Geometry for Optical Alignment

A comprehensive reference covering coordinate systems, trigonometric foundations, angular measurement, the paraxial approximation, reflection and refraction geometry, solid angles, beam path geometry, and practical alignment workflows for the optical table.

1Introduction

1.1The Role of Geometry in Optics

Light, in the regime where diffraction and interference effects can be neglected, travels in straight lines. This single observation — that light propagation can be modeled as geometric rays — is the foundation upon which optical engineering is built. Every calculation involving beam paths, component placement, image formation, and system alignment ultimately reduces to geometry and trigonometry [1, 2].

The ray model of light treats propagation as a set of directed line segments that obey two fundamental laws: the law of reflection and Snell's law of refraction. Both laws are geometric relationships between angles, surface normals, and refractive indices. From these two laws, the entire apparatus of geometric optics follows — lens equations, mirror formulas, prism deviation, fiber coupling, and the alignment procedures that bring real optical systems to life [1, 3].

For the working engineer, geometry is not an abstraction. It is the language of the optical table. When a laser beam drifts by 50 microradians and the target is one meter away, the resulting 50 µm displacement is a direct trigonometric calculation. When a window is inserted into a beam path at 45°, the lateral beam displacement is a refraction geometry problem. When two steering mirrors are used to walk a beam onto a new axis, the procedure exploits the geometric decomposition of beam position and angle as independent degrees of freedom [7, 8].

This guide consolidates the geometric and trigonometric tools that underpin all of optical alignment. It is prerequisite material — the mathematical foundation upon which the Lenses, Mirrors, and Optical Positioning topics build.

1.2Historical Context

The geometric treatment of light has ancient roots. Euclid's Catoptrics (circa 300 BCE) established the law of reflection and the principle that light travels in straight lines. Hero of Alexandria later showed that the law of reflection follows from the principle of shortest path. Ibn Sahl, working in Baghdad in 984 CE, discovered the law of refraction in the context of lens design for focusing light. Willebrord Snell independently rediscovered the refraction law in 1621, and René Descartes published an equivalent formulation in 1637 [1, 2].

The synthesis of geometric optics as an engineering discipline accelerated with the development of telescopes, microscopes, and precision instruments in the 17th through 19th centuries. Today, every optical designer works within the geometric framework — using coordinate systems standardized by international convention, sign rules codified by ISO, and alignment procedures refined through centuries of practice [3, 5].

2Coordinate Systems and Conventions

2.1Cartesian Coordinates

The right-handed Cartesian coordinate system is the standard reference frame in optics. By convention on an optical table, the z-axis lies along the direction of beam propagation (the optical axis), the x-axis is horizontal and perpendicular to the beam in the plane of the table surface, and the y-axis is vertical (beam height above the table) [3, 4].

Figure 2.1 — Right-handed Cartesian coordinate system on an optical table.

The right-hand rule determines the positive rotation direction for each axis. A point in 3D space is specified by the ordered triple (x, y, z). Distances between two points follow from the Pythagorean theorem:

3D Distance

d = \sqrt{(x_2 - x_1)^2 + (y_2 - y_1)^2 + (z_2 - z_1)^2}

In many alignment problems, the beam remains at constant height (y = constant) and the geometry reduces to two dimensions in the x-z plane [7, 8].

2.2Polar and Cylindrical Coordinates

Many optical components exhibit rotational symmetry about the optical axis. Polar coordinates (r, φ) are more natural than Cartesian (x, y) for these cases [1]:

Polar–Cartesian Conversion

x = r \cos\varphi, \quad y = r \sin\varphi \qquad r = \sqrt{x^2 + y^2}, \quad \varphi = \arctan\!\left(\frac{y}{x}\right)

Cylindrical coordinates (r, φ, z) extend polar coordinates along the beam propagation direction — the natural system for beam cross-sections, aperture vignetting, and round optic mounts.

2.3Spherical Coordinates

Spherical coordinates (r, θ, φ) use a radial distance and two angles: polar angle θ (from +z, 0 to π) and azimuthal angle φ (from +x, 0 to 2π) [1, 6]:

Spherical–Cartesian Conversion

x = r \sin\theta \cos\varphi, \quad y = r \sin\theta \sin\varphi, \quad z = r \cos\theta

Solid Angle Element

d\Omega = \sin\theta \, d\theta \, d\varphi

2.4Sign Conventions in Optics

Two conventions dominate: Real-is-positive (Hecht, Pedrotti) and Cartesian (engineering/ray-tracing). Mixing them is a reliable source of sign errors [1, 2, 4].

