One more thing to consider when working from your head or extensively modifying reference:
Since most surfaces aren't evenly curved like spheres and cylinders, the specular highlights will usually occur on or in a corner (see muzzoid's example above). Being able to conceptualize forms as being made up of planes will help immeasurably in inventing or transposing lighting.
**Finished Work Thread **Process Thread **Edges Tutorial
Crash Course for Artists, Illustrators, and Cartoonists, NYC, the 2013 Edition!
"Work is more fun than fun."
"Art is supposed to punch you in the brain, and it's supposed to stay punched."
The short answers is no. It's because the angle of incident and reflection change with distance and position. Here is a diagram to show you what angle the planes will need to be in order to show a specular highlight. The plane's "normal" will always be in the middle of the angle of incident and reflection. Another way to check this is to grab a small hand held mirror. Turn on your studio light and hold the mirror in different locations yet turned so you can see your light in the mirror's reflection. You'll see that you have to keep changing the mirrors direction (normal) in order to see the light.
Agreed. The trends to the light source almost always fan through a range of angles in a picture. The only exception occurs when you have a light source that is both infinitely distant, so that the rays are parallel, and in the plane of the picture (such as, directly overhead), so that the rays do not converge to a vanishing point. And even then, the planes facing the light source will fan through a range of angles due to perspective.
If you apply the principle to planar objects you'll soon see why the highlights so often fall on edges, valleys and points - the greater the range of angles a surface curves through, the greater the chance that one will be the critical angle needed to catch the highlight.
Man this makes how ive been doing it feel primitive, id also like to thank you guys for the really really indepth answers. This is definately a thread to book mark for future reference. (and it is also one of those rare useful discussions in the art discussions part).
Ive learnt alot !
I was going to do some 3d models to demonstrate but i guess there is no real need any more .
Id like to add a little nitpick to the usual definition of speculars, as mirror-like reflections of the lightsource; they're in fact the mirror-like reflections of the brightest spot(s) present in the environement of a given object.
A real life example: One day my bathroom was mainly lit by a 2:00 pm blue sky with a couple of sunbeams coming through the corner of the window and falling on a little portion of a pure red towel that was hanging on the door.
In this situation that little patch of colored sunlight was so bright that you had a red squary little specular on all the reflective enough objects of the bathroom added to the expected specular made by the skylight through the window.
So three notes:
-there can be several speculars at the same time on a same object (watch shop owners definitely understood this aspect of specular reflection)
-objects that produce light arent the only ones that create speculars.
-speculars can alter the colors of an object in surprising ways and can be a realistic excuse to introduce out of scheme colors in a painting.
Finally i wanted to give a hopefully helpful link, shamelessly stolen from Farvus in an old thread.
Its a little freeware/software really easy to download, that generates face models based on the race, gender, age options you select.
This software should be of interest regarding specularity because, once open, in the second tab "view", you can opt out the skin texture and put a check in the box "shiny", wich gives you this:
You can then of course turn the head in every direction you want and see, as you do so, how the speculars will slide across the complex features of the head.
Obviously all speculars arent so sharp on a real face because not all the areas of the face are as glossy as the nose for instance, but its a useful resource nonetheless especially to understand how speculars follow expectable (or not) paths on form.
And yes, no matter how round and smooth a form is, a planar compromise of it can be found with some work.
Speaking of which, the same model set to flat diffuse shading mode:
(Not very "synthesistic" planes, i'll admit!)
I think you should also make studies from life, photos, etc because its a different but very important way of processing complex informations.
As far as im concerned, its always helped me remember things better.
Hope this helps!
And thanks to Farvus for the link.
Last edited by SM; April 29th, 2009 at 10:33 AM.
Since the angles required for specular highlights shift dramatically as the positions and distances change, I think it would be highly impractical to have to draw the light plane, facing-the-viewer plane, and find the middle of that every single position change. The formula works great as a means to estimate more accurately. As stated before, I think I'll take all this into consideration as a way to fudge it in a more convincing manner and I won't slice my wrists off if the tangent plane isn't absolutely perfect.
