a somewhat more advanced lighting question.

[battlebattle]

ive been doing alot of studies lately and i think my technical skills are finally at a point where i dont have to fear white canvas anymore. ive been doing alot of practice without lines just raw painting as forms and i think its greatly helped my understanding of light and its relation to forms.

my new problem is that when doing my own paintings im having trouble "doing" without "thinking" when it comes to the subject of how all the colors of an object change depending on the light of everything around it.

say for example you have a blue ball and place it into a red box and light the box from the top with a yellowish.white light bulb.
the bounce light of the red walls would appear to make the ball more purpleish instead of its true blue color. the reflective light from the bottom would light the balls bottom with a little more bright red because the saturation and brightness mixed with the proximity of the light to the ball would sort of bypass the new mixed purpleish color the ball has taken on. the top would be red tented lighter vauled and warmer colored version of the purple.

but how would you approach this on a larger scale than a controlled study? for example in this picture

say i wanted to paint a figure into the foreground near the rocks facing the ocean who was wearing a very pure red cloak, with the sky being so much further away than a ball in a box does the sky still effect the overall hue of the red that far to the ground? i assume his cloak in that scene appear more burgundy/maroon in color, but how do you calculate by how much? or is the sky even that heavy of a factor at all or would the majority of the lighting would come from the warmer orange of the sun and the bounce light from the water/ground/purple rocks behind the figure? im sure the sky would play a part to some degree but with all these factors when trying make a composition for a entire scene including props and figures ive overthought my brain to a point where i feel like i have no idea how to approach this anymore.

any discussion,advise,reading,demos, or video tutorials would be highly welcomed.

thanks for reading my long ramble =\

[JeffX99]

These aren't really things you can "calculate"...there is no "correct answer"...and once you had one answer the scene would have changed anyway, perhaps significantly. So, what to do?

This is where experience and hundreds of paintings in many situations of many subjects comes into play. While you can't "calculate" a solution what you can do is orchestrate the scene, light and color so that it looks accurate (if that is what you're after) and makes the statement you want to make. In order to be able to do that effectively you just have to go through the training and develop the experience.

And yes, the sky can be a major factor in lighting a scene...or not...it just depends. You're beginning to encounter the vast gulf between colored spheres in boxes theory and the real world. There is no substitute for painting outdoor scenes and environments, you just need to get out there and do it if you want to understand what is happening with the color and light.

[Bowlin]

It sounds like you need to find more information. Have you tried finding figures in similar lighting situations? (the most I immediately see are very dark)

https://www.google.com/search?hl=en&client=firefox-a&sa=N&tbo=d&rls=org.mozilla:en-US:official&tbs=simg:CAESZRpjCxCo1NgEGgIIBwwLELCMp wgaPAo6CAESFIIE-wOdBKQE-QOBBP0DnAT8A5UHGiC1nSBJybCYWPWWh5eripT7lXbPW3mXc6e UmPeqlsnFhwwLEI6u_1ggaCgoICAESBDge4scM&tbm=isch&ei =DU3tUOS5IJHa8ASl54HICw&ved=0CGAQsw4&biw=1600&bih= 1033

[Elwell]

I agree and disagree with Jeff. Yes, experience is a big part of solution, yes, there is no such thing as a "correct" answer beyond what works in the context of the picture. But, you can indeed "calculate" these things, to a degree, with the proper tools and information. So, to take your example, you have the information you need to answer your questions within the picture you've posted. If you want to paint a figure near those rocks, the figure will be under the same lighting conditions as those rocks. What are those lighting conditions? Well, we have two light sources, the sun (direct, from the left) and the sky (diffuse and from above). The sun in very low in the sky, so its color is shifted way towards orange-red, and it's overall strength is rather low. The light from the sky is purple-blue, and effects every up- or side-facing plane that doesn't directly face the sun, or has the sun's light blocked from it. Since both light sources are so far away (effectively infinite), their distance doesn't come into play, and there will be little fall-off (inverse square law). Try to determine the local colors of the objects in your scene, and analyze how those colors have been shifted by the lighting conditions. In the case of the foreground rocks, they're probably a fairly light, fairly neutral gray (the background rocks are a different, darker sort of stone). Any other object under those conditions will have its light and shadow colors shifted to the same degree.

