How to Measure M²: A Practical Guide to Laser Beam Quality
M² is the single number that tells you how tightly your laser can focus compared with a perfect Gaussian beam. Measuring it correctly, per ISO 11146, is where most labs stumble. This guide walks through the physics, the procedure, the classic mistakes, and the instruments that turn a half-day caustic scan into a routine check.
The one-minute answer
To measure M², focus the beam with a clean, aberration-free lens and record the beam width at ten or more positions along the axis: roughly half within one Rayleigh length of the waist, half beyond two Rayleigh lengths, per ISO 11146. Use second-moment (D4σ) widths, fit the results to a hyperbola, and extract the waist diameter and far-field divergence. M² is then π · D₀ · Θ / 4λ, the product of your measured waist and divergence compared against the diffraction limit.
A motorized M² system automates that scan in minutes; instantaneous instruments capture every plane in a single frame for live monitoring. The full walkthrough below covers both.
What M² Means and Why It Matters
Every laser beam obeys the same trade: the smaller its waist, the faster it diverges. A perfect TEM₀₀ Gaussian beam sits exactly at the diffraction limit of that trade. M², the beam propagation ratio, measures how far your real beam sits above it, and because it multiplies both the waist and the divergence, it follows the beam through every lens in the system.
One number that survives every lens
The beam parameter product, waist radius times divergence half-angle, is invariant through ideal optics. M² is that product normalized to the Gaussian limit of λ/π. An M² of 1.0 is physically perfect; real lasers range from just above 1 for good single-mode fiber sources to hundreds for raw diode bars.
- Definition: M² = π · D₀ · Θ / 4λ, with D₀ the waist diameter and Θ the full divergence angle.
- Focused spot: the smallest achievable spot scales directly with M²; doubling M² doubles the focused diameter and quarters the peak intensity.
- Rayleigh length: the depth of focus zₖ = πw₀² / M²λ sets where your process stays in tolerance.
Why one measurement is never enough
A single beam profile, however pretty, cannot give M². A multimode beam can look Gaussian at one plane and blow up two Rayleigh lengths later. Only the caustic, the beam width as a function of propagation distance through a focus, separates a good beam from a lucky snapshot.
- Cutting and welding: M² sets achievable intensity and kerf width at the workpiece.
- Fiber coupling: coupling efficiency into single-mode fiber collapses as M² climbs above 1.
- Nonlinear optics and microscopy: harmonic conversion and two-photon excitation scale with focused intensity, so M² enters squared.
- Acceptance testing: M² is the contractual beam quality number on nearly every laser datasheet.
Interactive: The Beam Caustic Explorer
Drag the slider to change M². The red envelope is your beam focused through a lens; the dashed blue envelope is the ideal Gaussian from the same optics. Green dots mark the ISO 11146 measurement planes: five inside one Rayleigh length, five beyond two.
The ISO 11146 Procedure, Step by Step
ISO 11146 exists because informal M² measurements scatter wildly. Follow its recipe and two labs will agree on the same laser; skip a step and the number becomes negotiable. Here is the procedure in practice, with the mistake each step is designed to prevent.
| Step | What to do | The mistake it prevents |
|---|---|---|
| 1. Create a caustic | Focus the beam with a clean, well-aligned, aberration-free lens so an accessible artificial waist forms along your measurement axis | A lens with spherical aberration or a dirty surface adds its own M² to the result; the lens must be far better than the beam under test |
| 2. Condition the signal | Attenuate with wedges or reflective attenuators to sit comfortably in the camera’s linear range, and subtract background | Saturation clips the profile and ghost reflections add false wings; both corrupt second-moment widths dramatically |
| 3. Sample the right planes | Record at least 10 positions: about half within one Rayleigh length of the waist, half beyond two Rayleigh lengths | Points clustered near the waist cannot constrain the divergence; points only in the far field cannot locate the waist |
| 4. Measure D4σ widths | Compute the second-moment (D4σ) beam diameter at every plane, on both principal axes | FWHM and knife-edge widths understate the wings of multimode beams; ISO 11146 is defined on second moments for exactly this reason |
| 5. Fit the hyperbola | Fit D²(z) = A + Bz + Cz² and extract waist diameter, waist location, and divergence per axis | Reading the smallest measured spot as the waist misses it between sample planes and biases M² low |
| 6. Report per axis | Quote M²x and M²y separately, with wavelength, and note any astigmatism in waist positions | A single averaged number hides astigmatic beams whose two axes focus at different depths |
Measuring M² on Real Lasers
The standard is the same everywhere; the practical setup is not. Select your laser type to see what changes, and which Axiom Optics instruments are proven there.
The bread-and-butter measurement
For CW and quasi-CW lasers from the UV through the NIR, a turnkey M² station is the straightforward answer: a fixed lens, a camera on a motorized rail, and software that steps through the ISO plane set, fits the hyperbola, and reports M² per axis with full uncertainty. Attenuation is usually the only thing left to engineer, and reflective wedge attenuators keep the profile faithful at high power.
The CinSquare M² automates the complete ISO 11146 workflow from 400 to 1100 nm, driven by RayCi software for analysis and reporting, with CinCam CMOS profilers available separately when you want to build the rail yourself.
Beyond silicon’s reach
At 1.3 and 1.55 microns and across the fiber laser bands, silicon cameras go blind and the measurement moves to InGaAs. The procedure does not change, but detector linearity and background subtraction matter even more, because InGaAs sensors run warmer and multimode fiber sources carry power in broad wings that second-moment widths must capture faithfully.
The CinSquare M² VIS+SWIR spans 400 to 1700 nm in one instrument, and the CinCam InGaAs serves as a plug-and-play SWIR profiler for the planes-on-a-rail approach.
