In an ideal model of an optical system, light follows a planned path: from the source, through the appropriate optical elements, to the detector or image plane. In a real device, part of the energy may be reflected off an unforeseen surface, scattered on a rough surface, reflected off a mechanical element, transmitted through a gap, or directed to the detector by a path other than the one intended in the design. That is stray light.
So the problem is not merely that there is extra light in the system. What matters is where it goes, at what power, in what spectral range, and how it affects the operation of the whole system. That is why stray light analysis is not just an add-on to the optical design — in many advanced systems it is part of the verification of the entire opto-mechanical assembly.
What is stray light?
In the simplest terms: stray light is light that reaches a place or element it was not meant for. This does not necessarily mean an outright design error. Sources of stray light can include:
- reflection off a mechanical element;
- scattering on an optical surface;
- light coming from outside the field of view;
- multiple reflections inside the housing;
- unwanted transmission through an element;
- light passing through a gap;
- thermal emission in infrared systems;
- scattering on contamination or material imperfections.
The key point is that stray light does not have a single cause. That is why its analysis requires taking into account not only the optical elements, but often the entire geometry of the device.
Where does unwanted light come from?
Reflection
Even a surface with a low reflectance can direct part of the energy in an unwanted direction. If the light then reflects off another surface, a further path is created. In a complex system there can be many such interactions.
Scattering
Real surfaces are not perfectly smooth. There can be micro-roughness, surface structure, coatings, particles, and material defects. As a result, part of the radiation is scattered outside the direction predicted by an ideal-surface model.
Multiple reflections
A ray that does not reach the detector directly may reflect off one, two, or many surfaces and then reach the detection area. Each individual reflection may be a small effect — the problem arises when many such paths add up at the same place.
Light from outside the field of view
A source located outside the nominal field of view can nonetheless generate radiation that reaches the detector through reflections or scattering. That is why analysing the main optical path alone is not always enough.
Gaps and leaks
An unwanted path can also arise from the mechanical geometry. A gap between elements, a poorly designed baffle, or a manufacturing hole can let light into an area that should be shaded.
Stray light vs. useful light
It is worth distinguishing two things: useful light — the energy the designer wants to deliver to a specific place — and stray light — the energy that gets there by another, unwanted path. If a detector that should receive a specific signal also receives extra radiation, the total power at the detector is approximately P_det = P_signal + P_stray, and the detector does not „know” that the second part comes from an unwanted path.
From the system's point of view, this means an increase in the background. If the useful signal is small, even a relatively small amount of stray light can matter. That is why the problem should be analysed not only in terms of absolute power, but also the ratio of unwanted light to the useful signal.
Why is stray light hard to predict?
In an ideal optical schematic the main path is easy to draw: source → lens → filter → detector. A real system may look more like: source → lens → reflection → housing → scattering → baffle → reflection → detector.
That is precisely why stray light is a system problem — it is not always on the main optical path. It can use mechanical elements, the housing, internal surfaces, edges, mounts, baffles, windows, filters, and lens surfaces. Stray light analysis therefore requires taking into account the real geometry and the non-sequential nature of the propagation — TracePro uses non-sequential Monte Carlo ray tracing for this.
The effect of stray light on device performance
The effect depends on the type of system.
- Imaging system — reduced contrast, degraded image quality, loss of detail, additional background, veiling glare.
- Measurement system — increased background, a change in the measured value, reduced sensitivity, a lower signal-to-noise ratio.
- Laser system — unwanted power paths, additional reflections, energy scattering, potential safety issues.
In infrared systems there is one more particular problem: thermal emission from elements inside the device. TracePro can take temperature, wavelength-dependent properties, and detection conditions into account when analysing IR systems.
Stray light in imaging systems
Imagine an objective pointed at a scene. The designer may correctly design the lenses, apertures, coatings, focusing, and detector, and yet additional background may appear in the image. Why? Because part of the light can enter the system outside the main field of view, reflect off an internal surface, be scattered, reflect off another element, and reach the detector.
So there need not be a problem with the basic optical design — the problem may lie in the interaction of the optics with the mechanics. TracePro is described as a tool for analysing the non-imaging aspects of imaging systems, including stray light, polarisation, thermal load, volume scattering, and fluorescence. An example of such an analysis in practice → Stray light analysis in a space instrument.
Stray light in infrared systems
In the infrared, stray light analysis can be even more complex, because not all sources of radiation have to be external light sources — elements of the system can themselves emit thermal radiation. This concerns, among others, housings, mounts, retaining rings, mechanical elements, and internal surfaces of the device.
In MWIR and LWIR systems such radiation can fall within the detector band, which leads to an increased background, reduced contrast, and worse detection of low-contrast objects. In such applications stray light becomes part of the whole system performance budget, not merely a mechanical problem.
Stray light in laser systems
In laser systems the problem can be especially significant because of the high energy concentration. A small part of the radiation reflected off a surface can be directed in an unwanted direction. If it then hits a detector, an electronic element, another optical component, the housing, or an area accessible to the user, it can change the operation of the system.
