Technique Tuesday - Specular vs. Diffuse Reflections

Hey everybody, let me start by saying thanks for the "Rave" reviews for my Lighting Class over at Kelby Training! As with anything though they are always questions that arise from the viewers. Today I want to discuss "Specular" vs. "Diffuse" reflections. In the first lesson the question was raised about my remarks regarding how the size of the light source affects the specularity on the skin.

Here is an excerpt from the question I received, "I was a little confused on one thing. David seems to be making the point that the size of the light source will affect the degree to specular reflection on the skin. ... I read “Light: Science and Magic” a few months ago, and my understanding is that reflections or glare on the skin like this is more a function of the texture of the skin rather than the size of the light source. The size of the light source will affect the harshness or softness of the shadows, but skin glare is something you can control only though the angle of incidence, polarization (of the light source or a filter on the lens), or changing the degree to which the surface diffuses the reflection (i.e. powder makeup, etc.). ... I’m just not seeing how the size of the light source would control the amount of glare coming off the brides face."

OK, here we go and hang on as it might at first seem a bit complicated ----

Light: Science and Magic: is a great book - I own it myself, but we need to look beyond the absolutes of "angle of incidence" = "angle of reflectance". Sure it's true, but a person's face is composed of what I call a bunch of "micro surfaces" all oriented in a different direction. That's why things like make up cut the shininess of the skin. It's an application of even more "micros-surfaces" - so small in fact that there seems to be no shininess at all, but under a microscope, you would see the same behavior. So what we are discussing here is truly a diffused highlight as opposed to a specular highlight.

So now let's define a diffused highlight. It's a highlight that when a beam of light hits the diffused surface, the beam is reflected off into any number of directions randomly. Point a flashlight at a white board at any angle and you always get a diffused light return. Sure the shape of the light pattern is different as you rotate through the angles of incidence, but the angle of reflectance is still random because of the orientation of the "micro-surfaces."

Now let's discuss "Specular" reflections. Do the same experiment, but this time point the flashlight at a mirror. Notice that now you can see the reflection of the flash light quite plainly as you move through the angles of incidence. It is only when you are perpendicular with the mirror surface that we get nearly complete return of the light to our eyes - read that as specular return on the subject's face.

What we are trying to accomplish in our portrait lighting technique is bring the density of the "specular" highlight more in line - density wise - with the diffuse" highlight. In other words, we want detail in the speculars. Maybe this is where I wasn't clear in my video tutorial. Please read on. So how can we do this? It's simple, by using a larger light source and by bringing it closer to the subject.

But how does a larger light source reduce specularity? Let's do a little "mind exercise". Suppose you had a mirror 8x8 feet large in a room (our specular surface) and at the bottom right hand corner of the mirror was a white card (our diffuse surface.) Next I want to illuminate that surface with a light source. I choose my 200 W.S. Quantum, pointed directly at the mirror going through a 1x1 foot or a 1 square foot soft-box and I fire away.

Now let's take a spot meter reading of the 1 square foot hot spot on the mirror. For the sake of discussion, let's say it meters F64. Now do the same for the white card. Remember, the white card is giving us a much less efficient return to the meter since the light is being spread in all directions by those randomly oriented "micro-surfaces". Hence, the meter reading from the card will be substantially less - let's say F8.

OK, this is where it's gets a bit trickier. Now, lets say I manage to increase the size of my soft-box to 2x2 feet or a 4 square feet area. Now lets fire the strobe and take our measurement. The still intense specular reflection for the strobe would be what? It's easy the calculate - we have the same amount of light (200 W.S.) spread over 4x the area so the intensity has to measure 1/4 the original reading or two stops less - F32. The white card on the other hand still sees 200W.S of light and gives us F8.

Now, let's do this exercise one more time. Let's increase the size of our 200W.S. light box to 4x4 feet or 16 square feet of area. Let's fire away again and take our measurements. What do we get? Using the same reasoning as before we can easily calculate the new "brightness level" of our 16 square foot light source will be diminished by another 2 stops or now reading F16. What's our card read - still F8 - since it is still seeing 200W.S. of light.

So what's happening here is that the larger light source, as it spreads out over the specular surface is being reduced proportionally as it's size increases. Conclusion - the larger the light source relative to the subject reduces the intensity of the specularity in a very controllable fashion. In fact, the ratio of "specular" to "diffuse" highlight can be controlled to a point of maintaining detail in the specular highlight. Pretty cool, when you understand the concept and know that we have a lot of control over our finished image.

I learned this from my friend Dean Collins in 1983 when I took a very informative week long class from him. It's for this reason that I choose large light sources most of the time for my wedding images. It all goes back to Dean's explanation of the "specular" vs. "diffuse" highlights. I did a quick "google" search on Dean Collins and specular highlights and come up with this very thorough article over at the Strobist site discussing the same concepts with a slightly different perspective. It's well illustrated and really worth the read - lot's more good info there. Here is the Stobist link.

I hope this explains my comment on the tutorial and again, thanks for watching.