Designing micro-objective lenses requires balancing optical performance with extremely tight mechanical constraints. To overcome the challenges, this article introduces a practical and repeatable workflow for creating robust micro-objective designs in Ansys Zemax OpticStudio, achieving high-quality results in compact imaging systems.
Along the way, we will highlight common challenges, such as limited diameter, wide field of view requirements, as well as strict tolerances, and show how to manage them effectively using OpticStudio’s modeling, optimization, and verification capabilities.
Authored By Berta Bernad
Introduction
Micro-objective lenses enable high-quality imaging in devices that must remain exceedingly small, including smartphones, robotic surgical tools, endoscopes, inspection cameras, micro-drones, and AR/VR wearables. These applications all require strong optical performance within a very limited amount of space. Meeting those requirements means addressing four recurring challenges: extreme miniaturization, wide field of view, high spatial frequency contrast, and tight manufacturing tolerances.
To ensure feasibility, we translate application needs into measurable performance targets, such as wavefront error and MTF at the sensor’s Nyquist relevant spatial frequencies, ensuring that the lenses and the sensor are properly matched for contrast at fine details. We also account for CRA (Chief Ray Angle) limits and microlens behaviour on modern CMOS sensors to avoid colour shading and corner darkening.
This article demonstrates a practical workflow overview: selecting an architecture, building a stable starting design, evolving the merit function, and finally validating robustness with Quick Yield and Monte Carlo tolerancing.
Workflow Solution
Define System Requirements
The first step in any micro‑objective design is to translate the application needs into precise, measurable system requirements. For extremely compact imaging systems, such as the one used here as example, a micro-objective lens system used in robotic surgical tools, this step ensures that all subsequent optical and mechanical decisions remain feasible as the design evolves.
Key requirements for the example system include:
-
Mechanical envelope:
Overall system diameter ≤ 2.5 mm (including mechanics)
Maximum lens diameter ≤ 2.0 mm
Image space working distance ≥ 1.0 mm
-
Optical performance: F
F number ≈ F/5
80° full field of view
Operation across visible wavelengths (F–d–C)
Sufficient contrast at ~100 lp/mm to match 5 µm pixel sensors
-
Sensor compatibility:
Image circle diameter ≤ 1.6 mm
Chief ray angle ≤ 14° to avoid shading and color artifacts
These requirements define the boundaries within which all future design decisions will be made, guiding the selection of the optical architecture, the choice of glass types, and the structure of the merit function during optimization. In OpticStudio, they are captured using the Requirements Editor, which acts as a persistent reference as the design evolves.
Fig. 1. Requirements Editor.
Build a Starting Design
With requirements defined and captured, the next step is to create a starting design using a straightforward, solvable optical architecture. For micro-objectives, a good starting point is often a sequence of a plano-concave (PCV) lens, a stop, and a plano-convex (PCX) lens, because this structure provides early control over field behaviour while keeping the system simple enough for initial optimization. A marginal ray height solve near the image plane helps manage ray bending and maintain reasonable ray bundles across fields and apertures. You can find this starting point file in the download section (micro-objective-starting-point.zar).
Fig. 2. Starting Design Lens Data & Layout.
Optimize the system
Perform the first optimization loops
We start optimization using the default merit function for RMS Spot Size, ignoring lateral color, and assuming axial symmetry.
Fig. 3. Merit Function Editor.
After the first local optimization, we can achieve only a general focus at the image plane. To improve that, we make the thickness from the last surface to the image variable and optimize again. When evaluating the system performance using FFT MTF plots, we can see how the sagittal or tangential performance is diverging or how peripheral fields are degrading faster than expected.
Fig. 4. Design after the first local optimization.
Add more degrees of freedom
As a next step, we expand the design space by adding a third PCX element, to improve aberration control while keeping the system well behaved during optimization. Key parameters, such as air gaps or thicknesses, are constrained to ensure stable convergence. Full optimization settings and results are provided in a reference file to support reproducibility.
Now, with the third lens in place, we should see immediate improvements in early optimization runs, including smoother MTF curves across the fields and better control of peripheral field performance. Additional metrics like field curvature and distortion highlight off-axis and geometric limitations, while the Seidel diagram helps pinpoint which primary aberrations are driving these limits, guiding further design decisions.
Fig. 5. Design after adding a third lens.
As the three-element architecture begins to stabilize, the next step is to expand the optimization criteria to include a broader set of aberrations that impact real-world performance. At this stage, the goal shifts from simply achieving focus and basic field coverage to shaping the optical behaviour across wavelengths, fields, and spatial frequencies.
So far, color correction has intentionally been left out of the optimization to avoid over-constraining the system. Once the base form is stable, it is appropriate to add chromatic aberration controls, typically longitudinal chromatic aberration (LCA) and lateral color terms.