In high-power applications, stray light analysis can also be part of assessing the paths of potentially hazardous radiation. TracePro is used to analyse laser systems, covering reflection, absorption, scattering, and beam propagation in the real opto-mechanical geometry.
Why might sequential ray tracing not be enough?
In classical sequential ray tracing we specify the order of surfaces a ray is to pass through. This is a very effective approach for many imaging-system design problems. Stray light, however, has a different nature — we do not know in advance that a ray will go S1 → S2 → S3 → S4; instead it may go S1 → housing → S7 → S3 → baffle → S9 → detector.
That is why stray light analysis often requires non-sequential ray tracing, in which a ray can interact with successive elements according to the real geometry of the model, rather than following only a previously defined sequence. We discuss the difference between the modes in the article What is ray tracing?.
Non-sequential ray tracing analysis
In stray light analysis, the ability to model the real geometry of the device is hugely important. A single model can include lenses, mirrors, filters, baffles, the housing, mounts, mechanical elements, and absorbing and scattering surfaces. The program then traces the propagation of the rays and their interactions with the individual elements.
TracePro uses Monte Carlo ray tracing to simulate the distribution of light, and the material and surface properties can account for, among others, absorption, reflection, transmission, and scattering. This is especially important for stray light, because many problems arise precisely outside the nominal optical path.
TracePro as a tool for stray light analysis
TracePro from Lambda Research is one of the tools aimed directly at stray light analysis. The vendor positions the program as an environment for illumination design, non-sequential ray tracing, and stray light analysis. The program was originally developed among other things for stray light analysis in space telescopes and remote-sensing systems, and is now also used in other classes of optical systems.
An important feature is the ability to combine CAD geometry, optical properties, ray tracing, and path analysis in a single environment. The current version also includes tools such as Importance Sampling, Ray Sorting, and the Path Sort Table, which help identify problematic ray paths.
Modelling reflection and scattering
You cannot reliably analyse stray light if all surfaces in the model are treated as ideal. In a real device, surfaces have defined reflection, transmission, absorption, and scattering properties. TracePro lets you assign surface properties that depend on, among others, wavelength and angle of incidence, and supports various BSDF models.
This lets you replace the simplification „a black surface does not reflect” with a more realistic model: „a surface has a defined reflectance and a scattering characteristic”. This difference can matter greatly for analysing weak optical paths.
BSDF, BRDF, and BTDF — why surface data matters
One of the important concepts in stray light analysis is BSDF (Bidirectional Scattering Distribution Function) — a general family of descriptions of light scattering by a surface. Within it, one distinguishes, among others, BRDF (Bidirectional Reflectance Distribution Function), which describes the distribution of scattered light in reflection, and BTDF (Bidirectional Transmittance Distribution Function) — in transmission.
TracePro lets you use various BSDF models, including ABg, Gaussian models, and tabulated models. This has great practical importance: if you have real measured surface data, you can use it to create a surface property — Lambda Research also describes the ability to import measured scatter data using the Surface Property Generator tool. As a result, the model can be based not only on an assumption, but on data from the real component.
How to find the source of stray light?
Simply stating „too much light is reaching the detector” does not solve the problem — the designer has to answer where it came from. This is often the hardest part of the analysis. For a simple system, visualising the rays may be enough; for a more complex system, path sorting is needed.
For example, you can look for:
- rays that reached the detector;
- rays reflected off a specific surface;
- paths containing a specific element;
- rays making a specific number of reflections;
- sources responsible for the largest part of the unwanted energy.
TracePro provides the Ray Sorting and Path Sort Table tools for this.
Ray Sorting and Path Sorting
Imagine 100 units of energy reaching the detector. The designer wants to know where that energy comes from. Path analysis can let you split it, for example, into: the direct optical path, one reflection, two reflections, scattering, and a path through a specific mechanical element.
So it is not just about visualising „there are some rays here”, but about answering „this specific group of paths accounts for a defined part of the unwanted signal”. That is the information on which a design decision can be based.
How to reduce stray light?
Once the main path has been found, you can move on to eliminating it. Typical strategies:
- Increasing absorption — surfaces that should not reflect light are designed to reduce their contribution to multiple reflections.
- Changing the geometry — sometimes the most effective solution is to change the shape or position of an element.
- Adding a baffle — a baffle or another mechanical element can block a specific path.
- Changing the surface finish — so that its reflective and scattering properties are more suitable.
- Changing the optical coating — the right coating can reduce reflection or change the transmission characteristic.
- Removing a gap — if the problem results from a light leak, the solution may be to change the mechanical geometry.
The most important thing, however, is to first find the dominant path and only then modify the design.
The importance of the mechanical geometry
One of the most important lessons from stray light analysis is that an optical system does not end at the lenses — the housing is also part of the optical model. This concerns, among others, walls, mounts, rings, apertures, baffles, mounting elements, filters, windows, and covers.