Above all, the most important thing I've taken out of this technical approach (not to dismiss the other useful info in this thread, however) is that specular highlights work off tangents. Though I sort of did it in the original picture, it was by fluke and by no means good estimates at all. Had I known this, I wouldn't have placed most of what I did in the original Joan of Arc picture.
Well, this became a great thread; better than I expected. I am fresh out of questions as they have been answered thoroughly. Thank you for all the contributions!
Maybe when you have put this stuff into practice you could post some examples back on this thread?
As Briggs requested, here's the new version of Joan of Arc has been working on. Please ignore everything (including the Italian mobster angel in the picture if you can...) but the lighting on the armour. Does it look metallic? Are my estimations for specular highlights pretty accurate or are there some missing or completely out of place?
yeah mate, great improvment.
Now your only problem is that you havent drawn in all of the bounce lighting that should be there, and there are some value problems, so i played around with a the curves a bit.
Also it has a bit to do with your rendering method, i think you are safe to start to move into textured brushes. Try out some ones that just put a little grain into your strokes, it helps remove the digital feel.
Sorry for reviving this thread but I thought of another conundrum related to this topic over the past few months. I figured it might be worthwhile for others to know. DavePalumbo's example with the clay ball made me think.
This question is speaking entirely based on the diffuse reflections of objects and not the specular highlights.
Let's say we have a red clay ball under a certain lighting. It has a nice gradient since it's a fairly dull surface. For the sake of simplicity, the brightness of the lightest part of its diffuse reflection (that is not the specular) is 75% brightness. If we were to polish that same ball so that it still exhibits diffuse reflection (obviously with better specularity), would the lightest part of the diffuse reflected area still be 75% or would it be higher?
This question spawned because there was a point where an object had the same local colour as skin but shinier. I didn't know if I should colour pick the skin and just add a more obvious specular highlight or if I should heighten the value of the diffuse reflection at the same time. It was also very hard to reproduce this exact scenario in real life.
Thank you in advance.
I'll post results later.
EDIT: Okay, I've done a quick test by rubbing vaseline on one side of my hand and leaving the other completely dry. I placed a light on top of the hand and made sure they have a region where both are hit by the same intensity of light. There were a lot more specular highlights on the vaseline-rubbed side but the diffuse reflections on both sides seemed to be the same value. I would assume this is the logical conclusion.
Why didn't I think of doing this!?
Last edited by Alex Chow; September 5th, 2009 at 11:34 AM.
This is why objects get darker when they are wet. Or woods get darker when they are varnished.
You can see that when the stones are wet they have more specular reflection but the diffuse is darker, this is especially obvious on the second image when the little remaining puddles are very dark.
The answer may depend on how you define "diffuse reflection".
I've been using the terms diffuse and specular myself, but this discussion is a good example of how the alternative terms body reflection (instead of diffuse) and surface reflection (instead of specular) might avoid possible confusion.
frog from itchy's rocks give off a rock-coloured body reflection and a sky-coloured surface reflection. When he says that the dry rocks give off a stronger diffuse reflection he is including both the body reflection and the "fuzzy" surface reflection, which the numerous microfacets scatter in all directions instead of a single angle of reflection. On the wet rocks this surface reflection is much more concentrated but not necessarily greater in total, in which case the underlying body reflection would be essentially unchanged. Similarly on Alex's vaseline-smeared hands, the surface sheen would be concentrated but the flesh-coloured body reflection would I think be unchanged. (However if in either case the surface reflection is not merely concentrated but actually increased, then there would have to be a slight corresponding decrease in the body reflection).
This is one area where I think just gathering reference and painting from observation a ton is helpful. You gain a intuitive feel for it after doing it a bunch that I think the pure academic discussion doesn't give you.