For this site, all equations follow the Cartesian sign convention: light propagates in +z, distances along +z are positive, heights above axis (+y) positive, angles from normal positive when CCW.

Quantity	Cartesian Convention	Real-is-Positive
Object distance (s)	Negative (left of element)	Positive
Image distance (s')	Positive (right of element)	Positive (real)
Focal length (convex)	Positive	Positive
Focal length (concave)	Negative	Negative
Radius of curvature (center right)	Positive	Positive

Table 2.1 — Comparison of the two primary sign conventions in geometric optics.

3Trigonometric Foundations

3.1Trigonometric Functions in Right Triangles

For an angle θ in a right triangle with sides O (opposite), A (adjacent), H (hypotenuse) [1, 2]:

Trigonometric Definitions

\sin\theta = \frac{O}{H}, \quad \cos\theta = \frac{A}{H}, \quad \tan\theta = \frac{O}{A}

\csc\theta = \frac{H}{O}, \quad \sec\theta = \frac{H}{A}, \quad \cot\theta = \frac{A}{O}

Pythagorean Identity

\sin^2\theta + \cos^2\theta = 1

In optical alignment, right-triangle geometry appears constantly. A beam at small angle θ to horizontal rises by d·tan θ over distance d. For a beam striking a tilted mirror, the reflected height involves the tangent of twice the tilt angle [3, 7].

3.2The Unit Circle and Radian Measure

The radian is the angle subtended by an arc whose length equals the radius. A full revolution is 2π radians [1]:

Degree–Radian Conversion

\theta_{\text{rad}} = \theta_{\text{deg}} \times \frac{\pi}{180} \qquad\qquad \theta_{\text{deg}} = \theta_{\text{rad}} \times \frac{180}{\pi}

Radians are the natural unit for optics because the small-angle approximation (sin θ ≈ θ) is only valid in radians. Every paraxial optics formula assumes radian measure. Mixing degrees into a radian-based formula is a guaranteed error [1, 4].

3.3Inverse Trigonometric Functions

Inverse Trigonometric Functions

\theta = \arcsin\!\left(\frac{O}{H}\right), \quad \theta = \arccos\!\left(\frac{A}{H}\right), \quad \theta = \arctan\!\left(\frac{O}{A}\right)

The two-argument arctangent function, atan2(y, x), resolves quadrant ambiguity and returns angles in (−π, π]. It is the preferred function in computational optics [3].

3.4Key Identities

Sum and Difference Formulas

\sin(\alpha \pm \beta) = \sin\alpha \cos\beta \pm \cos\alpha \sin\beta

\cos(\alpha \pm \beta) = \cos\alpha \cos\beta \mp \sin\alpha \sin\beta

Double-Angle Formulas

\sin(2\theta) = 2\sin\theta\cos\theta

\cos(2\theta) = \cos^2\theta - \sin^2\theta = 2\cos^2\theta - 1 = 1 - 2\sin^2\theta

The double-angle relationship is critical for mirror alignment: when a mirror tilts by α, the reflected beam deflects by 2α [1, 7].

Half-Angle Formulas

\sin^2\!\left(\frac{\theta}{2}\right) = \frac{1 - \cos\theta}{2}, \quad \cos^2\!\left(\frac{\theta}{2}\right) = \frac{1 + \cos\theta}{2}

4Angular Measurement

4.1Units of Angle

Unit	Symbol	In Radians	In Degrees	Typical Use
Degree	°	π/180 = 0.01745	1°	General optics, component specs
Arcminute	′	2.909 × 10⁻⁴	1/60°	Angular resolution, astronomy
Arcsecond	″	4.848 × 10⁻⁶	1/3600°	Precision alignment, telescope pointing
Radian	rad	1	57.296°	All paraxial optics formulas
Milliradian	mrad	10⁻³	0.05730°	Beam divergence, military optics
Microradian	µrad	10⁻⁶	2.063 × 10⁻⁴ °	Laser pointing stability, metrology

Table 4.1 — Angular units commonly used in optics.

