PDA

View Full Version : What are the 3 primary pigments?


kbuckland
01-19-2004, 09:22 AM
Hello all!

My mind has been racked with color theory latley. My current palette is the "Gruppe" palette, a warm and cool of each primary. I'v used it for about a year now and have been, for the most part, satisfied. However, I would like to expierement with a three color palette using only the primaries + white. Any suggestions on which three pigments would match the primaries the closest?

Thanks,
Kyle
www.kylebucklandfineart.com

Eugene Veszely
01-19-2004, 09:28 AM
Depends on who you talk to :)

Some say Cyan, Magenta and lemon yellow ... others have much better knowledge :)

Helen Zapata
01-19-2004, 10:33 AM
Here is the Primary Palette CMY Project (http://www.wetcanvas.com/Community/Projects/browse_details.php?proj_id=420)

This might help. It's amazing what you can do with only the primary colors.

Helen

Richard Saylor
01-19-2004, 10:46 AM
Depends on who you talk to :)

Some say Cyan, Magenta and lemon yellow ... others have much better knowledge :)
I've tried that, and lemon yellow makes wonderful greens, but the reds (Magenta + Yellow) tend to be muddy. One of the main problems with three-color painting is getting good greens and reds. It tends to be a compromise, but I think a more neutral yellow works better than lemon.

I get good results in watercolor with W&N Quinacridone Magenta (PR 122), Winsor Blue GS (PB 15), and either Winsor Yellow (PY 154) or Transparent Yellow (PY 97). I've found these to work better than the primaries recommended by W&N (Permanent Rose, Winsor Blue RS, Winsor Lemon).

The jury's still out on the best three primaries for gouache and acrylic. (I don't do oils any more.) In any event, CMY has always worked much better for me than RBY.

kbuckland
01-19-2004, 01:23 PM
I appreciate the fast, friendly, and informative info. :clap: This is a wonderful website!Thank you,

Kyle

Marc Sabatella
01-19-2004, 02:17 PM
I'd recommend searching out a rather long thread from a couple of months ago that started in this forum when I asked a similar question; the title was something about the color wheel, I think. The discussion went all over the place, but the gist of what I took from it is this:

We all learned in school that red, yellow, and blue were primaries because they can't be mixed from other colors. This quite simply isn't true.

The construction of the eye is such that it responds most directly to red, green, and blue; all other colors are intepreted by our eye as mixtures of these three. The three colors that that are most primary in terms of how our eye actually responds to them are cyan (which reflects equal parts green and blue), magenta (which reflects equal parts red and blue light), and yellow (which reflects equal parts red and green light). The reality is, no pigments actually behave this way at all, so the theory doesn't apply as well as we'd like. In the real world, pigments what they reflect what they want to reflect, and it's usually less than theory would prefer, so no combination of three pigments will enable you to mix everything. But a pigment that is more or less cyan colored, like phthalo blue, and a more or less magenta one, like one of the quinacridones, plus a yellow pigment like cadmium yellow lemon, will probbaly yield the widest range of colors of any three color palette you might pick.

Whether that's enough is another matter. I personally found the violets I could get this way not intense enough - the phthalo was just too green - and also found it too hard to get the range of oranges I wanted, as the magenta was too blue. Not only was the fullest intensity orange I could get a bit weak, but there wasn't a lot of room for error mixing the duller earth tone types of colors this way.

Now, combine this with the fact that phthalo blue is particularly difficult to work with - and there really aren't a lot of other good options that are close to cyan in hue - and I pretty much gave up on this palette. I'm coming closer to something I like, though. Instead of phthalo blue, I am using ultramarine. Obviously, I get cleaner violets this way. It also gives my duller greens by default, which is almost always how I want my greens anyhow. For the relatively rare cases where I want a more intense green, I also have phthalo green on my palette, which I mostly use to "punch up" a green mixed from ultramarine and yellow. Also, to have an easier time with the oranges, I added a color between the magenta and the yellow. Right now, it's naphthol. It's a help, but it seems closer to the quinacridone rose than need be, so I may switch to cadmium orange. Or I might add - or switch to - a more orange leaning yellow than the azo/arylide yellow I'm using now. Because while oranges are easier now, they are still the hardest. Part of this is inherent in the nature of how we perceive color - slight changes in hue or intensity in the warm colors is perceived by our eyes as a much bigger difference than similar shifts in coolr colors. But I still not sure I'm encountering any major problems that I don't think keeping my brushes a little cleaner when mixing oranges wouldn't cure.

