|Color and Light||1.1||Do Assignment|
|Color and Light||1.2||Do Assignment||1.3 | 1.4|
|Illusion of Space and Atmosphere||1||Do Assignment|
|Personal Art||1.1||Do Assignment|
Specular highlighting is one of the few aspects that I believe cannot be answered by referencing and it's, not surprisingly, annoying me.
I've drawn still-lifes of metallic objects and I still cannot apply them in a manner that doesn't look like crap even with reference.
Why is referencing not the answer? Let's take it from the new CHOW with the topic "Joan of Arc" which, in most people's minds, would include metallic armour. I could Google "Medieval armour" and copy but that would mean I acknowledge the viewer, the shape of the armour, and the positioning of hte lighting to be exactly the same as the photograph with nothing left for the imagination to work with. And, not surprisingly, when I start changing any of the three factors even slightly, my specular highlights look completely misplaced.
I did read this and it's probably the most information I could get out of anywhere on this topic (this was pretty good too).
What runs through your mind when you decide upon specular locations? What do you ask yourself that tells you "Nope, that doesn't go there", "That should be there", "That specular is too large", "That specular is too small", "It's too strong", or "It's too weak"? Am I missing information that wasn't stated in the two pages I linked?
And attached, here's my self-rejected "Joan of Arc" draft rendered a little bit just for this thread (rejected because of bad design and a stiff pose). The second image is the material I'm trying to achieve; not as specular-ridden as some metals but takes in some diffuse lighting on top of speculars. An explanation on specular mistakes would be appreciated; a paintover would be a blessing but not expected (no, I won't submit an editted copy for CHOW).
Thank you in advance.
Last edited by Alex Chow; April 26th, 2009 at 03:50 PM.
I think you've pretty much understood the basics, and in fact the figure you've posted is a good start.
Specular highlights are nothing more than a reflection of a strong light source - in the case of the sun you will get very small and bright round highlights. On an overcast day the highlights will be less bright, and much broader and smoother, because they will be reflecting the whole sky. So these two different daylight conditions will give rise to very different kinds of highlights.
Take a look at the specular reflection on this foliage:
The broad highlights are a reflection of the overcast sky. On a sunny day you would get very small round highlights instead, with maybe some blue sky added too.
Highly reflective surfaces such as metals simply reflect their environment, so a suit of armour would reflect the sky mostly (or if it's indoors, the room around it).
The other important consideration is that most specular surfaces are far from being perfect mirrors, so the reflection is blurred and distorted by the imperfections in the surface - which is obvious in the photo you posted (the reflections here are from a photographer's softbox or similar soft light source, hence the broad white refection).
I would not take photo reference of real armor objects to study this.
That's hard because you probably try to get too much detail in the painting.
I have a nice example of some guy who painted a scene with armor in it a couple of decades ago: here
Try to understand what he did.
I've picked the largesst image I could find. Remember the original is roughly 230cm by 300cm, so even this image is small compared to that.
Title of the work: The Blinding of Samson. You can find out the rest yourself if you want to know more about it
hope this helps....
it doesnt really explain the theory behind it but this is probably the easiest way to think about it. Keep in mind that this isnt technically correct, and you should know that if you read that first link you posted, but using this simplified method you can get something that is much easier to replicate and for the most part will look pretty correct.
Last edited by Muz; April 27th, 2009 at 09:05 AM.
Direct reference only works if your light source is the same as the referenced light source. This is true for most things but metal in particular, as it's largely the reflection that defines the material and, if you're doing it right, the shape too.
For example, if you were to paint the armour helmet you posted into a daylight scene it would look like stone, because the lighting would be wrong.
I've no quick fixes but I recomment that instead of simply referencing you check out multiple references to actually understand why metals appear like they do. Think of where you light source(s) is / are, how bright they are, what colour the ambient light is, how reflective the surface is; does it have dents? Does it have rust? What surroundings will be reflected? etc. etc.
Hmm some interesting information, I should be reading more of it
I disagree with the people who advise against using photo refs.
Obviously the photo is only a means of understanding the theory involved, and assuming that you don't actually have a helmet at home to hold and rotate and study directly, photo refs are the next best thing. You're not looking at it to copy it exactly, but to see how light falls on that surface and apply that to your own painting. The only real differences between that surface and any other is the light will turn from highlight to shadow much quicker the shinier something is and have more dramatic extremes in the values (lighter highlights and darker shadows), and environment will show a more pronounced effect in color, bounced light, etc. (for example, the underside of the helmet is receiving alot of bounced light off the light colored surface that it's sitting on)
placement of highlights is actually no different from placement of highlights on a dull object, on skin, on anything else. The only difference is the intensity and the rate which it turns from light to shadow.
Psssssst! just use the DODGE tool!, dodge does highlights almost magically, I only use it for metal effects because it`s so easy, it looks bad on everything else.