Glass selection is also postponed until radii and thicknesses converge, at which point material properties can be used effectively to finetune performance.
Validate robustness
Quick Yield
Micro-objective systems are inherently sensitive to small deviations in curvature, spacing, and alignment. Before increasing architectural complexity, early robustness checks are essential. OpticStudio’s Quick Yield tool provides a fast sensitivity assessment by introducing small perturbations and observing relative performance variations. This highlights unstable elements and spacing choices before full tolerancing is attempted. At this stage, Quick Yield is used qualitatively to confirm that the design responds smoothly to perturbations and does not exhibit catastrophic performance drops.
Fig. 6. Quick Yield results using generic commercial grade tolerance preset.
Refine the merit function for advanced control
After confirming structural stability, the merit function is expanded to reflect real imaging performance:
Chromatic aberration control (LCA and lateral color)
Explicit CRA constraints to ensure sensor compatibility
Glass bounds to prevent impractical or unmanufacturable material choices
Transition from spot based metrics to MTF based optimization, aligning with how imaging performance will be evaluated in the final system
Add a doublet
To further improve chromatic and off-axis performance, a doublet is formed by pairing two adjacent elements.
Before allowing the glasses to vary, perform optimization with the current materials. This prevents the optimizer from searching too aggressively in the glass catalog before the doublet’s geometry is properly shaped, and it usually yields visible improvements in: axial color, lateral color, MTF consistency across wavelengths, and overall curvature balance.
Once the doublet geometry is stable, open the glass variables. Because a doublet relies on differential dispersion, letting the optimizer vary the glasses allows it to find the most effective pairing.
Checking chromatic aberrations ensures that chromatic corrections have not introduced unintended degradations elsewhere in the system. At this point, the design typically reaches a level of performance that makes it ready for material substitution and manufacturability checks. See the Micro-objective_doublet_subs.zar file for reference.
Fig. 7. System after glass substitution optimization.
Requirements check
If we go back to the requirements, we can see how the current design meets the original system requirements:
Total diameter remains within Ø 2.5 mm, with lens diameters ≤ Ø 2.0 mm
Achieved F number ≈ F/4.8
Full 80° field of view across the visible band
≥ 50% MTF at 100 lp/mm (sagittal MTF for max field needs further optimization)
Image circle controlled to Ø 1.6 mm
CRA ≤ 14°
≥ 1.0 mm image space working distance
Distortion is monitored rather than constrained for this application and is currently about 20% at full field.
After confirming that the nominal design meets all system requirements, the next step is to evaluate how the lens will perform when real-world manufacturing variations are introduced. This is where OpticStudio’s Tolerancing tools become essential.
Tolerancing
To begin, create a tolerance file by starting from a default tolerance set, such as the generic commercial grade preset, then modify it based on design experience. Typical adjustments include adding or tightening tilt/decenter limits, applying realistic radius tolerances based on vendor capabilities, and refining cemented or closely spaced elements where micro-objective designs tend to be most sensitive.
Once the tolerance model is defined, run a full tolerance analysis including a Monte Carlo simulations, which provides distributions of as-built performance metrics, allowing yield and robustness to be assessed directly. This tolerance analysis is a critical step in micro-objective design because the extremely small element diameters and steep ray angles can amplify manufacturing errors.
The histogram provides direct insight into whether the design is robust enough for production, whether it needs further desensitization, or whether certain tolerances must be tightened or relaxed to balance yield and cost.
Fig. 8. Histogram.
Conclusion
This design exercise demonstrates how a compact micro-objective can achieve high optical performance when requirements are clearly defined and constraints are introduced progressively. Successful outcomes rely on establishing a solid starting point, refining the architecture step by step, and using diagnostic tools, such as the Seidel analysis, MTF evaluation, and Quick Yield, to guide decisions with clear understanding. Ultimately, robust performance comes from treating image quality and manufacturability together, ensuring the final design not only meets its optical targets but can also be produced reliably at scale.
References
- O’Shea, D. C., & Bentley, J. L. Designing Optics Using Zemax OpticStudio®. SPIE Press, 2024. https://doi.org/10.1117/3.100004
- Gross, H. et al. Systematic design of microscope objectives – Part I: System review and analysis. Advanced Optical Technologies, 8(5), 2019. https://doi.org/10.1515/aot-2019-0002
- Methods for analyzing MTF in OpticStudio. Ansys Optics Knowledge Base. https://optics.ansys.com/hc/en-us/articles/42661772706835
- High CRA vs Low CRA CMOS Sensors: Impact on Lens Design Performance. June 19, 2025. https://sunex.com/2025/06/19/high-cra-vs-low-cra-cmos-sensors-impact-on-lens-design-performance/