Key conversion factors: 1 radian ≈ 57.3°; 1° ≈ 17.45 mrad; 1 mrad ≈ 3.44 arcminutes; 1 arcsecond ≈ 4.85 µrad.

4.2Pitch, Yaw, and Roll

The orientation of an optical component is described by three rotational degrees of freedom [3, 7, 8]:

Pitch (tip): Rotation about the x-axis. Tilts forward/backward, changing the beam's vertical angle.

Yaw (tilt): Rotation about the y-axis. Steers left/right, changing the beam's horizontal direction.

Roll: Rotation about the z-axis. No effect on centered beams through flat mirrors, but critical for polarizers, waveplates, and cylindrical lenses.

Figure 4.1 — Pitch, yaw, and roll on a kinematic mirror mount.

A kinematic mirror mount provides pitch and yaw adjustments — two degrees of freedom sufficient to control reflected beam direction in 3D [7, 8].

4.3Angular Resolution and Precision

Angular Error to Positional Error

\Delta x = L \tan(\delta\theta) \approx L \cdot \delta\theta \quad \text{(for small } \delta\theta \text{ in radians)}

Worked Example: Angular error propagation

Problem: A laser beam has pointing stability of ±50 µrad. Positional wander at 2 m?

\Delta x = 2.0 \times 50 \times 10^{-6} = 100 \text{ µm}

Interpretation: ±100 µm wander — acceptable for some setups but catastrophic for fiber coupling (core ~5–10 µm).

This linear scaling — angular error × distance = positional error — is one of the most frequently used relationships in optical engineering [4, 7, 8].

🔧 Optical Geometry Calculator — Beam Displacement Mode →

5The Small-Angle (Paraxial) Approximation

5.1The Approximation and Its Derivation

The small-angle approximation replaces trigonometric functions with first-order Taylor series terms [1, 2, 4]:

Paraxial Approximation

\sin\theta \approx \theta \qquad \tan\theta \approx \theta \qquad \cos\theta \approx 1

where θ must be in radians. From the Maclaurin series:

Taylor Series Expansions

\sin\theta = \theta - \frac{\theta^3}{3!} + \frac{\theta^5}{5!} - \cdots

\cos\theta = 1 - \frac{\theta^2}{2!} + \frac{\theta^4}{4!} - \cdots

\tan\theta = \theta + \frac{\theta^3}{3} + \frac{2\theta^5}{15} + \cdots

Dropping all terms beyond first order yields the paraxial approximation. Retaining the second term gives the third-order aberration correction (Seidel theory) [1, 5].

Figure 5.1 — The small-angle approximation on the unit circle.

5.2Error Analysis

Angle (°)	Angle (rad)	sin θ exact	sin θ ≈ θ error	tan θ exact	tan θ ≈ θ error	cos θ exact	cos θ ≈ 1 error
1	0.01745	0.01745	0.005%	0.01746	0.005%	0.99985	0.015%
5	0.08727	0.08716	0.13%	0.08749	0.25%	0.99619	0.38%
10	0.17453	0.17365	0.51%	0.17633	1.02%	0.98481	1.52%
15	0.26180	0.25882	1.15%	0.26795	2.35%	0.96593	3.41%
20	0.34907	0.34202	2.06%	0.36397	4.27%	0.93969	6.03%
30	0.52360	0.50000	4.72%	0.57735	10.3%	0.86603	13.4%
45	0.78540	0.70711	11.1%	1.00000	27.3%	0.70711	29.3%

Table 5.1 — Accuracy of the small-angle approximation at representative angles.

Worked Example: Paraxial error at representative angles

At 5° (0.087 rad): sin error = 0.13% — negligible. At 10° (0.175 rad): error = 0.51% — acceptable for engineering. At 15°: error = 1.15% — borderline. At 20°+: use exact trig. The conventional threshold is ~10° where sin θ error is ~0.5% [4].