When the answer must arrive every second
A motorized scan freezes one moment of a laser’s life. On a production line, during warm-up studies, or while actively aligning a resonator, you want M² as a live number that updates as you turn the screw. Instantaneous instruments split the caustic across the detector so every frame contains all measurement planes at once, turning beam quality into a real-time process variable.
The CAM SQUARED delivers exactly that: single-frame M² with no moving parts, fast enough to watch thermal lensing develop and mode hops happen as they occur.
M² where the metal melts
In laser powder bed fusion and welding, the number that matters is the focused spot at the build plane, which combines the source’s M² with everything the delivery optics add: thermal focus shift, scanner aberrations, protective window degradation. Measuring the caustic through the actual processing focus catches problems a source-side measurement never sees.
The Focus Beam Profiler is built for exactly this, characterizing the focused spot and caustic inside additive manufacturing machines, with the broader context covered in the laser additive manufacturing application notes.
Diagnosing a Bad M²
M² tells you how far from perfect the beam is, but not why. Two beams with identical M² = 1.8 can have completely different problems: one carries higher order transverse modes from the resonator, the other is a clean fundamental mode ruined by spherical aberration in a collimator. The fix is different in each case, and the caustic alone cannot distinguish them.
Wavefront sensing closes that gap. A single-shot measurement with a HASO Shack-Hartmann sensor decomposes the beam’s phase into Zernike modes, naming the astigmatism, coma, or spherical aberration responsible, and WaveView predicts the resulting M² and PSF directly from the wavefront. Pair the diagnosis with a deformable mirror and the aberration half of the problem becomes correctable in closed loop, recovering focused intensity without touching the resonator.
Which M² Method Fits Your Laser? Answer Five Questions
Recommendation
The Axiom Optics Beam Quality Lineup
M² Measurement Systems
CinSquare M²
Turnkey motorized ISO 11146 caustic scanner for 400 to 1100 nm, with automated fitting and reporting.
View CinSquare M² →CinSquare M² VIS+SWIR
One instrument from 400 to 1700 nm, covering visible sources and the fiber laser and telecom bands.
View CinSquare VIS+SWIR →CAM SQUARED
Instantaneous M² with no moving parts: every measurement plane in a single frame for live beam quality.
View CAM SQUARED →Focus Beam Profiler
Caustic and focused spot characterization at the work plane inside additive manufacturing machines.
View Focus Beam Profiler →Profilers, Software & Wavefront Diagnosis
CinCam CMOS
Compact laser beam profilers for UV, visible, and NIR, the building block of a manual caustic rail.
View CinCam CMOS →CinCam InGaAs
Plug-and-play SWIR beam profiler for telecom and fiber laser wavelengths.
View CinCam InGaAs →RayCi Software
Profiling and M² analysis software with ISO-compliant second-moment computation and reporting.
View RayCi →HASO Wavefront Sensors
Single-shot Zernike diagnosis of the aberrations behind a degraded M².
View HASO range →WaveView
Wavefront analysis software that predicts M², PSF, and MTF directly from a measured phase map.
View WaveView →Measuring M²: FAQs
Context decides. Single-mode fiber lasers and good HeNe sources typically measure between 1.05 and 1.2. Well-designed DPSS and solid-state lasers land between 1.1 and 1.3. Multimode fiber lasers for cutting run from several to tens, trading beam quality for power, and raw diode bars reach into the hundreds on the slow axis. The right question is not how close to 1.0, but whether the value supports the focused spot and depth of focus your application needs.
Because the propagation theorem behind the hyperbola fit is only exact for second-moment widths. FWHM and knife-edge measures ignore or understate the faint wings where multimode beams carry real power, so they produce optimistic M² values that do not predict actual focusing behavior. The cost of D4σ is sensitivity to noise and background in those same wings, which is why careful attenuation, background subtraction, and a clean region of interest are part of the standard.
Not from one ordinary profile: a single plane gives a width, and any M² can produce that width somewhere along its caustic. You need the width as a function of distance through a focus. There are two legitimate shortcuts. Instantaneous instruments like the CAM SQUARED optically fold many planes into every frame, so one exposure genuinely contains the caustic. And wavefront-based methods reconstruct the full complex field from a single Shack-Hartmann measurement, from which propagation and M² follow computationally.
The native waist is usually inaccessible, buried inside the resonator or fiber, and its Rayleigh length can stretch for meters, making the required far-field planes impractical to reach. Focusing with a known, high quality lens creates an artificial waist at a convenient location with a Rayleigh length of centimeters, so the full ISO plane set fits on a rail. Since M² is invariant through ideal optics, the measured value transfers directly back to the source, provided the lens itself contributes negligible aberration.
Five account for most bad data: an aberrated or misaligned focusing lens that donates its own error to the beam; detector saturation or clipping at the region-of-interest edges; unsubtracted background inflating the second moments; sampling planes clustered in one region instead of the ISO split around the waist and far field; and quoting a single M² for an astigmatic beam whose two axes differ. Each one biases the result in a way that a hyperbola fit will happily hide.
It means the beam propagates at the diffraction limit, which is the relevant kind of perfect for focusing. It does not certify a clean intensity profile, freedom from pointing jitter, polarization purity, or spectral behavior, and a beam can carry modest wavefront error yet still measure near 1 within experimental uncertainty. M² is one essential number in a beam’s health record, best read alongside the profile itself and, when something looks wrong, a wavefront measurement.
Need a Trustworthy M² Number?
Axiom Optics has equipped laser builders, integrators, and process engineers with beam diagnostics for over a decade. Share your wavelength, power, beam size, and how often you need the measurement, and the team will recommend the right M² system, profiler, or wavefront sensor, with attenuation and integration guidance included.
Talk to a Beam Diagnostics ExpertOr call the Cambridge, MA lab at (617) 221-6636