole content of the Fraunhofer regime — geometry has become spectroscopy of the hole, and every named aperture pattern is one evaluation of this transform. *Verified: the edge-phase bound $\pi N_F$ and the Fraunhofer scale $a^2/\lambda=2.0$ m for $a=1$ mm, $\lambda=500$ nm in verify_aq-optics-fraunhofer-diffraction.py.*

Go deeper · Derivation B · 1 minThe single slit as a worked Fourier transform

Derivation BThe single slit as a worked Fourier transform

This card carries out the transform of equation $(1)$ for the simplest non-trivial aperture, a uniform slit, turning the abstract "far field is a Fourier transform" into the concrete $\mathrm{sinc}^2$ pattern that anchors the rest of the diffraction track. The aim is to do the integral honestly and read off the exact minima condition — a single elementary integral does all the work.

Step 1. Take a slit of full width $a$ with uniform transmission, $A=1$ for $|\xi|<a/2$ and zero outside, infinite in the orthogonal direction. Equation $(1)$ collapses to one dimension: $U(\theta)\propto\displaystyle\int_{-a/2}^{a/2} e^{-ik\xi\sin\theta}\,\mathrm{d}\xi$ . Why: separability lets the trivial $\eta$ -integral factor out, leaving the whole angular structure in a single $\xi$ -integral over the open strip.

Step 2. Integrate the exponential: $\displaystyle\int_{-a/2}^{a/2} e^{-ik\xi\sin\theta}\,\mathrm{d}\xi = \frac{e^{-ika\sin\theta/2}-e^{+ika\sin\theta/2}}{-ik\sin\theta} = \frac{2\sin(\tfrac12 ka\sin\theta)}{k\sin\theta}$ . Writing $\beta=\tfrac12 ka\sin\theta=(\pi a/\lambda)\sin\theta$ , this is $a\,\dfrac{\sin\beta}{\beta}=a\,\mathrm{sinc}\,\beta$ . Why: the symmetric limits turn the complex exponential into a real sine, and factoring out $a$ exposes the pure sinc — the Fourier transform of a rect is a sinc, exactly.

Step 3. Square the amplitude for intensity: $I(\theta)=|U|^2=I_0\,(\sin\beta/\beta)^2$ , with $I_0$ the on-axis value where $\sin\beta/\beta\to1$ . Why: detectors register $|U|^2$ , and the on-axis limit $\lim_{\beta\to0}\sin\beta/\beta=1$ fixes the normalisation so $I(0)=I_0$ .

Step 4. Locate the zeros: $\sin\beta=0$ with $\beta\ne0$ means $\beta=m\pi$ , i.e. $a\sin\theta=m\lambda$ for $m=\pm1,\pm2,\dots$ . The first minima at $\sin\theta=\pm\lambda/a$ bracket a central lobe of angular full width $2\lambda/a$ . Why: the dark fringes are where the aperture splits into an even number of cancelling half-period zones — the Fourier zeros of the top-hat — and their spacing is the diffraction width $\lambda/a$ . *Verified symbolically ( $\int_{-a/2}^{a/2}e^{-ik\xi\sin\theta}\mathrm{d}\xi = a\sin\beta/\beta$ via sympy) and numerically (first minimum $0.363^\circ$ , full width $12.66$ mrad for $\lambda=633$ nm, $a=0.10$ mm) in verify_aq-optics-fraunhofer-diffraction.py.*

Limiting cases

Case	Result	Meaning
Far field, $N_F=a^2/\lambda z\ll1$	field $=$ Fourier transform of the aperture	pattern shape frozen in angle; only its physical size grows with $z$
Near field, $N_F\gtrsim1$	quadratic phase retained	Fresnel regime — structured, distance-dependent patterns (Cornu, zone plates)
Geometric limit, $N_F\gg1$	image $\to$ aperture's shadow	very many zones contribute; ray optics with faint edge fringes
Lens in the beam	pattern at $z=f$	lens cancels the aperture curvature; Fraunhofer at finite distance
Vanishing aperture, $a\ll\lambda$	nearly flat transform	a near-point hole radiates into all forward angles (one broad lobe)
Slit of width $a$	$I_0\,\mathrm{sinc}^2\beta$ , minima $a\sin\theta=m\lambda$	the archetypal transform pair (rect $\leftrightarrow$ sinc)

Common trapA lens removes diffraction, and "far field" just means far away

Two half-truths, both fatal. A lens does not abolish diffraction — it merely supplies the quadratic phase that the aperture's own curvature was carrying, relocating the Fraunhofer pattern from an unreachable infinity to the tidy focal plane. The diffraction spreading $\lambda/a$ is still there; it is what limits the focal spot and the instrument's resolution. And "far field" is not a vague "far enough": it is the sharp quantitative condition $N_F=a^2/\lambda z\ll1$ , equivalently $z\gg a^2/\lambda$ . A large aperture at moderate distance can sit squarely in the Fresnel regime while a pinhole a hand-span away is already Fraunhofer, because the criterion scales as $a^2$ — so always compute $N_F$ , never eyeball the distance. Confuse either point and you predict a sharp shadow where a diffraction pattern lives, or place a lens expecting a point where a sinc awaits.