Oh, I also have burnt sienna on my palette, mostly for underpainting.

cunparis
01-20-2004, 06:03 AM
Hello all!

My mind has been racked with color theory latley. My current palette is the "Gruppe" palette, a warm and cool of each primary. I'v used it for about a year now and have been, for the most part, satisfied. However, I would like to expierement with a three color palette using only the primaries + white. Any suggestions on which three pigments would match the primaries the closest?


Marc gave some good info. Also you might want to check out the book "Fill Your Oil Paintings With Color and Light" (I think that's it) by Kevin MacPherson. He uses:

- cad yellow light
- alizarin crimson
- ultramarine
- pthalo green

he says to start with the first 3 and then add green later on (first learn to mix your own greens). I made a color wheel and I liked it. I held my wheel up to the paintings in his book and I could see my colors in his paintings. That was cool. Then I tried:

- cad yellow light
- cad red medium
- cobalt blue hue (PB29)

I found the oranges weren't much better, the greens were the same, and the purple was very muddy. So I prefer the first palette. With the aliz crimson I could add a touch of yellow and get a good red. So then I tried:

- cad yellow light
- alizarin crimson
- cobalt blue hue (PB29)

I like this because the cobalt blue hue is slightly more opaque than the ultramarine, and mixes about the same. Not sure how real cobalt blue would work. I have already started experimenting with this limited palette and I like it. Colors are not as bright & intense as with a full palette, so I'm not sure I'd keep this palette forever. But I think it's really good practice to use it for a while.

Kevin says in his book that if you use a full palette, you will typically find the closest color on your palette and use it, even if it's not exact. When you have to mix all your colors from scratch, you're more likely to get an exact match. This makes a lot of sense. Depends on how you like to paint. If you want to paint more expressionist, then you want real bright pure colors that don't necessarily match the real colors.

-michael

JamieWG
01-20-2004, 08:49 AM
Hello all!

My mind has been racked with color theory latley. My current palette is the "Gruppe" palette, a warm and cool of each primary. I'v used it for about a year now and have been, for the most part, satisfied. However, I would like to expierement with a three color palette using only the primaries + white. Any suggestions on which three pigments would match the primaries the closest?

Thanks,
Kyle
www.kylebucklandfineart.com

Helen gave you a link to the CMY palette project, which tells you the manufacturers' recommendations for the primaries in their brands. Perhaps doing a couple of paintings with those colors would be a good start for you.

The true primaries are cyan, yellow and magenta. As Marc said, these can sometimes be difficult to work with, but a wonderful learning experience. It takes a lot of mixing to get the *real* primaries under control. So, many elect a split primary of reds, blues, and yellows as you've been doing, or the McPherson palette. These are far easier to control, but are in fact not the true primaries in terms of color theory. (That may or may not matter to you in terms of practical applications.)

If you use alizarin, be sure that you've selected one of the permanent alizarins, rather than the fugitive alizarin crimson. Gamblin and Winsor Newton both make good ones, and I'm sure other paint manufacturers do also.

Jamie

Patrick1
01-20-2004, 10:06 AM
I'm sure most people here have seen this a zillion times, but here's what Handprint has to say:

http://www.handprint.com/HP/WCL/color4.html#primary

http://handprint.com/HP/WCL/palette5.html#primaries

http://handprint.com/HP/WCL/palette4c.html

kbuckland
01-21-2004, 01:33 PM
Hello! Thanks again for the informative input. Here is my progress thus far: I tried Rose Madder, Cad. Lemon, and Pthalo. The range was impressive however the secondaries lacked in some places, namley the Oranges and Purples. I then decided to add the secondaries and switch my blue to Ultramarine sense Viridian is close in character to Pthalo Blue. So my palette consisted of Rose Madder, Cad. Orange, Cad Yellow Lemon, Viridian, French Ultra. Blue, and Dioxine Purple. With this I found a more structured solution, however I still could not get thoes nice clean warm yellows that I was used to with Cad Yellow Med. I am now using :

DioxinePurple RoseMadder CadOrange CadYellowMed

FrenchUltramarine

Viridian

CadLemon

-I guess it would be hard for an artist to be too picky when it came to his or her tools :)

My online gallery is mostly landscapes painted with the Gruppe Palette, however I plan to add two still -lifes to the gallery which are the results of this fourm and my expierements.Please visit.

