Question on what Richard Schmid wrote in Alla Prima

[andymania]

In Alla Prima, Richard wrote (and im paraphrasing)....."a 40 degree change in form can be shown with a simple color change rather than a value change." Why? Because we have more colors at our disposal than values..." So my question is: Any subtle changes in value within lights, halftones and darks can be indicated with just a shift in hue of the same value? I did notice it is extremely hard to capture many values with paint as opposed to a B+W medium like charcoal or graphite.

[velderia]

That's interesting.

[andymania]

Yeah i found that interesting too. I think I can finally just stick to no more than 5 values rather than trying to capture all 255 greys...lol:yayca:

[Smarty]

yer, i think as long as there is a perceived change in the subject it kind of pieces together. Implying more subtlety in the values. It's really to simplify your values to make a stronger composition. Also i think the hue change probably works best shifting towards the hue of the shadows.

[dpaint]

If you look at most master paintings the values are limited to four or five steps so the way you turn the form in the lights or in the shadows is with hue or saturation shifts, saving the value shifts for big divisions between light and shadow

[andymania]

but within those 4/5 steps (many master paintings as we know are very detailed) any other subtle value changes (assuming no glazing is done) are done by hue/temperature shift right? Oh wait just reread your post dpaint. ok i get it.

[andymania]

ok i get it.

[dpaint]

Its not about details. Look at them in black and white and see how many places in the painting are the same Values. Good paintings start with good value plans color doesn't matter.

[briggsy@ashtons]

I'd love to know if it is literally true that it is possible to make a form turn effectively without a value change. This notion also came up recently at Rational Painting (http://rationalpainting.org/index.php?topic=2547.0), where the idea met with scepticism, despite being apparently promoted by David Leffel and Nelson Shanks as well as Schmid. In Schmid's (rather gorgeous) paintings the details of the form are actually obtained by fine tonal modelling, as far as I can see (see attachments).

[dpaint]

Here is an example of what I mean. These paintings are by John Asaro
When you see them in gray scale there is nothing off about them. But look at the color choices he made are not natural at all except in the values

[dpaint]

So the point is that a good value plan will carry the painting and you can shift your color to almost anything you want as long as the underlying values are sound. So the color of a cape on a figure or their skin color doesn't matter and your choices can now have more impact then using natural color. Where people go wrong is they focus on the other aspects of color like hue and saturation but the values are off and that detracts from the strength of a painting usually.
Velasquez painted with red yellow and two blacks and white , Ansders Zorn, red yellow black and white. Corot painted landscapes without the benefit of modern greens and blues. Those paintings strengths come from accurate drawing and strong value plans they manipulate reality to conform to the artists vision.

[Baron Impossible]

So the point is that a good value plan will carry the painting and you can shift your color to almost anything you want as long as the underlying values are sound. So the color of a cape on a figure or their skin color doesn't matter and your choices can now have more impact then using natural color. Where people go wrong is they focus on the other aspects of color like hue and saturation but the values are off and that detracts from the strength of a painting usually.

That's true. When painting digitally I'll often arrange my colours by value, then quickly pick dependant on value alone (broadly speaking). The hue variance combined with underlying values then produces a more painterly effect. Of course it doesn't work so well with wildly differing saturated hues if you're aiming for realistic rendering but for mild variances I find it improves the output.

[andymania]

AHHHH..... Briggsy.....it seems that Scmid uses more than just 4-5 values as he claims. I see more in those b+w conversions.

[Elwell]

In Alla Prima, Richard wrote (and im paraphrasing)....."a 40 degree change in form can be shown with a simple color change rather than a value change." Why? Because we have more colors at our disposal than values..."
I'd like to know the direct quote rather than your paraphrase, to see if that's actually what he says. The other thing is that, while it's a very good book, Schmid isn't always very precise in his color terminology. The fact is, changing "color" (that is to say, hue and/or chroma) without also changing value is very hard to do with an open palette like Schmid uses.

[dpaint]

The values are much closer when you show them in grayscale. I just copied your image and changed the mode

[briggsy@ashtons]

andymania and dpaint

I used a grayscale conversion plus some automatic segmentation software to make it easier to see where the tonal gradients are occurring. The gradients are there but the number of steps is arbitrary. The segmentation software also seems to have increased the absolute contrast, although much less than you might possibly think from close up - try looking at the last image from more than ten feet away.