[battlebattle]

thanks for the responses guys. i should have mentioned the examples i used were just to explain the concepts in my head i was struggling with. im not actually using that picture for anything i just picked it because it was a great example of atmosphere,bounce light, and weird lighting drasticallyaltering the overall color of what was probably a tan rock or light brown rock.

i paint 100% digitally so painting landscapes from life is almost impossible, but i have been painting way more still life images. as of posting this have found colored paper in my house im going to use to try changing bounce/lamp light. but i do value studies from life with pencil and paper way more now.

ive been doing a still life painting as a warm up then doing environmental speed painting studies. as of recently right after doing a landscape study trying to recreate a similar landscape from imagination trying to capture the same vibe as the study i just did.

[Elwell]

Generalities are collections of specifics. That's what (good) reference is for. You asked a question about a specific example, but the thought process I explained is universally applicable.

[JeffX99]

I agree and disagree with Jeff. Yes, experience is a big part of solution, yes, there is no such thing as a "correct" answer beyond what works in the context of the picture. But, you can indeed "calculate" these things, to a degree, with the proper tools and information. So, to take your example, you have the information you need to answer your questions within the picture you've posted. If you want to paint a figure near those rocks, the figure will be under the same lighting conditions as those rocks. What are those lighting conditions? Well, we have two light sources, the sun (direct, from the left) and the sky (diffuse and from above).

Absolutely. Good point in bold. Just to clarify when I mentioned "orchestrating" the scene I was referring to having the experience to see, understand and use the information at hand - the "proper tools". Anyway, talking about the same stuff really.

[JeffX99]

ive been doing a still life painting as a warm up then doing environmental speed painting studies. as of recently right after doing a landscape study trying to recreate a similar landscape from imagination trying to capture the same vibe as the study i just did.

You probably won't want to hear this but imo this approach is a waste of time. I can't imagine anything less direct than painting landscapes digitally from photos then trying to do the same from imagination. Landscapes are right outside your door, if you really have an interest in environments and landscape then study from the source.

[Bowlin]

Here is a good tutorial on modeling.

http://androidarts.com/art_tut.htm

[Bowlin]

I agree and disagree with Jeff. Yes, experience is a big part of solution, yes, there is no such thing as a "correct" answer beyond what works in the context of the picture. But, you can indeed "calculate" these things, to a degree, with the proper tools and information. So, to take your example, you have the information you need to answer your questions within the picture you've posted. If you want to paint a figure near those rocks, the figure will be under the same lighting conditions as those rocks. What are those lighting conditions? Well, we have two light sources, the sun (direct, from the left) and the sky (diffuse and from above). The sun in very low in the sky, so its color is shifted way towards orange-red, and it's overall strength is rather low. The light from the sky is purple-blue, and effects every up- or side-facing plane that doesn't directly face the sun, or has the sun's light blocked from it. Since both light sources are so far away (effectively infinite), their distance doesn't come into play, and there will be little fall-off (inverse square law). Try to determine the local colors of the objects in your scene, and analyze how those colors have been shifted by the lighting conditions. In the case of the foreground rocks, they're probably a fairly light, fairly neutral gray (the background rocks are a different, darker sort of stone). Any other object under those conditions will have its light and shadow colors shifted to the same degree.

I've seen you make comments in previous threads that students often don't have ENOUGH reference material. But in this thread your encouraging just one found reference picture is enough for all your modeling and rendering needs? Is this a common method for artist, to just use one picture? I keep thinking maybe I'm going about this the all wrong ways.

[JeffX99]

What Tristan is saying is every scene, especially outdoor scenes, has its own lighting conditions that dictate how you handle adding any elements into the scene. I was saying basically the same, with the addition that outdoors the light is constantly changing as well. Ultimately you just need to understand the basics well enough to make a consistent, believable image (if you're going for a realistic image).

Of course one image only informs you of one lighting situation.

