Colours, I will strangle you!!

[JeffX99]

I'm not a woman either...but I sure LIKE'EM! (I didn't get that either...)

Trial and error is not the way to go about it either - observation, study, finding examples in other artist's work that appeals to you, etc. Color is a personal thing, much more so than value - value is an objective property while color is subjective and personal (which is why so much more is written regarding color theory and systems). In the end color is very maleable - but you need to observe and understand it to a certain degree before you can be expressive with it.

Edit: Cross post - yeah, it is a good discussion, and always of interest.

[Kweckduck]

Alright, progress report. I started out with reading Fletcher's Colour Control and his method of choosing colours is very interesting. I tried it out on the sketch I had done the day before yesterday and had more success than I did before. Which doesn't mean it was good, but at least it wasn't awful. But around page 46 he completely lost me when he starts talking about the intensity of colorus and such. He uses names of colours that are unknown to me and without any visual examples I would have to look up every colour and try and decode what he is trying to say. So at this point I'll switch over to the books I got from the library and see how far those get me.

I'll also do the values first from now on and do a colour underpainting afterwards in an attempt to simplify or at least not confuse myself with what is colour and what is value.
I'll take it to my sketchbook from here. Glad I was able to start an interesting discussion.

@Micronomicon: Somehow I find colorpicking to be a way of cheating. I don't want to be dependent on other people's pictures in order to pick my colours. Of course I do realise that I might have to forgo my principles at first in order to develop such skill. Still, I'll avoid it as long as I can.

@Brigssy: Thanks for the links. I'll be sure to give them a proper look-over. I think I might have come across it at some point, but the graphs looked completely beyond my realm of understanding that it scared me away. Since there seems to be a need for it, I'll start carrying asperines with me at all times.

@Dpaint, Jeff, Dile: I understand that painting from life is the best, since it's unfiltered in a sense. But I also understand the need for using photo reference. Drawing out in the open would feel like information-overload to me, I don't think I would be able to handle it very well in terms of frustration. I can't even say how many times I've smashed my pencil into my sketchpad after trying to get the same thing on paper again and again and again. Form and value have a certain element of physics to them, but colour just seems so elusive to me. A picture seems easier to study to get the basics down. Moving outside to finetune the basic rules you learned from the photos. Observation from life can be done constantly to catch things that cameras can't.

@ JJacks, Queen: What what? *Innocent smile*

[sanya]

http://www.itchstudios.com/psg/art_tut.htm
this site might help. scroll to the stuff about skin tones. add areas of higher saturation to make your colors pop.

Wow, that is a fucking RAD tutorial! Thanks for sharing!

[briggsy@ashtons]

Color is a personal thing, much more so than value - value is an objective property while color is subjective and personal (which is why so much more is written regarding color theory and systems).

This is a remarkable discovery, Jeffx, and I'd like you to tell me more about it. Am I correct then that of the three Munsell dimensions of object colour, value is objective but hue and chroma are not objective?

[JeffX99]

I'm saying that color can be a very subjective statement and interpretation whereas value cannot. The value scale is, pardon the pun, black and white (of course everything between) - the color spectrum is nearly infinite because it is like the value scale but raised to some high power.

What I'm getting at is if you get the values wrong, you're screwed - there is much more freedom of expression with color however. Which is basically why so much is written and studied regarding color and light phenomena but little can be debated about values.

Hope that clears up my point?

[m8y]

How can one hope to strangle that one has no "grasp" of?

Hehe, reminded me of this line from the World of Warcraft southpark episode, when they were talking about the obese basement dweller. "How do you kill that which has no life?".

As an example of what Jeff is saying about how our eyes/brain are much more sensitive to value than to colour .. this is why chroma subsampling is such an effective image compression tool especially for video. You can effectively halve or even quarter the colour information without too much perceptible degradation.

[JeffX99]

Briggsy...did that make sense? Of course all of the slots in the Munsell color space can be quantified and numbered and are thus "objective". But that has little to do with interpretation and expression either when painting from observation or imagination.