[andymania]

Elwell here is the exact quote:

"It is uneccessary to use all values in a subject. Color changes can frequently be used instead. This substitution of color for value is often not only more pleasing, but also makes more sense. Why? Because we have many more colors at our disposal, so using a color saves a value. Superb examples of that use of color are found in the paintings of Mary Cassatt. She had the uncanny ability to portray form with almost no perceptible darkening-a feat impossible without color."

-I kind of question the last sentence...convey form without any darkening??? How is that possible?

[Kraus]

So lets say we have a pair of tits (oYo). The closer the surface depth-wise to your eye, the lighter it would be. Say we make the tits yellow... According to Schmid i would be able to use orange of exactly the same value for the surface that is rounding itself into cleavage and make it look depthy without actually making anything darker?.. Hmm, makes logical sense.. Orange seems 'darker' than yellow, i guess...

Also, is it possible to perceive form through our perception of depth alone, or does our depth perception solely depends on light bouncing..

[MatejaPetkovic]

andymania and dpaint

I used a grayscale conversion plus some automatic segmentation software to make it easier to see where the tonal gradients are occurring. The gradients are there but the number of steps is arbitrary. The segmentation software also seems to have increased the absolute contrast, although much less than you might possibly think from close up - try looking at the last image from more than ten feet away.

I am interested in that "automatic segmentation" software, can you tell me name of it?

Thanks in advance

M

[PieterV]

So lets say we have a pair of tits (oYo). The closer the surface depth-wise to your eye, the lighter it would be.

Value changes depend on the angle of light as in relation to the angle of the surface and the viewing angle.
Distance between viewer and surface only becomes valid at large distances as the light loses brightness trough the atmosphere.


Say we make the tits yellow... According to Schmid i would be able to use orange of exactly the same value for the surface that is rounding itself into cleavage and make it look depthy without actually making anything darker?.. Hmm, makes logical sense.. Orange seems 'darker' than yellow, i guess...

If it seems darker, it is darker.
In the computer world there a differences in 'brightness', 'lightness' and 'value' (hsb, hsl, hsv)
High chroma orange is darker then high chroma yellow. This is because colors reach their maximum saturation at different values. So you cannot say that you're going to use orange of the same value that looks darker, value in paintings has to be determined empirically.


Also, is it possible to perceive form through our perception of depth alone, or does our depth perception solely depends on light bouncing..

It could be done with atmospheric perspective, linear perspective, stacked perspective, (etc), edges, line-thickness, stereoscopic images...

[Flake]

Guessing that it's Image > adjustments > Posterize in Photoshop

Slider lets you choose how many levels.

You may get better results by converting to grayscale and doing any contrast adjustments before you go to 2 or 3 levels.

[briggsy@ashtons]

I am interested in that "automatic segmentation" software, can you tell me name of it?

Thanks in advance

M

Posterize does something quite similar, but I used an ancient Photoshop plugin called Mask Warrior.

http://www.imagiam.com/index.php?option=com_content&task=view&id=21&Itemid=58
http://www.imagiam.com/index.php?option=com_content&task=view&id=21&Itemid=58&limit=1&limitstart=1

[Line]

What Schmid is suggesting is that you can look at color alteration as another tool to your arsenal, in your attempt to create volume. There are instances where you look at something and you realise that that large, for example, forehead, is one grey tone, yet you see it curve. Instead of going for a solution via tone (and in cases of subjects that are a big 'grey' thing, it's tough to render the halftone), you can use color temperature or change the color completely. It's a way to create contrast.

An excersice I had done that opened my eyes to this (before getting Schmid's excellent book) was to paint a cast, that was literally washed with soft lights. With the exception of some deep indentations, the whole thing looked like a grey spot, against and equally lit and grey spot (the wall behind it). I was using a complementary color scheme. I managed to paint it by using the change in hue/temperature and suggest volume even in areas that were the same tone. I'm no master at it, but as a first attempt, it was an eye opener and showed me that there is more to painting than adding white or black.

Try it and see for yourself, then Schmid's suggestion will make sense.

[BigJoe14]

I'd like to know the direct quote rather than your paraphrase, to see if that's actually what he says. The other thing is that, while it's a very good book, Schmid isn't always very precise in his color terminology. The fact is, changing "color" (that is to say, hue and/or chroma) without also changing value is very hard to do with an open palette like Schmid uses.