Worked examples

1 · The single-slit pattern on a laboratory wall (confidence builder)

Red helium–neon light, $\lambda=633$ nm, passes through a slit of width $a=0.10$ mm. Find the first-minimum angle, the angular full width of the central lobe, and how wide the central bright band is on a screen $L=2$ m away.

Step 1. The first dark fringe is at $\sin\theta=\lambda/a=633\times10^{-9}/1.0\times10^{-4}=6.33\times10^{-3}$ , so $\theta_1=\arcsin(6.33\times10^{-3})=\mathbf{0.363^\circ}$ ( $6.33$ mrad). This is one Fourier zero of the top-hat aperture.

Step 2. By symmetry the $m=-1$ minimum sits at $-6.33$ mrad, so the central lobe spans an angular full width $\Delta\theta = 2\lambda/a = 2(6.33\times10^{-3}) = \mathbf{12.66\ \text{mrad}}$ . Every diffraction estimate you will make begins with this $\lambda/a$ .

Step 3. On a screen a distance $L$ away, the central bright band runs from $-L\tan\theta_1$ to $+L\tan\theta_1$ ; in the small-angle limit its full width is $L\,\Delta\theta = 2.0\ \text{m}\times12.66\times10^{-3} = 0.0253\ \text{m} = \mathbf{25.3\ \text{mm}}$ . A tenth-of-a-millimetre slit paints a two-and-a-half-centimetre band — the hallmark reciprocity of diffraction.

Step 4. Check the regime: with half-width $a/2=0.05$ mm, $N_F=(a/2)^2/(\lambda L)\approx2\times10^{-3}\ll1$ , so the far-field $\mathrm{sinc}^2$ is fully valid at $2$ m. First minimum $0.363^\circ$ ; central width $12.66$ mrad; band $25.3$ mm on a $2$ m screen.

2 · Where does the far field begin, and why the lab needs a lens (exam level)

A circular aperture of radius $a=1$ mm is illuminated at $\lambda=500$ nm. Estimate the distance beyond which the Fraunhofer approximation holds, and explain why a bench experiment uses a lens instead.

Step 1. The Fraunhofer condition is $z\gg a^2/\lambda$ . The scale distance is $a^2/\lambda=(10^{-3})^2/(500\times10^{-9})=\mathbf{2.0\ \text{m}}$ — the distance at which $N_F=1$ ; a clean transform needs $z$ several times larger.

Step 2. At $z=8$ m, $N_F=a^2/\lambda z=2.0/8=0.25$ , so the edge phase is $\pi N_F=0.785$ rad — not yet $\ll1$ . Even eight metres out you are only on the threshold of Fraunhofer.

Step 3. That being impractical on a bench, a converging lens is inserted: it imposes $e^{-ik(x^2+y^2)/2f}$ , exactly cancelling the aperture's quadratic phase, so by equation $(4)$ the back focal plane holds the pure Fourier transform regardless of $f$ . "Fraunhofer at infinity" and "Fraunhofer in the focal plane" are the same pattern. Scale distance $a^2/\lambda=2.0$ m; $N_F=0.25$ at $8$ m; a lens brings the same pattern to $z=f$ .

Physics of the portrait

Fraunhofer diffraction, the Fraunhofer lines, and the diffraction grating all carry one name — a Bavarian orphan who became the finest optician of his age by learning to see the far field with unmatched precision.

Origins · 1 minJoseph von Fraunhofer and how this was discovered

Joseph von Fraunhofer

1787–1826 · Bavaria — master optician and glassmaker

Joseph von Fraunhofer was orphaned at eleven and apprenticed to a Munich mirror-and-glass maker; when the workshop collapsed in 1801 he was dug from the rubble by a rescue party led by the Prince-Elector Maximilian Joseph, whose patronage bought him books and instruments. He rose to run the optical shop at the Benediktbeuern glassworks, where he turned glassmaking into a quantitative science — but in doing so he needed a way to measure the dispersion of his glasses precisely. In 1814 that search led him to map hundreds of dark lines crossing the solar spectrum, the absorption features now called the Fraunhofer lines, which he used as fixed wavelength standards (and which decades later founded astrophysical spectroscopy). To measure wavelengths outright he built the first precision diffraction gratings — ruled wires, then lines scored on glass — and worked out the far-field theory of their patterns, the diffraction regime that bears his name. His achromatic lenses and the great Dorpat refractor were the best telescopes on Earth. He died at just 39, his health undermined by years of heavy-metal fumes in the glass furnaces with tuberculosis the immediate cause; his epitaph reads Approximavit sidera — "he brought the stars nearer."