The best reference I can think of (and nearly everyone has access to) is a car. Preferably a clean one that has a good somewhat reflecting paint on it.
This will give you a good idea of how little bends in the metal will alter the reflection of light. Also, it comes closer to the way an armor would look as the one in a reference picture, since most of the reflecion on armor is the environment and not the lightsource.
So just pick up the sketchbook or photocamera (I'm not against reference) and make a study of light on a car. I think it will help more than taking armor reference photo's which are mainly taken in artificial light.
Just my 2c.
if that very same page had not been linked to by two people already, among whom the OP himself.
A lot of replies. Thanks everyone. Just wanted to see if there was gonna be more before I reply.
Yes, to be clear, I did read http://www.huevaluechroma.com/021.php but maybe linking it on a tiny word may not have been the clearest way of saying it.
muzzoid's method is something I see some (see: quite a lot of) artists doing it (or fudging it in similar ways) but I also realized that it didn't fit into the technical definition of a specular highlight. I tried very hard not to "cheat" but I think somebody on this forum once said that if it looks good, it's right. This might be the case because just trying to place specular highlights the technical way is incomprehensibly difficult without 100% perfect reference.
I do have one more question...
Dave, can you clarify what you meant by this? As far as I know, a really really dull object would have different placements of highlights only because it would lack an evident specular highlight. A shiny object would show the specular highlight and, in most cases, it would not be at the centre of the diffuse highlight of the object. In that case, the shiny object would exhibit different placements of highlights than the dull one.Originally Posted by DavePalumbo"The only real differences between that surface and any other is the light will turn from highlight to shadow much quicker the shinier something is and have more dramatic extremes in the values (lighter highlights and darker shadows), and environment will show a more pronounced effect in color, bounced light, etc. (for example, the underside of the helmet is receiving alot of bounced light off the light colored surface that it's sitting on)...
...placement of highlights is actually no different from placement of highlights on a dull object, on skin, on anything else. The only difference is the intensity and the rate which it turns from light to shadow.
Last edited by Alex Chow; April 27th, 2009 at 08:48 PM.
the thing is, that the human eye doesnt know how a specular works on a static image, it only knows if it looks off when you are in motion. So in illustration you can get away with cheating. Your right, if it looks good that is all that matters.
But if you want to construct it peoperly you just have to do one thing. Draw a mental line that comes from the light, and work out where the right point is for that line to hit your eye. Again breaking it down into planes will help this.
if you really really want i can set up a 3d scene and do some animations on some objects to demonstrate how they work in relation to the position of the light?
No thanks. It's okay, though it would be very helpful for everyone including me if you have the time to do it. It's up to you really.
I've done tests myself in real-life already and noted the general pattern. I just wasn't satisfied after I realized how even the slightest changes can affect the specular highlights which made, for example, custom armour designs in a particular scene unbelievably difficult to comprehend if I go by the formal definition. The amount of variables to consider throws all my tests out the window unless I am a blacksmith who can create the damn armour right in front of me.
You're part right and I think I see where you're confused. In most cases, the highest intensity of the highlight (what would be the specular highlight on a shinny surface) is not dead center from our viewpoint. It is, however, always dead center from the viewpoint of the lightsource. This is because the light is hitting the object and then bouncing back so we can see it. If we were the light source, it would bounce back directly dead center. Since we're most likely not the light source, it comes to us on an angle and so appears to be off center. This doesn't make it work differently from a dull surface though, it just makes it more obvious when your placement is off.As far as I know, a really really dull object would have different placements of highlights only because it would lack an evident specular highlight. A shiny object would show the specular highlight and, in most cases, it would not be at the centre of the diffuse highlight of the object. In that case, the shiny object would exhibit different placements of highlights than the dull one.
Think of the light like banking a cue ball of the rail in billiards. When you hit the ball straight at the rail, it bounces straight back at you. If you were to hit from an angle, however, it would bounce off at an identical angle and continue moving. On a curving surface, this affects where we perceive the precise center of the highlight to be (which translates to the specular highlight in shinny objects). If you were to find the point exactly half-way between your eye and the lightsource and then draw a line which makes a perfect tangent to the surface of the object, this will be the location of the highlight's center/specular highlight. That's a little funny to wrap your head around though, and far to complicated to actually plot with mathematical precision, which is why we mostly just take an educated guess and if it looks right, we call it right.
Now to explain further how dull and shinny objects are really no different in placement of highlights...
Think about why textures appear the way that they do. Let's use for an example a ball of smooth but dry clay. Raw and untreated, it will have a soft diffusion of the highlight until reaching the terminator at which point there is no light hitting from the primary lightsource and it becomes the shadow area. The light dims across the surface as it reaches the terminator because most of that light is bouncing off along another angle which doesn't intersect with our eye. If we take that same ball of clay and put a glaze of gloss on it, suddenly the highlight intensifies and we get a specular highlight. (And remember, that specular highlight is in the center of the highlight from the vantage point of the lightsource, but will appear to be off center from our point of view.)