From the Values chapter, page 85-87:

"Conservation of Values
The Masters who maintained simple value patterns in their paintings seldom used more than five values (except in the transition zones and soft edged between shapes). You can see this dramatically in black and white reproductions of works by Howard Pyle, Serov, Vandyke, and others. They were stingy with the number of tones they used and never employed more than were necessary. In many of his portraits, Sargent usually employed only three values in the light, two in the darks, and then added some incidental highlights and dark accents.

This economy or conservation of values is based on two ideas. The first is that a few clear-cut values in a painting will yield a more powerful visual effect (though not necessarily a more "artistic" one) than a profusion of small values. This is why Impressionistic painting, which as a rule pays little attention to strong value patterns, is not as effective in monochrome reproduction as it is in full color (Impressionism by its nature is concerned with other effects.)

The second idea is that it is unnecessary to use all values in a subject. Color changes can frequently be used instead. This substitution of color for value is often not only more pleasing, but also makes better sense. Why? Because we have many more colors than values at our disposal, so using a color instead saves us a value. Superb examples of that use of color are found in the paintings of Mary Cassatt. She had the uncanny ability to portray form with almost no perceptible darkening-a feat impossible without color.

Manet and Berthe Morisot were also good at that. They understood the weakness in overmodeling-the use of too many values to indicate form. They were careful to keep their designs strong by maintaining simple values-clearly establishing their major areas as belonging in either the light or the dark, and not invading those areas with needless or inappropriate values.

Fortunately, we have an unlimited supply of colors and ways to use them. For example, in areas of gradual darkening caused by contours, a turn in the form on a subject of as much as 40 degrees can be shown with color changes alone before a change in value becomes necessary. My personal guideline is to always check if changing the color temperature of my mixtures will do the job before I change my values.

Value relationships are certainly not all this cut and dried, and there is certainly room to make choices about emphasizing or restraining them to meet your artistic intentions, but keeping your values uncomplicated and few in number is a sound idea no matter what creative changes you do or do not make. It allows you to simplify the way you look at a subject and render it. I also believe-without question-that it is the basis of strong color and design."

[Graydon]

So what he is saying is that using a restricted value pattern is the key to strong paintings.

If this is the case, then I agree partly. However, I have seen full valued paintings that are quite strong, such as those by David, Ingres, Bouguereau and Bargue.

If he is saying that color can turn form, then I am not so sure. We would have to prove that the value of each hue or chroma change is identical. On the other hand, I have seen chroma and hue changes reinforce modeling, given a logical value structure. Both change according to the color of the light source, the proximity of the object to the viewer and any reflected light.

[Serpian]

Elwell: What is an open palette? What would the opposite be like?

A note on the question of value composition (not so much on colour): Couldn't you say that the idea that great paintings use large flat areas of similar value only partially true? I mean, if you compare the pictures below, yes, they all have a solid value structure that could be simplified to a few large value steps. But if you compare Sargent's portrait of Carolus-Duran and Hanna Pauli's portrait of Venny Soldan (both late 19th century) with Caravaggio's Taking of Christ and Caracci's Assumption (both Baroque), the amount of modelling if very different, and creates two very different moods or feels. One is rather flat, in an attractive way (I just loove the flatness of Carolus-Duran's jacket), the other is much more modelled and heavy on the chiaroscuro. One is much more naturalistic in one way, giving a simple-to-read quick impression of reality, the other is much more naturalistic in another way, giving us information of almost every tendon and muscle, through much more detailed and rounded modelling.

http://www.jssgallery.org/Paintings/Carolus-Duran/Carolus-Duran.jpg
http://www.uni-koblenz.de/~vladimir/konst/venny.jpg
http://www.artrenewal.org/artwork/589/589/3755/taking_of_christ-large.jpg
http://www.artrenewal.org/artwork/630/630/4035/assumption_of_the_virgin_mary-large.jpg

[Graydon]

I should have hue and chroma alone, without value. Color comprises all three.

It seems that in the late 19th century, some artists preferred lighting that would flatten the form and emphasize local changes.

[andymania]

I tried doing a study this weekend. Instead of changing values, i changed hues while keeping chroma and value consistent. it does absolutely nothing. just flat patches of color,

[andymania]

In that portrait of Carolus Duran, i see way more than 4-5 values. I keep hearing painters saying, like Schmid, that the key is to keep it 4-5 values. Then after you analyze their final painting, it ends up having 10. Simple blocking in might be ok with 4-5 values, but if you want some more detail,more value changes within big value changes, you need more.4-5 dont cut it.