Thanks again-
Kyle
www.kylebucklandfineart.com

JamieWG
01-21-2004, 01:45 PM
Kyle, this is too funny. You and I have arrived at very similar palettes. I too added the secondaries (notably cad orange and dioxazine, plus phthalo green) to the true primaries, for the same reasons as you, then switched out the phthalo blue and went back to French Ultramarine and Cerulean blues because they were so much easier to work and not so "hot" in saturation. Next I ditched the phthalo green to go back to viridian. Then I added a mid-yellow cadmium. I go out with a full palette of about 35 colors, but the ones I use most are like your palette, plus a transparent red oxide and yellow ochre. I add in a cad red light for portraiture/figures too.

Be sure you check the permanence on your rose madder. Some are fairly fugitive.

Jamie

h20color
01-21-2004, 07:06 PM
Here is my input--to add to the mix

I use watercolor but I've switched to a clean secondary palette that leaves less chance of a muddy work using these colors. You can always add that particular color that you feel you need. I get bright colors less mud. You can dull the mix down by adding it's correct complement or the old standby Payne's Gray. This palette seems to offer the best of complementary colors. Any two complementary colors that make a pure gray will work.

This provides pure pigment for each of the colors and is pretty much balanced
PY=prim yellow, PB=prim blue, PV=prim violet, PR=prim red blah, blah, blah

Hanza Yellow (PY97)
Cadmium red orange (PR108)
Quinacridone rose (PV19)
Ultramarine blue (PB29)
Phthalocyanine cyan (PB17)
Phthalocyanine green YS (PG36)


Phthalocyanine turquoise (PB16) is close to a true cyan hue useful in landscape painting.


very close to some of the palettes listed.

Remember it's just another option.

Bad Dobby
01-24-2004, 05:26 PM
Here is Stephen Quillers response to that question:

http://www.dick-blick.com/items/049/10/04910-0000-3ww.jpg

Also, this painters wheel is helpful too:

http://www.handprint.com/HP/WCL/IMG/cwheel.gif

MsLilypond
01-26-2004, 03:35 PM
I have read the closest to the true primaries, is hansa yellow, pthalo blue & quin. magenta. I got the information off handprint.com and even though the site owner uses watercolours the paint & pigment info should apply to any paint medium

cunparis
01-26-2004, 04:09 PM
I have read the closest to the true primaries, is hansa yellow, pthalo blue & quin. magenta. I got the information off handprint.com and even though the site owner uses watercolours the paint & pigment info should apply to any paint medium

I agree with those colors. My thought is that it's nice to come up with a 3 color primary palette that includes colors you'd use a lot. I think that's why cad yellow light, alizarin crimson, and ultramarine are so popular for the 3 color palette.

I've done a lot of experimenting and for my last painting I used cad yellow medium (hue), alizarin crimson, and cobalt blue (hue). I really like those three. The yellow medium is really bright, yet I can easily tone it down with a bit of white..

Bad Dobby
01-27-2004, 02:06 PM
I think another way to answer this seemingly simple but complex question is to say: there are 3 primaries in theory ... and 3 primaries is practice. Primaries can be any 3 colors that you choose. There is no 1 answer. MacEvoy in his Handprint site states that you need several pigments (and those can vary between brands and transparency vs opagueness desired) to achieve the most saturated colors for secondary and tertiary colors. That using 3 primaries (say Hansa Yellow, Quin. Rose, Pthalo Blue) may give you a good overall mixing range; however it will result in lower saturation for the secondaries and tertiaries. So he proposes what he calls "The [6 color] Secondary Palette."