🔧 Small-Angle Approximation Explorer →

5.3Applications in Optics

The paraxial approximation enables the linear formalism of geometric optics [1, 2, 5]:

Thin Lens Equation (Cartesian)

\frac{1}{s'} - \frac{1}{s} = \frac{1}{f}

Gaussian Beam Divergence

\theta = \frac{\lambda}{\pi w_0}

The ABCD matrix formalism describes paraxial rays as (y, θ) vectors transformed by 2×2 matrices. The linearity depends entirely on the paraxial approximation [1, 4, 6].

5.4When the Approximation Fails

The paraxial approximation fails when angles exceed ~10°–15°: high-NA objectives (0.95 NA → 72°), 45° mirror reflections, large-apex prisms, and tightly focused beams. Use exact Snell's law and numerical ray tracing [4, 5].

6Geometry of Reflection

6.1Law of Reflection

The angle of incidence equals the angle of reflection. Incident ray, reflected ray, and surface normal all lie in the same plane [1, 2]:

Law of Reflection

\theta_i = \theta_r

Angles measured from the surface normal — not from the surface [1, 4].

Figure 6.1 — Law of reflection (left) and angular doubling (right).

6.2Reflection from Flat Mirrors

Angular doubling: Mirror tilts by α → beam deflects by 2α. The single most important geometric fact for alignment [7, 8].

Lateral displacement: Mirror translated by d perpendicular to surface → beam displaced by 2d.

Image parity: Single mirror reverses handedness. Even reflections restore it; odd reflections reverse it [1].

Worked Example: Angular doubling at distance

Problem: Steering mirror 0.3 m from target, tilted 0.5° (8.73 mrad).

Beam deflects by 2α = 1.0° = 17.45 mrad.

\Delta x = 0.3 \times \tan(1.0°) = 0.3 \times 0.01746 = 5.24 \text{ mm}

Interpretation: Half-degree tilt produces 5+ mm displacement at 30 cm — demonstrating the need for fine-pitched adjusters (80–100 TPI).

6.3Retroreflection Geometry

Corner-cube: Three mutually perpendicular mirrors return beam antiparallel regardless of orientation. Three orthogonal reflections negate all direction cosines: (a,b,c) → (−a,−b,−c) [1, 4].

Cat's eye: Lens + mirror at focal plane. Retroreflects through the same aperture (no lateral offset).

6.4Multiple Reflections

Periscope: Two parallel mirrors translate beam without changing direction. Height change = 2d·cos θ [1].

Two mirrors at 90°: Returns beam antiparallel in plane of incidence — the 2D corner cube. Self-aligning, used in interferometers [1, 4].

7Geometry of Refraction

7.1Snell's Law

Snell's law governs the change in direction at a refractive interface [1, 2]:

Snell's Law

n_1 \sin\theta_1 = n_2 \sin\theta_2

Light entering a denser medium bends toward the normal (θ₂ < θ₁). Under the paraxial approximation, this linearizes to $n_1 \theta_1 = n_2 \theta_2$ — the basis of the ABCD matrix formalism [4].

7.2Beam Displacement Through Parallel Plates

A collimated beam through a plane-parallel plate emerges parallel but laterally displaced [1, 2, 4]:

Lateral Beam Displacement

d = t \cdot \frac{\sin(\theta_1 - \theta_2)}{\cos\theta_2}

Lateral Displacement (Closed Form)

d = t \sin\theta_1 \left(1 - \frac{\cos\theta_1}{\sqrt{n^2 - \sin^2\theta_1}}\right)

Figure 7.1 — Lateral beam displacement through a plane-parallel plate.

Worked Example: Window displacement

Problem: HeNe beam through 5 mm N-BK7 window (n = 1.5151) at 45°.

\sin\theta_2 = \frac{0.7071}{1.5151} = 0.4668 \;\Rightarrow\; \theta_2 = 27.82°

d = 5.0 \cdot \frac{\sin(17.18°)}{0.8849} = 5.0 \cdot \frac{0.2954}{0.8849} = 1.669 \text{ mm}

Interpretation: A standard 5 mm window at 45° displaces the beam by ~1.7 mm. Must be accounted for when inserting or removing windows from beam paths.

🔧 Optical Geometry Calculator — Parallel Plate Mode →

7.3Total Internal Reflection

Critical Angle

\theta_c = \arcsin\!\left(\frac{n_2}{n_1}\right)

For θ₁ > θ_c, total internal reflection occurs — the operating principle behind optical fibers, Porro prisms, and pentaprisms [1, 6].