[Bowlin]

Oh... I was thinking more about the method involved. If it was one found copy right image that needed to be changed and drawing a figure from memory. And then trying to figure out the local color of the rocks and the local color of the figures skin and clothes, there just doesn't seem to be enough information to really go on. Like the OP says how do you calculate exactly what the lighting and shadow colors are going to be? What are the tools you use to calculate it? There's the Loomis scale separation chart that you can use to help guess values for the lighting, but I can't find anything for colors. It just looks like you have to mix a color, see if it looks right in the scene, if not, mix another one till the new local color looks like it belongs. Then ball park the light and shadow colors onto that local color. That's pretty much what the OP originally suggested.

[JeffX99]

Yes - that was my point - no calculations, just experience and knowing what works to achieve what you want. That isn't found in charts and such. Again you're thinking in absolutes and "exact" answers, "right way" etc. No two artists will paint, even from the photo, the same statement. Give it to ten and you'll get ten interpretations - if they're all competent then all ten will be good solutions. Give it to a 100 artists...same thing. In fact, most experienced artists select, decide and work from an essentially infinite set of possible solutions, all which may be right. The trick is to see and work toward the one you are most in the mood to execute at the time. These possibilities are multiplied infinitely when working outdoors.

[vineris]

Like the OP says how do you calculate exactly what the lighting and shadow colors are going to be? What are the tools you use to calculate it?

Experience. You go out and observe how different colours look under different lighting conditions, and from that you formulate some rules. Once you have some rules derived from observation, you can look at a photo and extrapolate that under these lighting conditions a red cloak will be shifted X amount darker and Y amount more blue. Sometimes you will come across information that helps you clarify those rules.

It just looks like you have to mix a color, see if it looks right in the scene, if not, mix another one till the new local color looks like it belongs.

This only works if you have enough experience to know what looks right in a scene. And most newbies don't, because they don't spend nearly enough time observing stuff and trying to derive rules from what they observe.

[Bowlin]

Give it to ten and you'll get ten interpretations - if they're all competent then all ten will be good solutions.

So I guess just say to the OP that any method you use it's going to be a stylized color composition?

[JeffX99]

So I guess just say to the OP that any method you use it's going to be a stylized color composition?

That is exactly what a painting is don't you think?

[briggsy@ashtons]

Hopefully our ten artists would come up with ten different interpretations even if our red-cloaked gentleman was right there in the scene. Modern colour theory won't tell you how you must paint him, but as Elwell described, combined with experience it will let you "calculate" what he would look like in that light.

To comment on just one point raised by battlebattle, a blue ball under a red light, or a red cloak under a blue light, won't necessarily result in a purplish reflection. The process is one of subtractive mixing, and so depends on the actual spectra of object reflectance and light involved. A red object that reflects hardly any blue (e.g. cadmium red) would mainly just be comparatively dark under blue skylight.

[Bowlin]

I was thinking more about the actual method involved in determining how to make a figure with a red cloak look like it belonged in that scene. Is the figure drawn from imagination/memory? Or does the OP want to use a photograph of a figure? Either method she took, she might not come up with cadmium red as the cloak's local color. And then how much does she stylize her paintings or how naturalistic does she try to be?

If she was trying to be a Matte Painter, how does she go about "calculating" the right color for a cadmium red cloak, other than a trained eye? What are the actual tools and calculations involved? I'm not saying it can't be done, just don't know how.

Besides, I don't know of any artist that starts with just one landscape picture and tries to paint a figure from out of their head without it being heavily stylized. There's just not enough information for the figure to look VERY naturalistic. Who alive does this, that we know for sure?

[briggsy@ashtons]

What are the actual tools and calculations involved? I'm not saying it can't be done, just don't know how.

Well if you follow the process of analysis that Elwell outlined, and have understood the basic principles of colour and light, for example from here, you should have a pretty good start! I suggest you begin with simple volumes like spheres or cubes before trying to tackle something as complex as a figure though. At least that's what I give students in workshops as an exercise after a couple of days of classes, and they seem to manage very well. Here's an example I posted here ages ago:

[JeffX99]

I was thinking more about the actual method involved in determining how to make a figure with a red cloak look like it belonged in that scene. Is the figure drawn from imagination/memory?

