The Dimensions of Colour - a colour theory discussion thread

[briggsy@ashtons]

The Dimensions of Colour
www.huevaluechroma.com

This started out well over a year ago as a tutorial intended for CA, but it just refused to stop growing. Eventually I took up an offer to have it made into a website as a second year student project at Billy Blue, a graphic design school where I teach drawing. But it's still meant for the community here, and I hope a bunch of you get a chance to work through it and maybe give me some feedback here. Huge thanks also to sciboy for generously hosting it on his server, where it's been sitting for the last week while I tinker with it. Still tinkering, but I think I can finally bear the thought of you guys seeing it now....


EDIT (Dec 11): I'd love for this thread to develop into a place where people come to discuss questions about colour theory. So please add any comments or questions you have, whether in connection with something on the site, or in relation to your own work, or about colour theory in general.

[darkwolfb87]

Thank you, this looks wonderful. I'll get you some feedback as soon as I can.

[Elwell]

David, I'm only about a quarter in, but this is fantastic. Have you posted it at Rational Painting?

[Farvus]

Looks like tons of helpful knowledge. It's really hard to find this kind of websites on the net so thanks a lot. I'll try to read it soon.

[Elwell]

OK, just got through the whole thing, it's great. You'll want to give it another copyediting pass, there are a few typos and coding problems still, but it's a fantastic resource.

[Sonaj]

That is very helpful and a great read, thank you

sticky?

[Elwell]

Oh, and I made this a sticky.

[thinairart]

Awesome, some really great info you've put together! The points you bring up regarding diffuse and specular reflection are well written and on target.

[briggsy@ashtons]

Thanks Tristan - very proud of my first sticky! I've just finished adding the corrections to the text for Part 10, which I didn't quite get onto last night. (Didn't imagine anyone would get through the whole thing in one sitting!) I'll do another pass through the whole thing soon, but at the moment I'd probably add more typos than I fixed. I did post it on Rational Painting just before turning in, but it looks like they're still thinking about it.

Thanks everyone for your comments so far. reidaj - I'm not sure but you may have the impression that there's more 3d rendering on the site than there is. The only image that used a 3d programme is the flash animation Ray made to show a light source moving around a red ball, for which he used Maya. It looks pretty right to me when you allow for the fact that he has used a fairly close light source. All of the other spheres were painted from scratch in Photoshop.

If anyone sees any problems caused by the code please post or PM me telling me where they are and what browser you're using - I 'll see if I can get someone who knows what they're doing to help.

[thinairart]

sorry about that Briggsy, my post turned out to be more confusing then I meant it to be

I was reminded of 3d rendering by the appearence of the spheres with specular highlights, and mistaking those illustrations for 3d images, that lead me to make an unrelated comment about the use of 3d rendering as a guide to the modeling of light. I edited my original comments to keep my post on topic.

Excellent guide, glad it has been made sticky.

[romance]

Cheers for this! Brilliant!

[Flynt]

Fantastic---I am so pleased to see someone put this together including citations.

Great!

[briggsy@ashtons]

Thanks for the comments, guys. Flynt, it would be great one day to see Dimensions of Colour among the links on that gorgeous site of yours.

[seth1]

Thanks so much for taking your time to donate this to us. I will be reading through it. Cant wait to get started.

[talbot]

Thank you for this wonderful resource! There looks to be a wealth of information here. Can't wait to get started. Thanks again!



P.S. You spelled "color" wrong.

[Jack the R]

Thanks to Briggsy, this has helped correct a few things I was told wrong, and has got me thinking of painting more of as an act of construction.

Regarding Principle 1, the shading series, shouldn't the hue of the shadow shift some in regard to the light? Is there a way to calculate the shift in hue?

Also -

Also note that, contrary to a widespread myth among painters (e.g. Loomis, 1947), the richest colour is not on the edges of the lighted area.

I think this term "richest" is new, are you referring to brightness or chroma? Either way, if the "richest" area isn't on the edge of the lighted area, would it be the full light area?

[briggsy@ashtons]

Thanks for the responses guys, and especially for the questions, Jack the R. I'd love for this thread to develop into a place where people discuss questions arising from the site, or any questions about colour really.

shouldn't the hue of the shadow shift some in regard to the light?