Maybe I can say it a different way...value can be separated from color and is much more finite - color is nearly infinite as well as being much more individual due to the very nature of ocular physiology (the distribution of cones across the retina - which varies between individuals and even between the sexes).

Edit: m8y - that's pretty interesting - I never thought of that aspect of color/value and compression.

[dpaint]

This is a remarkable discovery, Jeffx, and I'd like you to tell me more about it. Am I correct then that of the three Munsell dimensions of object colour, value is objective but hue and chroma are not objective?

I think what Jeff is trying to say is that hue and chroma are easier to design within the framework of an image than value. Most veteran artists will tell you that color doesn't matter. What they usually mean by that is hue and to a lesser extent chroma don't matter when you are painting. Even Value doesn'ty really matter when painting; designing those attribute do matter though.

An example is academic painting, no one with competent observational skill would say that chiaroscuro effects accur in natural settings, yet they look convincing when painted by someone who can design their lights and darks in a believable way. In real life shadows are rarely black or opaque and if you actually saw things that way you would run to the doctor thinking you were going blind.

It is even more so with the other two attributes of color, there are a range of believable hues for skin tones that when painted with knowledge look completely real yet when isolated would send you running for that doctor again if your skin really was that color.
What was it that Eugene Delacroix said? “I can paint you the skin of Venus with mud, provided you let me surround it as I will.”

Accuracy such as Munsell is a good basis for observational representational painting skills and I recommend it, but it is only a beginning; especially when trying to capture outdoor effects with the landscape or a model.

[briggsy@ashtons]

I'm saying that color can be a very subjective statement and interpretation whereas value cannot. The value scale is, pardon the pun, black and white (of course everything between) - the color spectrum is nearly infinite because it is like the value scale but raised to some high power.

OK, so value is one dimensional and colour is three dimensional. But, as dpaint says, both are open to subjective treatment for expressive purposes. Values can be varied expressively within their one dimension in a way that seems very similar to the way colour can be varied expressively within its three dimensions. (In fact, value is arguably the most important of the three dimensions by which colours can be expressively varied.) You can treat your values in a style that is high-key, low-key, high-contrast, low-contrast, etc. One of these tonal treatments might have maximum visual fidelity to the subject, but the same could be said for colour. So the leap to objective vs subjective still perplexes me!

It's interesting that although value seems to be a straightforward physical property, this is actually an illusion (middle grey is not mid way in physical reflectance between white and black); value, like hue and chroma, is a psychophysical property that is objective only in relation to a defined "standard human observer" under defined conditions, which you could say means that all three are subjective. For example, no value scale looks evenly spaced on all backgrounds; the formula for the value scale used in Munsell and L*a*b* is actually a fudge that was derived by fitting a curve to the average of the very different scales obtained experimentally on different backgrounds.

[kev ferrara]

By a quick calculation based on some recent personal study, I would guess that there are only about 5000 or so truly distinct "natural color schemes" which evoke a particular aesthetic response in humans population wide. (Values schemes are included as part of the color schemes.)

I would guess that there are another 1000 schemes or so that would classify as "cartoony" that would work in animation but might be a tad too colorful to still work as "realism." These would be on the borderline between truth and falsity.

I also would guess that there are tens of thousands of color schemes which do not evoke a particular response in humans because they have a confused or physically-false expression. For these types of plans, there might be a wide difference of reaction, with those with a naturally naive viewpoint finding more connection, while others who might have more sensitivity to the actual appearance of the natural world might find nothing to respond to.

EDIT: These numbers are ballpark guesses and the only point of them is to assert that the number of color plans is limited to a manageable number because they fall into categories. Just like the number of colortones is actually infinite, but fall snugly into 12 hues and 10 values or so.

[JeffX99]

OK, so value is one dimensional and colour is three dimensional. But, as dpaint says, both are open to subjective treatment for expressive purposes. Values can be varied expressively within their one dimension in a way that seems very similar to the way colour can be varied expressively within its three dimensions. (In fact, value is arguably the most important of the three dimensions by which colours can be expressively varied.) You can treat your values in a style that is high-key, low-key, high-contrast, low-contrast, etc. One of these tonal treatments might have maximum visual fidelity to the subject, but the same could be said for colour. So the leap to objective vs subjective still perplexes me!