Material	n at 632.8 nm	θ_c (air)	Application
N-BK7	1.5151	41.3°	General-purpose optics
Fused Silica	1.4570	43.3°	UV optics, laser windows
N-SF11	1.7786	34.2°	High-dispersion prisms
Sapphire (nₒ)	1.7659	34.5°	Harsh-environment windows
Diamond	2.4176	24.4°	ATR spectroscopy

Table 7.1 — Critical angles for common optical materials at 632.8 nm.

7.4Prism Geometry

Prism Deviation

\delta = \theta_1 + \theta_4 - A

Minimum deviation when ray passes symmetrically (θ₁ = θ₄) [1, 2]:

Minimum Deviation

n = \frac{\sin\!\left(\frac{A + \delta_{\min}}{2}\right)}{\sin\!\left(\frac{A}{2}\right)}

This provides an elegant method for measuring refractive index and connects prism geometry to dispersion.

8Solid Angles and Radiometric Geometry

8.1Definition and Units

Solid Angle Definition

\Omega = \frac{A_{\text{sphere}}}{r^2} \quad \text{[steradians, sr]}

Full sphere = 4π ≈ 12.57 sr. Hemisphere = 2π ≈ 6.28 sr [1, 6].

8.2Solid Angle of a Cone

Solid Angle of a Cone (Exact)

\Omega = 2\pi(1 - \cos\theta)

\Omega = \int_0^{2\pi}\!\int_0^{\theta}\sin\theta'\,d\theta'\,d\varphi = 2\pi(1-\cos\theta)

Figure 8.1 — Solid angle subtended by a circular aperture.

8.3Small-Angle Approximation for Solid Angles

Solid Angle (Small-Angle)

\Omega \approx \pi\theta^2 \approx \frac{\pi}{4(f/\#)^2} \approx \pi \cdot \text{NA}^2

Worked Example: Solid angle of a collection lens

Problem: 25.4 mm diameter lens at 100 mm from a point source.

θ = arctan(12.7/100) = 7.24° = 0.1264 rad

Exact: Ω = 2π(1 − cos 7.24°) = 0.05009 sr

Approximate: Ω ≈ πθ² = 0.05020 sr (error: 0.22%)

Interpretation: At f/# ≈ 3.9, the πθ² approximation works well.

🔧 Optical Geometry Calculator — Solid Angle Mode →

8.4Throughput and Étendue

Étendue

G = n^2 A \Omega \quad \text{[m}^2\text{·sr]}

Étendue (NA and f/# Forms)

G = \pi n^2 A \sin^2\theta = \pi A \cdot \text{NA}^2 = \frac{\pi A}{4(f/\#)^2}

Étendue is conserved in lossless systems (Liouville's theorem). No passive optical system can increase source brightness. Optimal coupling matches source and receiver étendue [6].

9Beam Path Geometry on the Optical Table

9.1Defining a Beam Path

The optical axis is defined by two reference irises placed as far apart as practical. Any beam has four parameters: two position (x, y) and two angle (θ_x, θ_y). Two irises provide exactly four constraints [7, 8]. Standard beam heights: 76 mm (3″), 102 mm (4″), or 127 mm (5″).

9.2Two-Mirror Steering (Walking the Beam)

The most fundamental alignment procedure uses two steering mirrors [7, 8]:

Mirror 1 (position mirror): Near Iris 1 — adjustments change beam position.

Mirror 2 (angle mirror): Near Iris 2 — adjustments change beam angle.

Iterate: (1) M1 for Iris 1, (2) M2 for Iris 2, (3) repeat until converged.

Figure 9.1 — Two-mirror beam walking. M1 controls position at I1; M2 controls angle at I2.

Each kinematic mirror provides pitch + yaw (2 DOF), giving 4 total — exactly matching the 4 beam parameters. Two mirrors are necessary and sufficient for complete alignment [7, 8].

9.3Z-Fold and Figure-4 Configurations

Z-fold: Two mirrors at 45° incidence. Simplest, most common.

Figure-4: Two mirrors at 67.5° incidence. More compact but steeper angles may introduce polarization effects.

Figure 9.2 — Z-fold and Figure-4 configurations.