Colours in the shadow zone very often do shift in hue compared to the light, but not necessarily, and certainly not because they are shadows. If the colours do shift in hue, it is because the hue of the dominant light source in the shadow zone is different to the hue of the primary light source. In addition, if the primary light source is distinctly coloured, the light source in the shadow region appears to shift towards the additive complimentary of that colour, by an instance of simultaneous contrast. This effect is rather inadequately summed up by the artists' rule of warm lights giving cool shadows and vice versa. In the discussion of "Principle One" I am just taking the simplest case first, where the primary and secondary lights are both the same colour, and perceived as white. For the more complex, and common, situation of lights of different hue, you need to take this in conjunction with the principles dealing with coloured light sources and the effects of multiple light sources.

I think this term "richest" is new, are you referring to brightness or chroma? Either way, if the "richest" area isn't on the edge of the lighted area, would it be the full light area?

Thanks. I think I got "richest" from the way I have heard it expressed verbally many times, in a context that implies highest chroma. Actually Loomis (1947, pp.152-3) variously refers to "brighter", "the most brilliant and pure", and the "most intense" colour being located at the edge of the lighted area, but it is clear from his diagram that he also means higher chroma but not higher lightness in the terminology I have been using.

In answer to your second question, the chroma of the diffuse reflection keeps increasing up to the full light (as shown in Figure 10.1), but this colour may be desaturated (1) by the addition of white or different coloured light at the highlight (which may be quite broad and "fuzzy"), and (2) by specular reflection of the environment, especially on the receding planes. So the actual situation is quite variable.

If you look through either of Idiot Apathy's Peer Project threads you will clearly see the difference between the spheres done by guys who used shading sequences based on the uniform saturation principle, as I recommended here, and others who worked on the theory of putting the highest chroma in the halflight (directly or indirectly following the recommendation of Loomis, I believe).

[Farvus]

In answer to your second question, the chroma of the diffuse reflection keeps increasing up to the full light (as shown in Figure 10.1), but this colour may be desaturated (1) by the addition of white or different coloured light at the highlight (which may be quite broad and "fuzzy"), and (2) by specular reflection of the environment, especially on the receding planes. So the actual situation is quite variable.

I have read more interesting parts from your website but one thing is still confusing to me. Sorry if this was already answered but I'm still not entirely sure. I read everywhere that shadow side is usually less saturated but when I look at different paintings, it's actually the opposite. (especially when it comes to pale skin).

Is it beacause this broad and weak part of highlight extends and dominates almost whole part of light side? Where it doesn't get pale there is real color of object and sometimes it can be only found at the edge of shadow. If it's true then this color close to higlight's white must be even less saturated than shadow side which creates the the opposite result?

At the same time maybe it could be lots of bounce light that saturates the shadow. Propably beacause there is bright surface around like in this painting. Shadow on her cheek gets some light from chest -http://artrenewal.org/images/artists...ne_Milbank.jpg

Basicly I don't see any mounting point when I want to know what closer to object's natural color - it could be shadow or light.

[Jack the R]

Thanks for the reply Briggsy. I'm going back through your site again so I can understand it. Give me another day . . .

What struck me most is the implication of principle 1 for glazing tonal paintings in a way that is technically accurate and realistic looking - and if you can do that, wouldn't it, by virtue of speed, be the superior painting technique? It caught my fancy anyway - I'd like to paint comics, and I need as much speed as I can get. I haven't managed to pull off the glazing process in a way that looks good yet, but I'm working on it.

[briggsy@ashtons]

Thanks for the question Farvus. Certainly the shadow colours in the portrait image you tried to link to are much higher in chroma than the uniform saturation relationship would predict, but do they actually look realistic to you? Apparently even ARC-approved artists are not obliged to copy colours precisely as they appear.

Below I've modelled what I think is a reasonably realistic-looking sphere out of Mrs Milbank's skin. Most of the shadow zone is represented with a low-chroma colour that I think does look like our heroine's skin in a less bright light. However, as you discuss, where the shadow zone receives light from other areas of skin we do get colours more saturated than the main series, because Mrs Milbank has coloured this light twice.

The dreary looking lass on the left of the same ARC page on the other hand is very obedient to the uniform saturation rule. I'm not for a microsecond suggesting that that makes it a better painting, but I think you'll agree that it does present a more realistic visual appearance.

[Farvus]

Thanks a lot. I think I showed bad example beacause it does look a bit unrealistic. Also talking about skin can be hard beacause it's not consistent "material". There are parts on a face tinted by blood vessels and so on.