History

1786

— David Rittenhouse strings a hair grating across two screws and records the diffracted orders

1814

— Joseph von Fraunhofer discovers and maps the dark absorption lines of the solar spectrum

1821

— Fraunhofer rules precision diffraction gratings and formulates the far-field (Fraunhofer) diffraction theory

1823

— Fraunhofer measures the wavelengths of light directly from his grating patterns

1835

— Friedrich Magnus Schwerd publishes a detailed quantitative analysis of slit and grating diffraction

1882

— Henry Rowland rules large concave gratings, transforming precision spectroscopy

Going further

Texts. Hecht, Optics, 5th ed., Ch. 10 (Fraunhofer diffraction) · Goodman, Introduction to Fourier Optics, 4th ed., Ch. 4 (Fresnel and Fraunhofer diffraction) · Saleh & Teich, Fundamentals of Photonics, 3rd ed., Ch. 4 (Fourier optics) · Born & Wolf, Principles of Optics, 7th ed., §8.3 · Loudon, The Quantum Theory of Light, 3rd ed. (diffraction and the uncertainty relation).

Provenance. Equations and numbers are standard textbook results. A verifier should recompute: the binomial path expansion giving the quadratic edge phase $\pi N_F$ ; the single-slit sinc from the elementary integral $\int_{-a/2}^{a/2}e^{-ik\xi\sin\theta}\mathrm{d}\xi$ with minima at $a\sin\theta=m\lambda$ ; the Airy first zero from the first root of $J_1$ ( $3.8317$ ) giving $1.22\lambda/D$ ; and the far-field distance $z\gg a^2/\lambda$ ( $a^2/\lambda=2.0$ m for $a=1$ mm, $\lambda=500$ nm). Sources: Hecht Optics Ch. 10; Goodman Fourier Optics Ch. 4; Saleh & Teich Ch. 4. All numbers are standard; none fabricated.

Prerequisites. Huygens–Fresnel Principle · Fresnel–Kirchhoff Diffraction Integral · Fresnel Number. Leads to. Single-Slit Diffraction · Circular Aperture Diffraction and the Airy Pattern · Rayleigh Resolution Criterion · Fresnel Diffraction Regime.

What actually distinguishes Fraunhofer from Fresnel diffraction?

One phase term. Both start from the same Huygens sum over the aperture; the difference is whether the aperture's own quadratic (wavefront-curvature) phase $\tfrac{k}{2z}(\xi^2+\eta^2)$ is kept or dropped. Its maximum across the opening is $\pi N_F$ , so when the Fresnel number $N_F=a^2/\lambda z\ll1$ it is negligible and the integral becomes a clean Fourier transform — Fraunhofer. When $N_F\gtrsim1$ that phase sculpts the light into the structured, distance-dependent Fresnel patterns. It is genuinely a single dimensionless number, not a vague notion of "far," that separates the two regimes.

How can a pattern be the Fourier transform of a hole — transform with respect to what?

With respect to position across the aperture, giving spatial frequency. Each far-field direction $\theta$ maps to a spatial frequency $f_x=\sin\theta/\lambda$ , and the amplitude there is $\int A(\xi)\,e^{-2\pi i f_x\xi}\,\mathrm{d}\xi$ , the ordinary Fourier integral. So "large angle" means "high spatial frequency": a hole with sharp edges is rich in high frequencies and diffracts widely. The aperture and its far field are a Fourier pair in exactly the same sense as a signal and its spectrum.

Why does a lens produce the Fourier transform in its focal plane?

A lens converts the direction of a plane-wave component into a position: each plane wave leaving the aperture at angle $\theta$ focuses to height $x'\approx f\sin\theta$ in the back focal plane, so the focal-plane coordinate maps diffraction angle — hence spatial frequency. The lens's quadratic phase $e^{-ik r^2/2f}$ also cancels the aperture's residual curvature, removing the near-field structure. The net effect is that the focal-plane field is the pure Fourier transform of the aperture, at finite distance $z=f$ rather than at infinity.

Does the single-slit sinc² apply to every wavelength and slit, or are there limits?

It is the scalar-diffraction, far-field result, superb for slits many wavelengths wide under coherent illumination. It assumes a uniform top-hat aperture, a scalar treatment, and $N_F\ll1$ . It begins to fail as the slit approaches the wavelength — there the vector, polarisation-dependent boundary physics takes over and transmission falls faster than any scalar estimate — and it softens under non-uniform or partially coherent illumination. For ordinary bench slits at visible wavelengths none of these caveats bite, and the $\mathrm{sinc}^2$ is exact to measurement precision.

If the far field needs tens of metres, how do spectrometers fit on a bench?

They cheat with a lens (or a concave grating that is its own lens). The condition $z\gg a^2/\lambda$ would demand impractical distances for a millimetre aperture, but a converging element supplies the missing quadratic phase and drops the identical pattern into its focal plane centimetres away. The physics is far-field; the geometry is compact.

Why is the central diffraction lobe twice as wide as the side lobes?

Because the zeros of $\sin\beta/\beta$ fall at every nonzero multiple of $\pi$ , equally spaced — except that the central peak straddles two of those spacings, from $\beta=-\pi$ to $\beta=+\pi$ , while each side lobe occupies only one. In angle, the central lobe runs from $-\lambda/a$ to $+\lambda/a$ (full width $2\lambda/a$ ) and each side lobe spans $\lambda/a$ . That double width, together with the peak holding roughly 90% of the energy, is why the central maximum so dominates the pattern.