1
$\begingroup$

This question already has an answer here:

I'm interested in camera lenses and how things like barrel distortion and lens flare happens, but know that one cannot just start there. I'm a math major and I've worked out of Strauss PDE book so I know about the wave equation and some stuff beyond as well. Are there any good books out there for me to self-teach so I can at least begin to understand all this?

$\endgroup$

Before answering, please see our policy on resource recommendation questions. Please write substantial answers that detail the style, content, and prerequisites of the book, paper or other resource. Explain the nature of the resource so that readers can decide which one is best suited for them rather than relying on the opinions of others. Answers containing only a reference to a book or paper will be removed!

marked as duplicate by Qmechanic Dec 1 '18 at 7:09

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

  • 1
    $\begingroup$ Lens flare is just internal reflections and one wouldn't go beyond ray optics and Fresnel equation for the modeling. If you want to simulate it, then you will either need an optical raytracing program or write the code yourself (there is the theoretical possibility of pen and paper, if you are really ambitious... it's "just" matrix-multiplications...). In general, photographic lenses are probably being designed without wave optics, since they are usually not operating close to the diffraction limit. $\endgroup$ – CuriousOne May 26 '16 at 0:01
  • 1
    $\begingroup$ If simulating raytraced graphics is your goal, look into Blender computer graphics software with Yafaray ray tracing extensions. Both are free and there is a Blender StackExchange. $\endgroup$ – paisanco May 26 '16 at 0:28
  • $\begingroup$ @CuriousOne Your last statement has certainly been true in the past, but I'm not sure it holds now. I'm not close to camera design, but I suspect that economics have shifted. The CCD chip is now a very expensive part of specialist cameras, so to get the most out of it means that you want the resolution of your optics to match the CCD cell sizes, i.e. pretty much the diffraction limit for typical camera lens dimensions. In the old days, there was no economic penalty of failing to approach the expensive-to-realize diffraction limit. But now the diffraction limit is much cheaper to approach. $\endgroup$ – WetSavannaAnimal May 26 '16 at 0:46
  • $\begingroup$ @WetSavannaAnimalakaRodVance: Most consumer lenses are not good enough to exploit the resolution of CCDs in the latest cameras. A lens that is as good or better than the sensor can easily set you back 2-10 times the cost of the camera body and I would estimate that the glass that a serious amateur (let alone a professional) carries around these days is almost two orders of magnitude more expensive than the silicon in the camera. I wish you were right, but the stuff in my camera bag clearly doesn't meet that high a quality standard. $\endgroup$ – CuriousOne May 26 '16 at 1:01
1
$\begingroup$

As @CuriousOne says, both the effects you refer to are explicable with ray optics. But both are subtle insofar that you won't find easy expressions for them in any undergrad physics course: you actually have to do accurate numerical ray simulations to see that spherical lenses impose geometrical distortion (barrelling and so forth). Very roughly, geometric distortion arises simply because Snell's law is $n_j\,\sin\theta_j = const$ at interfaces instead of $n_j\,\theta_j = const$. The latter, approximate form would mean that a lens system would impose conformal, linear maps on the object plane. The higher order terms in the sine Taylor series mess up conformalhood and linearity more and more as the ray angles increase, i.e. the system deviates from the so called paraxial approximation more and more.

There are rules of thumb - not very accurate - that designers used to estimate geometric distortion in the old days before computer design. All of these rules of thumb are grounded on the Seidel distortion measure, which is the geometric distortion that arises from spherical surface lenses calculated by assuming that Snells law is $n_j (\theta_j - \frac{\theta_j^3}{3!}) = const$. The five Seidel aberrations, of which geometric distortion is one, are treated thoroughly in Chapter 5 "Geometrical Theory of Aberration" in:

Born, M and Wolf, E, "Principles of Optics: Electromagnetic Theory of Propagation, Interference and Diffraction of Light", Seventh edition with errors corrected 2002, Cambridge University Press

Ludwig von Seidel was the originator of this treatment, although he is not well kenned outside the German speaking world. I don't think there are English translations of his works and Born and Wolf are his most thorough messengers.

Lens flare is almost always modelled numerically if it is modelled at all. There isn't really a systematic treatment as the phenomenon itself is a mess. It arises from scattering from unwanted centers in lens glasses, from multiple reflexions in poorly coated lenses and scattering off lens train hardware - e.g. edges of apertures. Simply put, some unwanted light path forms a bright spot, much brighter than the surrounding image, and the affected pixels saturate given the relatively high CCD sensitivity you need to glean the rest of your image, which is at much lower intensity than the flare. You need a really advanced optical design software, that can handle multiple beam paths and reflexions, to analyze this phenomenon. Most designers treat lens flare like electrical engineers treat electromagnetic interference: it's hard to model and best reduced by following "codes of practice"- standards that dictate quality of blacking of lens hardware (i.e. making them rough and black so that they don't scatter) and similar flare mitigation measures.

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.