[Anid Maro]

A long quote by Schmid regarding color and value...

This is actually a matter I'd been putzing around with on my own for quite some time. I've for a while been fascinated by the fact that each hue has its own distinct greyscale value (the technical term being, I think, luminosity) and have developed a lil' color wheel that shows approximately where each of 12 hues would fall in value when converted to greyscale (via luminosity in GIMP).

http://i134.photobucket.com/albums/q96/Anid_Maro/ValueBasedColorWheel.png

This wheel is reinforced by Dpaint's (http://www.conceptart.org/forums/showpost.php?p=2569410&postcount=10) posting of John Asaro's painting. However, I do not think either my wheel nor Asaro's painting are demonstrating the point Schmid is trying to get at. Schmid specifically points to the narrow range of values in Impressionist work and discusses how those artists managed to convey a change in form without alteration in value.

I am tempted to say however, when one thinks of the general obsession with high saturation amongst Impressionists, that Schmid is concerned with the use of tints and shades rather than the actual luminosity of each color. In that sense one may use a variety of highly saturated colors in order to obtain a broad range of values (as in Asaro's painting).

So, Andymania, when you use Photoshop or a similar program to convert that portrait of Carolus Duran you will indeed see many many different values. Yet it seems to me by looking at the painting that few tints or shades were employed, which fits with my understanding of what Schmid is speaking of. Further exploration of the effect would lead, I imagine, to something much like either Asaro's painting (if attempting a broader range of values) or Matisse's Green Stripe (http://en.wikipedia.org/wiki/Green_Stripe) (if attempting a narrower range of values).

[dpaint]

I think there is a general misunderstanding about the idea of four or five values in a picture.
I was always taught that those are full step values that make up those four or five values. Remember most painting systems only recognise no more than nine or ten steps in the value scale. In reality there are no steps, just a continuous gradation from black to white. You can slice it as thin as you want, but if you are using a nine step scale to measure values then those paintings are only using four or five values according to the divisions of the scale. Again this is a problem of explaining something to someone using a theoretical system describing a portion of reality.

[andymania]

Dp,

You mean those 4-5 are subdivided into another 4-5 steps right?

[dpaint]

What I mean is, if you accept the 9 or ten step scale and measure those ten steps against those paintings in black and white, they will fall into only a few places along the scale. So, the area of the cheek in Schmids painting may have slight gradations that don't exactly fall on the sixth place of the scale, but compared to the other steps, you can say the cheek overall is a six on the scale. It is no different than what you do when you paint from life, you are taking the millions of gradations possible and translating them to your paintings.

I read some where that most people can only discern a half step change in value, anything less and it is not apparent to them. Most artists can easily see quarter steps and very good artists see eighth step increments. If that is true, then most artist only have 40 steps in a value scale between black to white that they can easily recognise.
This fits in with my experience teaching workshops for the last 15 years. Maybe other people who teach traditional painting have other experiences they would like to share regarding this.

[Graydon]

Anid,

You have to be clear what you mean by tints and shades. These are painters terms. Color comprises hue, value and chroma. So the discussion has to be whether one element can be used without the other to convey form. So far, value seems to be the most important with hue and chroma as reinforcements. Schmid suggests that form can be modeled with hue alone in areas. And perhaps, when a subject is high key, hue plays a bigger role. In high key paintings, chroma could also figure but at values 8 and 9, only the yellows and yellow oranges are high in chroma.

[Anid Maro]

Indeed they are painter's terms. My apologies, I often switch between additive and subtractive color spaces frequently and tend to use terms from one in the other and vice versa. I'll put out some definitions so I can be clear with what I'm saying.

Tint = A color desaturated by addition of white paint. When working digitally this can be simulated by increasing the RGB values beyond what is necessary for full saturation of a hue.

Shade = A color desaturated by addition of black paint. When working digitally this can be simulated by decreasing the RGB values below what is necessary for full saturation of a hue.

Saturation = A measure of the "purity" of a color, or rather how strongly of a hue a color appears to be. Typically a measure reserved for an additive system (such as RGB).

Chroma = Another measure of the "purity" of a color, or as described by David Briggs: "the degree of visual difference from neutral grey". Typically a measure reserved for a subtractive system (such as CMYK, or paint pigments).