The reason is how light reacts to the surface. On a smooth and shinny surface, the light hits and reflects with little interference. Because of this, the highlight appear very bright. The light beam is striking and reflecting off, staying more or less concentrated and retaining most of its intensity. It's like when you angle the sunlight off of a flat shinny object and get a bright spot on the wall. Move the object and you move the spot, which is just a beam of light. Put that spot over your eye and it's blinding, because you've just made the angle perfect between your eye and the light source.
Now a rougher surface, in this case the dry clay, acts in exactly the same way. The difference is that on a very very very small scale, that surface is not smooth but actually covered in imperfections. What may feel and look like a reasonable smooth surface is actually made up of impossibly small bumps, each of which has its own little specular highlight and each of which is catching and reflecting light. The intensity of the light being reflected is weakened as micro-textural imperfections send it off into all random directions. Seen all at once, they give the appearance of a dull and smooth gradation from the intense center (where you're catching the most reflected light) to the weaker edges (where less is bouncing back at you).
Hopefully that's not too convoluted an explanation
Thank you everyone.
"There is no such thing as 'accurate drawing'. There is beautiful
drawing, and ugly, and nothing else." JAD Ingres, Ecrits sur l'art
Just a little diagram to illustrate Dave's point about the bouncing cue ball. It might help to (1) imagine a plane directly facing the light source, (2) imagine turning that plane on an axis [in the picture plane, and at right angles to the light fall] until the plane faces you, and then (3) imagine turning the plane exactly halfway back. Wherever that plane is at a tangent to the surface, that's where a highlight will be (exactly a tangent for a point light source; more broadly a tangent for a large light source). This will be at the same position as the brightest region of a pearly object (such as, well, a pearl), but not at the same position as the most brightly lit areas of a matte object, which occur where the surface directly faces the light source.
Last edited by briggsy@ashtons; April 28th, 2009 at 11:40 AM.
Something really "clicked" for me while reading Briggsy's post... thanks very much!
Last edited by CCThrom; April 28th, 2009 at 03:09 PM. Reason: disseminating bad information, sorry
Reading through this thread has taught me a lot, I thought I knew a lot about this, but I clearly didn't. Thanks everyone who posted understandable and clear information here on the subject; it's helped me, and many others learn! Many of the websites I have looked at have used way too many technical words, or rambled, so you miss what you were trying to learn in the first place.
It is the same, a highlight is a reflection of a lightsource but greatly diffused. The difference in appearance is just how the light bounces off of a surface (because of however rough or smooth it is on a miniature scale)Where that reflection of the light source shows up is related to, but not quite the same as the highlight
Thank you for the paintover and information, but I was slightly confused at this part. If you do not mind, can you please show me the three steps? I can't seem to replicate your results in other situations (not the Joan of Arc you've shown, but in another lighting situation), or I did but I don't think I'm doing it right.
Briggsy beat me to it but the eye's starting point is also a major factor in where the highlights will be. An example is if a child and an adult were looking at the same person in armor the highlights will be in different places to each of them. The horizon is line is the starting point (height from the ground plane) and plays a major factor in where hightlights will be. Here's my example using your character. I use the dodge and burn tools for this type of stuff. Helps get the values up to where they look right...then you can go in and paint over it.
And DON'T 'just use the dodge tool'. That nearly always looks incredibly cheesy.
You guys have been very helpful. I'll definitely redraw Joan of Arc (even though CHOW is over) and try to apply all of this. The answers are given in-depth and I feel ashamed to ask one more question in my mind. I'll do it anyways since we're this far
Taking everything into consideration, I would like to redirect this back to this quote by briggs...
Let's take this plane resulting from the three steps briggs has laid out and imagine a scene full of objects spanning from left to right, up and down. Not just one object, many objects spread out throughout the picture. Some are above the horizon, some are under. They are all shiny enough to exhibit a specular highlight and the lighting is perfectly consistent through all the objects.
For all the objects' specular highlights, will all of them share the same tangent plane from (3), even though positioning for each object is different? Or will I have to find the tangent plane for every object since the positions change?
If this is answered already through the paintovers and explanations, and it may very well have, then I apologize. I do feel an explicit answer will clinch my understanding of specular lighting without a question. If anybody is willing to answer this, even with yes or no, I will have no more questions... I promise!
Thank you in advance.
EDIT: Okay, I posted right after you posted briggs and didn't notice your post. Thanks for the diagram and all your help as it really solidifies my understanding of specular highlights. My question is still up in the air and it's really the only one I have left to ask. If it's answered, this thread can live on in my Bookmarks folder!
Last edited by Alex Chow; April 29th, 2009 at 12:20 AM.