I looked at this girl on the left. There are parts that look like it's obvious - for example on the neck and on her right arm. However some parts like on her shoulder blade or face get more pale and beacause of that the saturation drops down. These parts are close to the bone so maybe the skin gets more shiny and there is stronger highlight ? .

[briggsy@ashtons]

These parts are close to the bone so maybe the skin gets more shiny and there is stronger highlight ? .

Not so much because it is on the bone, I think, but because the surface in those two places is at the critical angle needed to bounce the specular reflection to our eye. Of course, the bone does in each case create the plane break that causes the surface to pass through that critical angle.


In the diagram below, you can see that colours from the girl's back and neck, excluding the highlight on the shoulder blade and the shadow area on the arm, lie along a line of uniform saturation (blue line). The highlight is less saturated because it approaches the colour of the light source. The shadow on the arm lies on a line of lower saturation than the colours in the light because of a difference in hue between the main light and the secondary light (showing up in the shift from creamy light to bluish shadow in the dress).

[Farvus]

Fantastic diagram. Now I see that. It's sometimes hard too see the same occurrence in everything beacause every object has different material properties. I think I'll need to play more with 3d studio max to see these things more clearly.

Thanks for taking your time to prepare and explain these things. I still don't have much knowledge about color but I'll try to join discussion if anything interesting comes out.

[fotojb]

Thank you!

[briggsy@ashtons]

I posted the image below on a thread started by andymania, but am reposting it here with a quote of his original question because the latter is such a CLASSIC example of how trying to follow an artists' rule of thumb can be far more confusing than a little rational analysis. It's also a great example of the confusion caused by the use of the terms "warm" and "cool" in the popular vague sense that could refer either to a difference of hue or or a difference of chroma. The lighting shown is actually the simplest lighting of all, where the main light and the secondary light are both white. The underlying diffuse reflection (i.e. skin colour) therefore follows a single line of uniform saturation and steadily increasing chroma from shadow into light, as shown by the blue line in the attachment. (If the last sentence makes no sense to you, read this page). As long the desaturated lighter colours are saved for the specular highlights they shouldn't look chalky.

OK Help me out on this one. Look at photo first.

Here is a photo of my friend in my living room that has a neutral cloudy day light source (I think this light source is cool rather than warm). What temperature is the light area and what temperature is the shadow? It is very hard for me to establish this.

What I do know:

Value: Light area is an average of 15-20% gray and shadow area is around 40%

Chroma: I realize that these colors are in a middle chroma range. That means not much intensity in the colors.

Light source temp: It is (I think) on the cool side but just a bit.

Local color: His flesh is slightly on the warmer side and the hues are not very saturated however I have a cool light on it so Im a little confused as to how I should interpret this temperature wise.

Overall Temperature: Light: warm???
Shadow cool??? But the degree of coolness is very little.

Now, what friggin colors should I use to block this out??? I am afraid of using too much white in the light area, since I am afraid of chalkiness and too much desturation.

Any advice greatly appreciated.

-Andy

Now my biggest problem

[briggsy@ashtons]

What struck me most is the implication of principle 1 for glazing tonal paintings in a way that is technically accurate and realistic looking - and if you can do that, wouldn't it, by virtue of speed, be the superior painting technique? It caught my fancy anyway - I'd like to paint comics, and I need as much speed as I can get. I haven't managed to pull off the glazing process in a way that looks good yet, but I'm working on it.

It appears to me that the difficulty is that, while glazing in theory produces colours in the right relationships, when you create these colours using a greyscale layer under a coloured layer, our visual system tends to recognize this and (unfortunately) to see it just that way. This tendency is very hard to avoid completely in traditional paints. Taking great care to keep the glazes to the exact areas where they are needed, and perhaps introducing a bit of hue and thickness variation into them, may help. Perhaps someone who uses glazes more extensively than I do can add some advice here.

[HunterKiller_]

Exceptional wealth of information. Much thanks to you, sir.
I better take it slowly or my brains gonna melt.

[Earendil]

I'm going to have a lot to process in the next days/months/years. This site is fantastic, I'm so glad you put it in your signature! Excellent, excellent, EXCELLENT!

[PelleTM]

Fantastic site, it will take some time to digest. If you want some feedback and are interested in improving upon the site, all I can say is try to add as many pictures and explainatory "animations" as you can - these help tremendously! Also perhaps add a glossary? (Couldn't find one). Again, thanks a lot. I also appreciate your other links.

[Giorge]

Waw this s&^% is fascinating , i bookmarked this page. Need to study it bather