Now contrasted against the "saturation/vivid color" approach to the 3/6 pigment palette color theory above; Artists like Scott Burdick (Oct 2003 The Artist's Mag.) states that the 3 primaries that you choose (in addition to white on your palette) should be limited to one "pure" primary and 2 "grayed" primaries to achieve great color harmony. He notes that most colors around us are grayed down, that there are fewer instances of a pure primary in our visual experience. He claims that using 3 saturated primaries can lead to an unnatural gaudiness. Sure you can gray them down, but that requires additional mixing and potential for dead colors or mud. He also states that you can even use 2 pure primaries and 1 grayed, but he preferes the former. Some examples that he cites for "Primary Power" are:

Cad red (pure red primary)
Ivory black (grayed blue primary)
Yellow ochre (grayed yellow primary)
(Burdick states that Zorn used the above palette)

Ultra. blue (pure blue primary)
Burnt sienna (grayed red primary)
Raw sienna (grayed yellow primary)

Cad yellow (pure yellow primary)
Viridian (grayed blue primary)
Venetian red (grayed red primary)

Burdick also unifies his mixes by adding various amount of one color (say yellow in a sunlit scene, or blue in a cool light scene, etc.) into the other 2 primaries. And occassionally, he will use the one pure primary without alteration as a high note or accent .. but in small amounts. He says that the limited palette approach is but one way to paint, and a nice departure from a larger palette from time to time.

I guess it all is up to the individual Artist and what they want to achieve. Obviously, you can't get a bright & cheery highly saturated full gamut painting using Burdick's model. But you will achieve strong color harmony and consistency.

Richard Saylor
01-27-2004, 03:17 PM
...Primaries can be any 3 colors that you choose...."
It is fashionably postmodern (and gives the impression of profundity) to redefine established terms in such a way that they become meaningless, but it becomes annoying after awhile. :(

Marc Sabatella
01-28-2004, 02:30 PM
It is fashionably postmodern (and gives the impression of profundity) to redefine established terms in such a way that they become meaningless, but it becomes annoying after awhile. :(

This can be true, but the handprint site has a real point. In the real world of real pigments, there is nothing particular magic about what we think of as the three primaries - *all* colors can be mixed or at least approximated from adjacent colors on the color wheel, and *no* set of three colors will allow you to mix everything perfectly. True, there are reasons why, theory, if pigments existed that reflected light in exactly this way, cyan, magenta, and yellow would be the best choices, but as it is, real magenta and real cyan pigments have no special properties that make thjem any more primary than other colors in the range of what they can mix. Yellow is special only in that it is so light that it is cannot be mixed from other pigments which are darker, plus the fact that our eyes are more sensitive to small differences in hue in the yellows than elsewhere. But a well chosen four primary system that included yellow could be just as effective - more so, actually - than a three primary system, even if cyan and magenta were not part of the four.

Richard Saylor
01-28-2004, 02:52 PM
This can be true, but the handprint site has a real point. In the real world of real pigments, there is nothing particular magic about what we think of as the three primaries - *all* colors can be mixed or at least approximated from adjacent colors on the color wheel, and *no* set of three colors will allow you to mix everything perfectly. True, there are reasons why, theory, if pigments existed that reflected light in exactly this way, cyan, magenta, and yellow would be the best choices, but as it is, real magenta and real cyan pigments have no special properties that make thjem any more primary than other colors in the range of what they can mix. Yellow is special only in that it is so light that it is cannot be mixed from other pigments which are darker, plus the fact that our eyes are more sensitive to small differences in hue in the yellows than elsewhere. But a well chosen four primary system that included yellow could be just as effective - more so, actually - than a three primary system, even if cyan and magenta were not part of the four.
Why do you always say that about yellow? Either one can mix yellow from other pigments or one can't. I also doubt that you could mix a convincing cyan or magenta from, say, ultramarine blue, cadmium red medium (or light), pthalo green, and the yellow of your choice.

Richard Saylor
01-28-2004, 05:18 PM
Just a few more thoughts on the subject...

1. There are no exact C, M, or Y pigments. This does not mean that CMY theory should be thrown out as useless, any more than the number pi should be thrown out merely because it has no exact numerical representation. (However, somewhere, I think it was in Indiana, someone proposed to the state legislature that pi be legally declared to be equal to 3. This happened quite a few years ago, of course.)