Value = A measure of how light or dark a color is. Often represented in relation to a gray scale in 10 steps.

Hue = What is typically seen as the "color" part of color.

At higher levels of saturation (where the most extreme reaches of a color's chroma are used) hue itself becomes a more powerful means of conveying form since the typical means of adjusting value (tints/shades or mixing in the complement) is no longer available and shifts in hue provide the only remaining method to do so.

All three aspects of color, hue value and chroma (what I refer to as saturation) are quite interrelated. Each hue has its own tendencies toward certain values (such as a yellow having a higher chroma at higher value steps) and in reverse each value step might be said to have its own tendencies toward certain hues (such as a value of 9 may only have a high chroma in yellowish hue). Adherence or deviance from this framework determines the range of chroma available to work with (for example, a yellow at a value of 2 won't look much like yellow at all but rather a more muddy sort of color).

Manipulation of the interrelation between hue value and chroma can yield some interesting results, like that Asaro painting that was posted earlier.

From reading Schmid's words, it seems to me that he is speaking more about using a change in hue for adjusting value in order to avoid the use of shades or tints (which would reduce the chroma) rather than the actual replacement of values by hue alone (which, given my words above regarding the relations between hue value and chroma, makes little sense to me). Furthermore such an understanding explains how the range of values in the Carlos Duran painting (that, aside from some highlights and cast shadows, appears to be painted with high chroma colors) can be consistent with Schmid's words (although I do not dispute Dpaint's explanation).

I hope that better describes what I was trying to say.

[andymania]

ahh thsnk you graydon and Anid. But it seems impossible to change form without a value change.

[dose]

At higher levels of saturation (where the most extreme reaches of a color's chroma are used) hue itself becomes a more powerful means of conveying form since the typical means of adjusting value (tints/shades or mixing in the complement) is no longer available and shifts in hue provide the only remaining method to do so.


The problem with this is that a hue shift in HSB or HSL also creates a value shift to the human eye. If you have Photoshop, open up the color dialog and push saturation to 100% and brightness to anything above 25%. Then adjust the hue while leaving saturation and brightness alone. As you're adjusting hue, watch the Luminance value in Lab, which is a fairly accurate measure of how light or dark a color appears to the human eye. You'll notice that the luminance changes along with the hue shift.

This is the same logic that I believe led Richard Schmid to write what he wrote. Using cadmium red instead of cadmium orange is a "simple color switch" in terms of the paint on his palette, but cadmium red is actually darker than cadmium orange.


All three aspects of color, hue value and chroma (what I refer to as saturation) are quite interrelated. Each hue has its own tendencies toward certain values (such as a yellow having a higher chroma at higher value steps) and in reverse each value step might be said to have its own tendencies toward certain hues (such as a value of 9 may only have a high chroma in yellowish hue). Adherence or deviance from this framework determines the range of chroma available to work with (for example, a yellow at a value of 2 won't look much like yellow at all but rather a more muddy sort of color).


Actually hue, value, and chroma can function quite independently. It's possible to have a high chroma/high value yellow and a low chroma/high value yellow. At lower values, yellow is unable to reach higher chromas and will approach what usually called "brown", though often a greenish brown for yellows leaning away from orange. A better term for "brown" or "a muddy sort of color" is "a low value, low/medium chroma yellow".

It's important to understand the difference between chroma in the "real world" and saturation on the computer. Chroma is an absolute scale from 0 (neutral gray) extending theoretically to infinity, though the highest chromas currently available in paint top out around 30. The idea is that any color at the second chroma- regardless of its hue and value- will have the same amount of "colorfulness" in it. Certain hues can achieve higher chromas at different values, but that doesn't mean it can't also achieve the lower chromas at that same value.

Unfortunately none of HSB, HSL, RGB, CMYK, or Lab model HVC exactly, and adjusting any single one attribute of those color spaces changes two or more dimensions in HVC. Similarly, in real paint, simply adding white or black usually shifts two or more dimensions as well (Adding the complement to a color almost always results in a shift in all three).



http://www.huevaluechroma.com/pics/8-6c.jpg
(from http://www.huevaluechroma.com/082.php)

In HSB (probably the most common color space used by digital artists) is normalized such that 100% saturation is the maximum chroma displayable for the given Hue and Brightness in the current RGB space. As such, 100% saturation of one hue at a certain brightness might actually be a different chroma than the same hue at a different brightness, even though they both read 100% saturation. Similarly, 100% saturation of one hue at a certain brightness might be a different chroma than another hue at the same brightness. The result is that HSB doens't map directly to Hue, Value, and Chroma, and you're usually modifying two dimensions in HVC when adjusting one of H, S, or B. When you lower saturation, most of the time you're actually increasing value and decreasing chroma at the same time. Shifting hue usually changes value and chroma at the same time.