If she was trying to be a Matte Painter, how does she go about "calculating" the right color for a cadmium red cloak, other than a trained eye? What are the actual tools and calculations involved? I'm not saying it can't be done, just don't know how.

Besides, I don't know of any artist that starts with just one landscape picture and tries to paint a figure from out of their head without it being heavily stylized. There's just not enough information for the figure to look VERY naturalistic. Who alive does this, that we know for sure?

The figure would be painted into the scene based on experience and a trained eye.

Again "calculation" is the wrong term to use because it implies two things: 1) that there is one correct solution 2) that there is somehow a formula that can be followed by anyone, regardless of experience to reach a correct solution. These two concepts are simply N/A when it comes to art.

Maybe these analogies will help...how does a writer "calculate" the resolution of conflict in a story? How does a musician "calculate" the next phrasing in a piece of music? The potential solutions for these scenarios are fluid and solved based on experience, intuition and intent.

[Bowlin]

Well if you follow the process of analysis that Elwell outlined, and have understood the basic principles of colour and light, for example from here, you should have a pretty good start! I suggest you begin with simple volumes like spheres or cubes before trying to tackle something as complex as a figure though. At least that's what I give students in workshops as an exercise after a couple of days of classes, and they seem to manage very well. Here's an example I posted here ages ago:

Yeah, that does sound like studying theory along with a trained eye.

(briggsy and Jeff) Well maybe "calculations" wasn't a good word to really bring into the discussion? But the more I think about it, there HAS to be some simple way of determining colors for other objects. Because your making a guess of the color of the light and we know it goes from warm to cool from each light source. Perhaps you can determine other colors like in this chart?

[Elwell]

Sort of, but it's more useful to slice or compress the color space horizontally rather than "unwrap" it. That way you can shift the center (neutral point) towards the color of the light source, and see how the other hues will shift (note that this only works on a modern color wheel that's based on visual complements).

[briggsy@ashtons]

But the more I think about it, there HAS to be some simple way of determining colors for other objects. Because your making a guess of the color of the light and we know it goes from warm to cool from each light source. Perhaps you can determine other colors like in this chart?

Sure. In terms of simplistic traditional colour theory you could come up with an answer of sorts by thinking of hue in isolation, and applying "rules" framed in terms of "warm" and "cool" and so on.

Alternatively in terms of modern colour theory you would understand the effect as a shift in colour space, involving chroma and value as well as hue, and you could use Photoshop layer modes as a "colour calculator" to estimate what these shifts would be (once you understand how different layer modes emulate different types of colour mixing). Please see this page:

http://www.huevaluechroma.com/104.php

[Bowlin]

Sure. In terms of simplistic traditional colour theory you could come up with an answer of sorts by thinking of hue in isolation, and applying "rules" framed in terms of "warm" and "cool" and so on.

Well I was thinking of my idea while using the Loomis separation chart, first to see how many value separations there are. Then determining the local color. But yeah, I wouldn't have any way of determining chroma with this chart. And I understand that some colors the hue shifts and value shifts wouldn't be in a straight line toward the light, it would end up being very stylized. But at least it's "some" kind of simple guide to help eyeballing it?

(I'll try figuring out the photoshop layer modes when I have more time, if I can! Your site always overwhelms me with all the different variables you have to understand)

Elwell - I can not figure out what your saying. Using the Yurmby wheel without unwrapping it? How do you slice or compress color space without unwrapping it??

[briggsy@ashtons]

OK then, if you want to keep it simple for now, at least remember that objects opposite in hue to the light tend to appear darker, as well as more neutral. When you get into it in more detail you'll see which colours tend to "funnel in" towards the neutral point, and which are prone to show more of a hue shift.

I'm sure Elwell just meant you should visualize the colour shifts on the "colour wheel" view of colour space, i.e. looking down on the hue circle, like in my diagram 10.8 (bottom right) and 10.9, rather than on the "unwrapped" outer surface of colour space that you used.

[Elwell]

Yes, exactly.