If a mechanical element can interact with light, it can also become a source of, or an element in the path of, stray light. That is why non-sequential ray tracing is so useful: it lets you analyse the optics in the context of the real opto-mechanical assembly.
The role of materials, coatings, and surface finish
Two elements of identical shape can behave completely differently if they have different optical properties. In a stray light model, the reflectance, absorption, transmission, and scattering, and the dependence of these quantities on wavelength and angle of incidence, can all be significant. TracePro lets you define material and surface properties, including BSDF models, and also use your own user data.
This is especially important as a project moves from the concept phase to production. At the start you can use estimated data, and later replace it in turn with: manufacturer data → measured data → a validated model.
Stray light and prototyping
Without simulation, stray light analysis often happens only after a prototype has been built. If the detector shows too high a background, the search for the cause begins — and the problem may lie in the optics, a coating, the housing, a mechanical surface, a gap, or the alignment of an element. Such diagnostics can be time-consuming.
Simulation lets you move part of this work to the design stage — the ability to identify problematic paths and test design changes before building a prototype is one of the main values of stray light analysis. This does not mean measurements can be completely replaced by simulation; the best process is: simulation → prototype → measurement → model correlation → next iteration.
A professional stray light analysis workflow
- Step 1 — defining the problem — where the detector is, what signal is useful, what stray light level is acceptable, what spectral range matters.
- Step 2 — preparing the geometry — the model should include not only the optical elements, but also the significant mechanical geometry.
- Step 3 — material properties — transmission, reflection, absorption, scattering; where possible, based on real data.
- Step 4 — sources — the source should match the real system (spatial and angular distribution, spectral data, measured ray files).
- Step 5 — ray tracing — for stray light, non-sequential analysis is key.
- Step 6 — detectors — we define where and what we measure.
- Step 7 — path analysis — we look for the dominant routes by which unwanted energy arrives.
- Step 8 — design modification — a change in geometry, surfaces, coatings, or baffles.
- Step 9 — re-simulation — we check whether the modification actually reduced the problem.
- Step 10 — correlation with measurement — after a prototype is built, we compare the model with the real device and update it if needed.
The most common mistakes in stray light analysis
- Analysing only the main optical path — stray light is difficult precisely because it can use unplanned paths.
- Ignoring the mechanics — the housing and mounting elements can significantly affect the result.
- Ideal surfaces — assuming every surface is perfectly black, perfectly reflective, or perfectly smooth leads to an unrealistic model.
- No scatter data — if scattering is significant, use BSDF data or an appropriate surface model.
- No path information — merely determining the stray light level does not tell you what to change.
- Over-simplified geometry — removing an element that reflects light in the real device can artificially improve the result.
- Treating a simulation result as a measurement — a simulation is a model; its reliability depends on, among others, the quality of the geometry, the material data, the source, and the computation settings.
Stray light as an opto-mechanical problem
The most important conclusion is simple: you cannot effectively analyse unwanted light by looking only at the nominal optical path. You have to look at the whole system. A lens is only one element — the model may also include the housing, baffles, mounts, filters, internal surfaces, coatings, gaps, and the detector. Stray light is a good example of a problem in which optics and mechanics have to be analysed together.
Where does Optiwave fit into all this?
Not every optical problem should be solved with the same tool. For stray light in large, complex opto-mechanical assemblies, the natural choice is TracePro, because Lambda Research positions it directly for non-sequential ray tracing, illumination, and stray light analysis.
Optiwave tools such as OptiFDTD or OptiOmega, on the other hand, are intended primarily for full-wave problems and photonics components, where the key phenomena are electromagnetic on a scale appropriate for a wave model. So FDTD should not be used as a replacement for stray light analysis of a whole opto-mechanical device. It is, however, possible to use different modelling levels in one process: an electromagnetic model of the component → the component's characteristics → a system-level model → propagation analysis in the real geometry. The choice of tool should always follow from the physics of the problem — we compare the methods in the article Methods for modelling optical phenomena.
Summary
Stray light is unwanted light that reaches a place where it should not be. It can come from reflections, scattering, multiple interactions with surfaces, light from outside the field of view, leaks through the geometry, and thermal emission in IR systems. Its effect can include reduced contrast, degraded image quality, an increased background, a worse signal-to-noise ratio, measurement errors, and unwanted energy paths in laser systems.
The key point is that stray light is often a problem for the whole opto-mechanical assembly, not a single lens. That is why its analysis requires an appropriate model of the geometry, the material and surface properties, and the ability to trace unplanned ray paths — and non-sequential ray tracing works especially well for this. TracePro combines 3D geometry modelling, optical properties, scattering, Monte Carlo ray tracing, and path-analysis tools, which makes it possible to move from the question „why is too much light reaching the detector?” to the much more useful answer: „which path causes the problem, and what should be changed in the design?”. We set out the criteria for choosing a package in the guide How to choose optical simulation software?.