Yes - you're absolutely right - value is also somewhat subjective but to a much lesser degree than color, simply because it is "one dimensional" vs. the much broader "dimensions of color" and their interactions. I was speaking in more general terms relative to the limitations of value vs. color. I tend to make that leap just as a simplification when I teach because it seems easier for people to grasp.

A useful analogy might be comparing value to the figure...if you're going for naturalism/realism you pretty much have to get things right - not a lot of room for interpretation vs. a tree in the landscape being like color...much more room for design, interpretation, scale, placement, etc.

Value is definitely the most important of the dimensions of color - we see value first and probably more consistently between people vs. color which is much more an individual perception. That said though, I think color tends to be the more expressive and emotional of the dimensions whereas value carries more accuracy/form information. Though in dpaint's insightful example of chirascuro (and it's more extremme form of tenebrism) there is certainly a great deal of emotional impact carried via value with little color. Interesting...
Hmmm...mood/emotion can certainly be carried by value as well, as in Isaac Levitan's limited value landscapes for example (actually I just scanned a bunch of Levitan - I always had the impression of limited key but in fact he seems to use limited values but at both ends of the range - almost in a high contrast way).

I guess my point is you can definitely treat values with some expression as mentioned, but if you get them out of balance within that construct, (high-key, low-key, etc.) things fall apart in a hurry - thus the more "objective" nature of value.

Again, great discussion - thanks for the reply!

[briggsy@ashtons]

I guess I'll have to be satisfied with "simplification", but for the reasons given in my last post I would have hoped for "oversimplification" at least!

Everything depends on what you are trying to achieve, so to that extent I agree that colour is subjective. If you aim for a vivid sense of light and atmosphere, then you need to build in certain abstract relationships between colours that are the basis of our perception of these things. These relationships can be established accurately from life either by pure observation or by observation assisted by theoretical understanding, and they can be maintained among modified or imagined colours either by careful visual judgment or by the latter assisted by theoretical understanding. If one values other things than these (or if one has what Kev just called a "naturally naive viewpoint"!), then one is much freer to treat colour in an arbitrary fashion.

[JeffX99]

Yeah - I'll give you that it may be an oversimplification of theory but, in my experience it is a more direct approach and easier for communication. Maybe I'm just not sophisticated enough but when I'm out there painting my main concerns are composition, value (light) and capturing a fleeting moment or effect of nature.

I think that the artists I've studied under, and admire throughout history have, not necessarily a "naive" approach to color but one that is "innate" (maybe a better word than naive). For the most part they work from observation in what I would say is a "natural" manner but even then they tend to push color or subdue it to suit their statement. So I guess I'm saying they tend to build in the sense of light and atmosphere more through observation and awareness than theory.

[Robotus]

By a quick calculation based on some recent personal study, I would guess that there are only about 5000 or so truly distinct "natural color schemes" which evoke a particular aesthetic response in humans population wide. (Values schemes are included as part of the color schemes.)

I would guess that there are another 1000 schemes or so that would classify as "cartoony" that would work in animation but might be a tad too colorful to still work as "realism." These would be on the borderline between truth and falsity.

I also would guess that there are tens of thousands of color schemes which do not evoke a particular response in humans because they have a confused or physically-false expression. For these types of plans, there might be a wide difference of reaction, with those with a naturally naive viewpoint finding more connection, while others who might have more sensitivity to the actual appearance of the natural world might find nothing to respond to.

I would love to know the process by which you came to that conclusion.

[kev ferrara]

I would love to know the process by which you came to that conclusion.

Lots and lots of comparative analysis and theory.

EDIT: To clarify, I guessed.

[briggsy@ashtons]

Jeffx88, I just added the italicized sections to my last post in the light of yours. Better?

[JeffX99]

Yeah briggsy, that's better! Almost makes it sound like color is subjective!

LOL - I'm just razzin ya - it's been a great discussion!

Edit: Hey! Who's JeffX88?!

[OmenSpirits]

Yeah briggsy, that's better! Almost makes it sound like color is subjective!

