Key equations and principles
Frequently asked questions (FAQ)
What are the Fresnel equations?
The Fresnel equations describe the amplitude reflection coefficient r and amplitude transmission coefficient t for light at the interface between two media, for s (perpendicular) and p (parallel) polarisations. The reflectance is R=|r|², while the transmittance must include the cosine ratio. For non-absorbing media, R + T = 1 holds strictly for both polarisations, expressing energy conservation.
Why must the transmittance formula include the cosine ratio?
Because the propagation direction changes on refraction, the beam cross-sectional area changes by the factor n₂cosθ₂/(n₁cosθ₁). Simply squaring the amplitude to obtain power gives the wrong answer, hence T = (n₂cosθ₂/n₁cosθ₁)|t|². Omitting this factor is the most common beginner’s error.
What is the Brewster angle?
When the angle of incidence is θ_B = arctan(n₂/n₁), the reflection coefficient for p polarisation is exactly zero; this is the Brewster angle, about 56.3° for air → glass. At this angle the s-polarised reflectance is still about 14.8%, so it is not a ‘zero-reflection angle’.
What are the conditions for the critical angle and total internal reflection?
A critical angle θ_c = arcsin(n₂/n₁) exists only when light travels from a denser to a rarer medium (n₁ greater than n₂). Once the angle of incidence exceeds θ_c, the refraction angle has no real solution and all the light is reflected back into medium 1, giving R = 1 and T = 0 — this is total internal reflection.
What happens at grazing incidence (approaching 90°)?
As the angle of incidence approaches 90°, the reflectance of both polarisations tends to 1 (complete reflection). This is an asymptotic process rather than a sudden jump to 1 at a specific angle, so the tool displays close to 100% rather than exactly 100% at 89.9999°.
How do I read the reflectance curves?
The upper plot shows Rₛ, Rₚ and unpolarised R over the full 0–90° range. The lower plot zooms in near the Brewster angle: the red Rₚ curve dips to zero here, the key region for suppressing p-polarised stray light and designing polarising beam splitters; the Rₛ curve remains well above Rₚ throughout.
Why can’t this tool calculate the reflectance of metals?
This tool is written for non-absorbing, non-magnetic media (real refractive indices), where R + T is identically 1. Metals require a complex refractive index n − ik, the Fresnel equations must be rewritten in complex form, and the reflectance no longer satisfies R + T = 1. For metals and absorbing media, please calculate separately using a complex refractive index.
What is the sign convention for the p component?
This tool uses rₚ = (n₂cosθ₁ − n₁cosθ₂)/(n₂cosθ₁ + n₁cosθ₂), giving rₚ = −rₛ = (n₂−n₁)/(n₂+n₁) at normal incidence. Physically the two sign conventions are equivalent; what matters is that the power reflectance and the normal-incidence value remain continuous. The reflectance itself is unaffected by this sign choice.
Can real coatings achieve zero reflection?
A single interface cannot eliminate reflection; air → glass at normal incidence still gives about 4%. In practice anti-reflection coatings (a single quarter-wave layer with refractive index √(n₁n₂), or broadband multilayer stacks) reduce the reflectance to a few parts per thousand. This tool gives the Fresnel reflectance of an uncoated bare interface.
How can I use it to estimate stray light in an optical system?
Simply sum the R of each air–glass interface. For example, an optomechanical system with six glass elements has twelve air–glass interfaces; before coating, the total surface reflectance is about 1 − (1−0.04)¹² ≈ 39%, which is the stray light that anti-reflection coatings must suppress.
What are the unit conventions for the calculated results?
The refractive index n is a dimensionless ratio, and angles of incidence and refraction are in deg. r and t are amplitude coefficients (dimensionless, may be negative), while R and T are power percentages. Internally the tool computes angles in radians to ensure accuracy.


