• Specular reflection Occurs when the material "echoes" the incident light, with same angle (relative to the surface normal line).
  • Diffuse reflection seems quite similar to Scattering. Yet the article says:

The incident light ray is somewhat reflected to every direction (i.e broadcasted), and it is more intense in the direction normal to the surface (Lambert cosine law).

However the article says

Few materials do not cause diffuse reflection: among these are metals, which do not allow light to enter; gases, liquids, glass,

And later:

A few materials, like liquids and glasses, lack the internal subdivisions which produce the subsurface scattering mechanism described above, and so give only specular reflection

I don't really see why Scaterring wouldn't count as Reflection though. What are the main differences between Scattering and Diffuse Reflection?

It goes on to contradict itself as far as I understand:

The vast majority of visible objects are seen primarily by diffuse reflection from their surface.[7][8] Exceptions include objects with polished (specularly reflecting) surfaces, and objects that themselves emit light. Rayleigh scattering is responsible for the blue color of the sky, and Mie scattering for the white color of the water droplets in clouds.

I'd also appreciate any light with the paragraph:

In this case, diffused rays will lose some wavelengths during their walk in the material, and will emerge colored.

from that same article as well.


1 Answer 1


I usually think of specular and diffuse reflections a little differently than the wikipedia article.

a) Specular reflection - smooth interfaces where geometric optics and ray tracing can be used. And angle of incidence = angle of reflection

b) Diffuse reflections - rough interfaces where specular reflection where for the majority parallel rays coming in you get a Lambertian type intensity intensity profile coming out.

c) Scattering - a specific process where for a plane wave or photons interaction with a particle, or material and you can compute the distributions of photons as a function of angle. e.g. forward scattering, back scattering, Mie scattering, Raleigh Scattering, Raman Scattering, etc.

d)Scattering - some light going in some direction I would not like it go, e.g. stray light in an optical system. In this case If you think of it an energy balance Total energy = Transmitted Energy + Reflected Energy + Absorbed Energy + Scattered energy.

e) Other things that can happen when the light hits the surface if the surface is structured in some way like diffraction from a grating, or interference, light emission if the material is excited etc.

So the problem with my definition above for diffuse reflection above is that lots of situations may not be exactly a Lambertian distribution. Realistically you need to consider what the length scale of the roughness compared to the wavelengths you are interested in. Depending on the situation is something more than than geometric ray tracing needed? Do I need use the wave nature of light to be able to get a solution? Can I take a shortcut if there is some structure or particle size or characteristic of the material. So the article is trying to talk about a lot of different cases and trying to do so generally without explaining the complexity. In some cases they are considering a white light source rather than a specific wavelength since that is what we often do for photography or other applications. In that context they can talk about subsurface layers, and absorption of a specific wavelength giving colors etc. because they are making some assumptions that are relevant to that particular material and light source.

BTW, if you think about how CGI and computer graphics have evolved they are concerned more about how good things look, rather than having an exact physics solution. So if they can take a shortcut that mostly convinces the viewer the scene is rendered realistically and it saves computational time for their purposes that may be good enough.

  • $\begingroup$ Uhmm I kinda disagree with this. Bc scattering and diffuse reflection are caused by the same thing: particles ~or less than lambda. So it is the same phenomena. $\endgroup$
    – Mah Neh
    Sep 18, 2022 at 9:21
  • $\begingroup$ Sort of. However, Scattering problems are usually something more specific and solvable. Surfaces can be complex - so diffuse reflection is kind of a catch all of the average phenomena. If you want to think of diffuse reflection to be the sum and interference of a lot of scattering events that is probably o.k. $\endgroup$
    – UVphoton
    Sep 18, 2022 at 14:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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