LOL - I'm just razzin ya - it's been a great discussion!

Edit: Hey! Who's JeffX88?!

evil twin.

[briggsy@ashtons]

Edit: Hey! Who's JeffX88?!

Whoops! I mean ... just testing your "observation". You got a pass on resubmission.

[Kamber Parrk]

Hey briggsy@ashtons

I've recently read James Gurney's Color and Light and, I'm in the process of working my way through your "The Dimensions Of Colour" site.

I'm pleasantly disturbed to learn from you that much of what I've been learning from my big stacks of how-to library books on (mostly for me) watercolor painting is wrong.

Sorry in advance if I haven't yet uncovered an answer already on your site(s), but could you recommend a good "short list" of how-to type books that present a practical application of the theory that you have presented to us?

(I think somewhere along the line you mentioned a painter named Stephen Quiller. Any others?)

Also, I have a downloaded copy of Chevreul's The Principles Of Harmony and Contrast Of Colours. Winslow Homer considered this work to be "his Bible." I'm a big fan of Homer. Do you consider this work to still be worth studying in the 21st Century?

[briggsy@ashtons]

Kamber Parrk

My intention with the Dimensions of Colour site was to give people the basic information they need to develop an active, analytical, problem-solving approach to colour, so I'm hoping that once you've finished absorbing it and James' book, your passion for "how to" books on colour will have cooled a little!

For basic watercolour painting technique, you've probably read enough already, but you might like to also explore some of the older manuals available free on archive.org. For advanced technique in ANY medium the best way to learn is to copy works, or parts of works, by artists you admire. For Homer, it can be very difficult to determine his technique visually, but fortunately there have been quite a few studies of his painting procedure. These are mostly in catalogues of major Homer exhibitions, which also contain great illustrations for copying. The most recent, which I suspect you already have, is this one, which I'm sure will reference the earlier ones:

Watercolors by Winslow Homer: The Color of Light (Art Institute of Chicago)
http://www.amazon.com/Watercolors-Wi.../dp/0300119453

Chevreul's importance regarding colour is sometimes exaggerated because of his historical role in communicating the ideas of contrast and afterimages to artists, including Homer and the impressionists, and perhaps also because he was one of the last scientists to advocate the naive (red-yellow-blue) colour wheel. (His work on the chemistry of fats however is awesome!). His Priciples of Harmony and Contrast preceded the Helmholtz-Maxwell revolution in the study of colour of the 1850's-60's, and he never updated his understanding. A really great book that communicated the implications of that revolution for artists is Modern Chromatics (1879) by the American physicist and amateur painter Ogden Rood. If you're willing to work your way through books from that era, that's definitely the one to start with:

http://www.archive.org/details/cu31924031167889

[Kamber Parrk]

briggsy@ashtons

Thanks! [Kinda figured you were more of tools not rules sort of guy.]

I do have Watercolors By Winslow Homer: The Color Of Light checked out of the local library. (Probably going to be fodder for a master study or two, but there is an extensive analysis of Homer's materials and techniques that I'm plowing through for the moment).

Thanks for steering me towards the Rood book! (Would really hate to dedicate too much time to wrestling with the wrong book of arcane 19th Century color theory just out of some sort of Homerian hero worship).

Just an aside: that U. of Wisconsin cadaver dissection you have linked through your site is utterly fascinating!

[Kweckduck]

So after starting with Fletcher's color control, which I found had something new to teach about how to choose colors, but lost me half-way through, I started reading Itten's "The Elements of Colour".

So far I haven't really come across anything earth shockingly new. Sure, I learned how the theory of complimentary colors came from the splitting of white light and experimenting with that, but in essence the theory behind complimentary colors isn't new. Since I'm reading it in Finnish it's slow going and haven't gotten incredibly far yet, so I'm looking forward to the rest of it.