"Tints" and "Shades" in computer programs are usually just equivalent to lowering Saturation to create a tint, and lowering Brightness to create a shade. They're easy ways for programmers to throw in a feature that looks helpful to artists, but they're usually based on convenient math for programmers, rather than on what would truly be most useful for the artist. Unfortunately, what would be most useful for artists would involve a lot of strange math and some interesting interface problems- and would probably involve a bit of a learning curve for the artist. As of yet, no one has been willing to tackle this problem, and generally stick to more straightforward transformations of RGB space that have been done by people before them.

[Anid Maro]

Fully agreed Dose. The change in Luminance (also referred to Luminosity from what I've seen) with hue shifts is exactly what I was speaking of and what I think Schmid was referring to when he was talking about altering color temperatures (hue) before altering value (adding white or black paint for tints or shades).

Regarding chroma, that's why I was careful to say that the HVC framework determines the "range" of chroma possible and not that it determines the chroma itself. That is to say; you can have a high value and high chroma yellow as well as a high value and low chroma yellow, but there is no low value and high chroma yellow (as seen in the Munsell slices you put up).

At one point I said that value could be thought to determine hue, but I should clarify that this is highly dependent on the level of chroma one is picking. If you are looking for a low amount of chroma, then it is easy to obtain a hue at any value you would like. However it is only at the higher reaches of chroma where value becomes a determining factor in what hues are available.

Regarding "tints" and "shades", as far as I can tell (which may be incorrect) the adjustment of saturation and brightness to create tints and shades (either by using S/B sliders, directly altering the RGB values, or digitally "mixing" white/black color into another color on a digital palette) is for all practical purposes the same as the mixing of actual white/black pigments into a real paint.

Also thanks for the clarification on the difference between chroma and saturation. I can see now I'd been using the word "saturation" inappropriately. I fell I quite well understand the actual workings (well, at least in a layman's sort of way), but there are nuances in the verbage that often escape me (and result in communication issues).

Color Theory is a friggin' beast that takes years to tame, or so it seems to me. :)

ahh thsnk you graydon and Anid. But it seems impossible to change form without a value change.

Actually that's what I was getting at. Changing the hue will actually change the value as well. So no, you cannot effectively describe the form of an object without a change in value, however there are other less direct methods of creating different values.

[artworking]

...is it possible to perceive form through our perception of depth alone, or does our depth perception solely depends on light bouncing..

I ask these questions-
1) Is it possible to create form in a painting through hue alone, without changing value (perceived lightness and darkness)?
2) Do our eyes also sometimes perceive light bouncing as an opposite hue (on the color wheel)?
I believe the answer to both is yes.

Richard Schmid puts a lot of energy into describing the effects of light on a subject as it relates to painting, "warm light=cool shadow, cool light=warm shadow". Ask yourself, "Is your subject in cool or warm light?" and "Is what you are painting in the light or in the shadow?" then, mix your colors appropriately. I am going to get really technical now... take a deep breath... because sometimes that rule is broken, I have observed, when working through the same value painted in the shadow to the end of the hues' shadow spectrum it becomes necessary to cross over to the other side and pick up a cool color (in the shadow). Schmid knows when it is the right time to do this and it seems to have something to do with light that is bouncing.

Just when I thought I understood cool light=warm shadow, I realized there was another way of creating depth by painting all the way around to the other side. Warning:This use of hue is not appropriate for all paintings and should be used only sparingly.


when a subject is high key, hue plays a bigger role

And yes Graydon, in high key it does seem to play a bigger role. Perhaps because our eyes see more detail in the light, but these rules work in the dark values as well.

Thank you for this topic. And thank you Richard Schmid for writing such an amazing book on painting technique.

I have my monitor set to millions of colors. Does any one know how many millions that could be? What ever that number is, my eyes are able to perceive even more.

[Black Spot]

Has anyone tried using a B&W filter on and drawing with random colours? If the values are right, does it matter what the colour is?