I think if you take two artists, one who understands the science of it, but who hasn't done studies yet, and another who has done a ton of studies but hasn't read the science behind it, the artist who has done the studies will probably have the more convincing final result. I know in the past I have ruined some art by over thinking the process.
More knowledge is always better than less, but I think there is a case to be made for just filling your brain with visual knowledge first, and the nuts and bolts part later.
wow so much info i never really thought about, i always went with feeling if the highlights were correct and not really going through all the calculating to see if it was correct.
thanks guys for opening my eyes, it goes to show that everything has an explanation and a purpose to why its there.
edit: hah jwillson i didn't read your post till i posted mine. I agree about what you said, and knowing how something works or why is there is also important i think.
Similarly, if a highly specular material is very rough (such as frosted glass or metal) the reflection is blurred, but the material does not become diffuse, it is still glossy even if the reflections are blurred.
Diffuse materials are made of particles that reflect light in a diffuse manner at the molecular level, the scattering is caused by the way the atoms in the material interact with photons.
frog from itchy, I agree with all of the facts that you state, but I still think that you are demonstrating how the term "diffuse reflection" can cause confusion. In your reply you use the term with purely the same meaning as body reflection. But the "diffuse reflection" coming off the dry rocks has two components, a sky-coloured surface reflection that could (on many rocks) be concentrated by polishing or varnishing, and a rock-coloured body reflection that could not. Think of a polished slab of granite, for example.
Some materials are made of atoms and molecules that, when they absorb photons, spit them out in a random direction - this is what causes diffuse reflection.
Other materials are made of atoms and molecules that when they interact with photons they reflect them at a specific angle (they absorb and re-emit them in fact), this is what creates specular reflection.
More complicated still, many materials have both properties, they will diffuse the majority of light that they reflect, but a fixed proportion of it will also be reflected directly - how this works is governed by quantum mechanics (specifically QED). These are the objects that will appear to get more glossy when polished, but they were always glossy to begin with (if only a little).
But to get back on topic, and also to clear up why this is important in terms of art, if a material is only capable of diffuse reflection then no amount of polishing will make it glossy. Think of oak for instance, you can polish it totally smooth and it will not have any specular reflection on it unless you add a coat of varnish - which of course adds a layer of glossy material on top of the wood.
So the dry stones in the photos above do not have a layer of sky reflection being diffused by surface irregularities, all of the light they reflect is being diffused by the atoms within those stones. It is only the layer of water that has this sky reflection, and that is only when they are wet.
Well, most rocks, like most substances except metals, are in your third category and have a light-source-coloured surface reflection and an object-coloured body reflection at the same time. When rough their surface reflection is seen as a faint "diffuse" sheen; when polished their surface reflection is seen as a concentrated "mirror like" reflection. My point is that this creates an ambiguity in the terms "diffuse" and "specular" reflection.
You're right, the dry stones in the second photo do have a faint and diffused sky reflection, and I was wrong to say that they don't, so they clearly fit in the third category.
Is this what causes the wet stones to be darker? I don't know but I'm going to try and look into it. My gut feeling is that the more matte or diffuse a surface is, the darker it gets when it's wet. Untreated wood for instance gets very dark when it is wet, whereas chrome is mostly unchanged.
Thanks frog D'Itchy, that's a relief!
I think that the usual explanation why a wet or varnished surface looks darker (and often also more colourful) is a double one. Yes, partly it's because the desaturating surface reflection that is diffused over the whole rough surface gets concentrated at small highlights, leaving just the body reflection to show through in between. The other reason is multiple reflection: light reflected off the surface, instead of proceeding straight to the observer, is often reflected back down again from the top boundary of the varnish/water layer, perhaps several times, so the surface gets more goes at absorbing whatever wavelengths it absorbs.
I just wanted to thank you all guys for these explanations. Something clicked in my head after racking my brains reading this thread. This is something I always thought too hard for me but you actually explained it perfectly with very understandable diagrams. Even though I now need to draw my ass off to practise it and learn how to render it well, at least I know where to place speculars, which means a lot. THANKS