But while reading some of the things that I had come across elsewhere and some other things some questions came to mind:

1. He uses examples like grey on a red background seemingly having elements of green, and on a green background seemingly having red elements. As I understand it this has to do with our minds looking to balance out the colours with the complimentary colours.
My question therefor is, if this is so (for I didn't really see it in the examples myself) why aren't shadows just shades of grey?
According to Itten, if you mix opposite pigments you will get grey. I haven't tried this myself, but I'll take his word on it. But why then have I always read that you ought to mix the opposite/complimentary color into the shadow? In my experience this would mean the shadow on a red ball would at least be slightly greenly tinted. Or were these people always talking about pigment/paint and did it actually mean I needed to move towards grey in the shade?

I don't know if that made any sense to anyone. The whole thing has got me properly confused.

2. He also states at some point that harmony of colors is achieved by having colours of the same darkness/lightness. But how do you define form without dark/light contrast? I most be missing something here, because I cannot see how this can be true.

If anyone can help me to the path of understanding, it would be much appreciated.

[Elwell]

Itten's mystical mumbo-jumbo has little to offer the representational artist, IMO. What little actual color science it was based on is outdated, and his continued influence is more a result of the mystique/legend of the Bauhaus than anything else.

[Kweckduck]

[dose]

Unfortunately, Elwell speaks the truth...

[JeffX99]

1. He uses examples like grey on a red background seemingly having elements of green, and on a green background seemingly having red elements. As I understand it this has to do with our minds looking to balance out the colours with the complimentary colours.

My question therefor is, if this is so (for I didn't really see it in the examples myself) why aren't shadows just shades of grey?

The whole thing has got me properly confused.

If anyone can help me to the path of understanding, it would be much appreciated.

This is true to some extent - but in what is essentially a "laboratory" experiment - or in some minimal op-art. The real world basically obliterates such subtle effects with its complex interactions of light, texture, form, etc. (Well, in landscape anyway...the effects may be more noticeable in still life and portraiture). I have seen this effect in nature as well though - but in certain circumstances - golden rolling hills in summer for example can be very high chroma (saturation) in afternoon/evening light, causing the shadows to take on an even stronger complement - but this was more of an "impression" by the eye/brain to "balance" the complements. Of course as an artist, interpreting these effects you can use them to your advantage.

Shadows aren't shades of grey because they are influenced by the ambient light in the scene - whatever colour it may be.

It can be confusing to try to understand the theory and technical aspects of colour without the practice, observation and interpretation from life. A really, really helpful device is to cut/punch a small hole in an index card to isolate colours - you hold it up to your eye so that only the colour passage you want to identify can be seen - this helps determine what the colour really is - and is often a big surprise!

Hope that helps some! Oh and, just for general purposes - for colour the kind of "compliment" is "complementary".

[duztman]

A really, really helpful device is to cut/punch a small hole in an index card to isolate colours - you hold it up to your eye so that only the colour passage you want to identify can be seen - this helps determine what the colour really is - and is often a big surprise!

to expand on this, you can make an index card divided in three portions, black, grey and white, with a hole in each. this can help isolate very dark or very bright colours by help of contrast.

but to second what every sane person has already pointed out: if you want to learn colours, there is no remedy but to landscape. reading books is not the answer, you need to program your neural pathways by actually performing the task.

[AndreasM]

I think the underlying principles governing all this theory is as follows:

Get the relationships right. Governed by drawing, is Value, Hue and Chroma. In that order. As soon as you are trying to balance these relationships in a picture, you are in fact composing. When "realism" is your goal, color and form are easily subordinated to tone, which is governed by drawing.

In working directly from nature, that's pretty much all you need to know...you have to trust your eyes and spend less time assigning ideas to the things you see.

Theory, I think, is most useful when you don't intend to work directly from nature, but need a solid reference that can act as a substitute to the optical world.

Kweckduck,
Make it simple man, you are just starting out. Restrict the range of information in the beginning and add more complications when you master the underlying principles. To be honest, I think you'd be better off working on your understanding of value, through observation. It's tricky enough to key your values in black and white, and drawing well at the same time. When you add color to the equation, it get's messy fast.

[Dile_]

In that order.

Value > Chroma > Hue. "Hue doesn't matter as much" as dpaint (I think?) pointed out. Chroma on the other hand does. Maybe we are thinking of different things though (since it'd be hard to have chroma without hue).