I recently came across an interesting question: “What is the best form of a singlet lens?” At first, it seemed trivial — the answer is clearly the plano-convex lens. That’s what many optical design texts suggest, since it minimizes spherical aberration and coma. But I soon discovered that this answer relies on certain assumptions.
The shape of a singlet lens is often characterized by the shape factor \(q\):
\[q = \frac{R_2+R_1}{R_2-R_1}\]where:
Many standard optical design textbooks identify the plano-convex lens as the “best form” when minimizing primary aberrations. The plot below, taken from Optical Systems Engineering by Keith Kasunic (and originally from Fundamentals of Optics by Jenkins and White), shows how aberrations vary with shape factor \(q\). Spherical aberration is minimized around \(q =0.714\), and coma is zero at \(q=0.800\). A plano-convex lens, with the curved side facing the collimated light, closely matches this ideal.
Figure from Optical Systems Engineering from Keith Kasunic
We can confirm this with a Zemax simulation. I designed a lens with the following constraints:
Below are the optimization results:
Surface Type | Radius | Thickness | Material |
---|---|---|---|
(OBJECT) Standard | Infinity | Infinity | |
(STOP) Standard | 76.889 | 5.000 | BK7 |
Standard | -475.000 | 124.324 | |
(IMAGE) | Infinity |
Not very surprising — it is a plano-convex — but things get more interesting when we consider the field of view.
The plano-convex lens performs well for collimated light (on-axis). But what if we introduce a 50° field of view and optimize again? The result is no longer plano-convex. It’s a meniscus lens. Below are the optimization results:
Surface Type | Radius | Thickness | Material |
---|---|---|---|
(OBJECT) Standard | Infinity | Infinity | |
(STOP) Standard | -53.012 | 10.000 | BK7 |
Standard | -28.987 | 94.699 | |
(IMAGE) | Infinity |
This is quite fascinating—let’s delve deeper into the transverse aberration coefficients for spherical aberration (TSPH), coma (TTCO), and astigmatism (TAST).
Surface | TSPH | TTCO | TAST |
---|---|---|---|
1 | 0.0649 | 1.404 | 6.753 |
2 | 0.0610 | -1.722 | 10.80 |
Total | 0.1259 | -0.318 | 17.55 |
Surface | TSPH | TTCO | TAST |
---|---|---|---|
1 | -0.229 | 3.185 | -9.841 |
2 | 3.3795 | -14.05 | 12.987 |
Total | 3.1504 | -10.87 | 3.146 |
To interpret these values meaningfully, it is helpful to recall how each type of transverse aberration scales with pupil coordinate \(\rho\) and image height \(h\):
\[e_y(\rho, h) = TSPH\cdot \rho^3\]\[e_y(\rho, h) = TTCO\cdot \rho^2h\]\[e_y(\rho, h) = TAST\cdot \rho h^2\]Astigmatism in the plano-convex lens is high — and since it scales with \(h^2\), it quickly dominates at large field angles. The meniscus lens, though it has more coma, maintains low astigmatism, which is more important off-axis. This can be seen in the transverse ray plot below, where Plano-convex lens demonstrate large astigmatism at \(50^o\) field angle.The meniscus lens shows smaller overall error — except near the edge. This suggests we can do better by shifting the stop and optimizing again.
If we stop-shift the meniscus lens, we can achieve a even better performance across all aberrations! The transverse aberration coefficients are further optimized.
Surface Type | Radius | Thickness | Material |
---|---|---|---|
(OBJECT) Standard | Infinity | Infinity | |
(STOP) Standard | Infinity | 12.052 | |
Standard | -151.505 | 20.000 | BK7 |
Standard | -41.130 | 97.046 | |
(IMAGE) | Infinity |
Surface | TSPH | TTCO | TAST |
---|---|---|---|
1 | -0.006 | 0.267 | -2.326 |
2 | 1.525 | -1.098 | 0.176 |
Total | 1.344 | -0.8318 | -2.150 |
Comparing the stop-shifted meniscus and plano-convex lenses:
That is why simple eyeglass lenses, camera objectives, and AR/VR correction optics like those in the Apple Vision Pro often use a meniscus form. Even early cameras from 200 years ago relied on this basic principle.
What starts as a “trivial” answer reveals a much richer story — once we account for field of view.