9.4Folded Beam Paths

Mirror Separation for Height Change

L_{\text{separation}} = \frac{|y_2 - y_1|}{\sin(2\alpha)}

For 45° incidence: L = height difference.

Worked Example: Two-mirror beam height change

Problem: Beam at 100 mm must reach 75 mm using Z-fold at 45°.

L = |100 − 75| / sin(90°) = 25 mm — very compact.

9.5Alignment Error Propagation

Alignment errors propagate geometrically. A decentered beam through a lens of focal length f acquires angular error [4, 7]:

Decentration Angular Error

\delta\theta = \frac{\Delta x}{f}

Cumulative Error

\Delta x_{\text{total}} = \sum_{i=1}^{N} L_i \cdot \delta\theta_i

Worked Example: Cumulative alignment error

Problem: Three mirrors: M1 (20 µrad, 1000 mm to target), M2 (10 µrad, 500 mm), M3 (15 µrad, 200 mm).

Δx₁ = 1.0 × 20 µrad = 20 µm; Δx₂ = 0.5 × 10 = 5 µm; Δx₃ = 0.2 × 15 = 3 µm

Total: 28 µm. First mirror dominates — errors early in the path have the largest impact [7].

🔧 Optical Geometry Calculator — Error Propagation Mode →

10Practical Selection and Workflow

10.1Step-by-Step Alignment Workflow

Systematic alignment procedure for free-space optical systems [7, 8]:

Step 1: Define beam height (76, 102, or 127 mm). Verify with rulers at multiple positions.

Step 2: Set laser output parallel to table at chosen height.

Step 3: Place two irises along beam path, far apart as practical.

Step 4: Walk beam with two mirrors. M1 for Iris 1 position, M2 for Iris 2 angle. Iterate.

Step 5: Add elements sequentially. Check downstream iris after each. Center lenses via autocollimation [7].

Step 6: Verify and lock. Tighten all mounts, re-check all irises.

10.2Common Pitfalls

Forgetting angular doubling. Mirror tilt α → beam deflection 2α. Most common geometric error.

Mixing sign conventions. Choose one, use it throughout.

Neglecting beam height. Unlevel beam arrives at different heights on different components.

Irises too close together. Poor angular sensitivity — maximize separation [7, 8].

Not centering on elements. Decentered beam acquires angular error Δx/f [7].

10.3Reference Geometry Quick-Reference

Formula	Expression	Use
Angle conversion	θ_rad = θ_deg × π/180	Radians for paraxial formulas
Positional error	Δx = L · δθ	δθ in radians
Angular doubling	Deflection = 2 × mirror tilt	Key alignment fact
Snell's law	n₁ sin θ₁ = n₂ sin θ₂	Refraction at interfaces
Critical angle	θ_c = arcsin(n₂/n₁)	TIR threshold
Parallel plate	d = t · sin(θ₁−θ₂)/cos θ₂	Window displacement
Solid angle (cone)	Ω = 2π(1−cos θ)	Light collection
Étendue	G = n²AΩ	Throughput conservation
Small-angle	sin θ ≈ tan θ ≈ θ, cos θ ≈ 1	Valid θ ≲ 10° (radians)

Table 10.1 — Essential geometric formulas for optical alignment.

References

[1]E. Hecht, Optics, 5th ed. Pearson, 2017.
[2]F. L. Pedrotti, L. M. Pedrotti, and L. S. Pedrotti, Introduction to Optics, 3rd ed. Cambridge University Press, 2017.
[3]J. E. Greivenkamp, Field Guide to Geometrical Optics, SPIE Press, 2004.
[4]W. J. Smith, Modern Optical Engineering, 4th ed. McGraw-Hill, 2008.
[5]R. E. Fischer, B. Tadic-Galeb, and P. R. Yoder, Optical System Design, 2nd ed. McGraw-Hill, 2008.
[6]B. E. A. Saleh and M. C. Teich, Fundamentals of Photonics, 3rd ed. Wiley, 2019.
[7]Thorlabs, “Beam Alignment Procedures,” Application Note. Available: thorlabs.com.
[8]Newport Corporation, “Practical Beam Steering and Alignment,” Technical Note. Available: newport.com.

← View Abridged Version 🔧 Optical Geometry Calculator →