2. The handprint site is so extensive and exhaustive that one can use it to support almost any opinion, as long as that opinion is not too unreasonable. In that respect it's similar to the Bible.

3. Suppose we have perfect red, green, blue, cyan, and magenta pigments. It is impossible, even in theory, to mix a yellow hue from them, since they combine according to the colors they absorb, not the colors they reflect. The same goes for cyan and magenta as for yellow. To see this, one can either use common sense or (if so inclined) algebra.

Marc Sabatella
01-28-2004, 07:16 PM
Why do you always say that about yellow? Either one can mix yellow from other pigments or one can't. I also doubt that you could mix a convincing cyan or magenta from, say, ultramarine blue, cadmium red medium (or light), pthalo green, and the yellow of your choice.

First, regarding yellow - both the reasons I gave for yellow being difficult to mix seem perfectly valid to me. The Handprint site is but one of the sources I've seen that talks about the unusual sensitivity the human eye has to changes in both hue and intensity in the yellow region of the spectrum. Thus, in order to mix something that *looks* yellow, you have to hit a relatively narrow range of wavelengths. Miss it by, say, 1%, and the result looks "off" yellow by more than if you missed an attempt to mix, say, a cyan by the same amount. And what this is going to mean in practice is that you are going to have to start with two colors that are already very close to the target wavelength on the color wheel - the further the source colors are from the target, the more the result is going to differ in intensity from what we perceive as yellow. Again, while the same is true of, say, cyan, our eye can tolerate a bigger loss of intensity in cyan than it can in yellow and still be satisfied that we have a match. The bit about the value of yellow being so much brighter than other colors was something someone else suggested as being a possible explanation for this phenomenon, and I have no idea how true it is, but certainly, you can't deny that mixing two colors is going to produce a color darker than the lighter of the two source colors, and since nothing is lighter than yellow among full intensity colors, we are definitely up against it trying to mix a yellow. Again, the further you get from the target yellow in your source colors, the darker the mix will be. Lightening with white might help in theory, but we'd still be fighting the loss in intensity.

As for mixing a cyan from the colors you mention, ultramarine and phthalo green produce as good a cyan in terms of how we actually perceive it as any actual cyan and magenta pigments will produce a blue or violet. Frankly, I'm less impressed by the violets I can get from cyan and magenta than by the cyans I can get from pretty much any blue and green. True, the mixed cyan won't behave as the theoretical cyan that reflects the blue and green regions of the spectrum equally - but nor does any existing cyan pigment.

Marc Sabatella
01-28-2004, 07:33 PM
1. There are no exact C, M, or Y pigments. This does not mean that CMY theory should be thrown out as useless, any more than the number pi should be thrown out merely because it has no exact numerical representation. (However, somewhere, I think it was in Indiana, someone proposed to the state legislature that pi be legally declared to be equal to 3. This happened quite a few years ago, of course.)


I don't think anyone is trying to throw out CMY theory - just be realistic about its applications to the real world. pi still *is* applicable to the real world, even though it has no exact terminating decimal representation. In the real world, the ratio of circumference to diameter really *is* pi, not 3. But in the real world, cyan and magenta *don't* mix a wider range of colors than any other two colors equally spaced on the color wheel.


2. The handprint site is so extensive and exhaustive that one can use it to support almost any opinion, as long as that opinion is not too unreasonable. In that respect it's similar to the Bible.


Yes, admittedly, some of us get lazy with that... :-)


3. Suppose we have perfect red, green, blue, cyan, and magenta pigments. It is impossible, even in theory, to mix a yellow hue from them, since they combine according to the colors they absorb, not the colors they reflect. The same goes for cyan and magenta as for yellow. To see this, one can either use common sense or (if so inclined) algebra.

Again, I perfectly agree regarding the difficulties in mixing yellow - although here, you've made the problem harder by choosing colors that have a wider gap around yellow than around most other colors. Make the green a yellowish green, the red and orangish red, so that both colors reflect some yellow, and a mixture would indeed leave some yellow - as much as a mixture of actual cyan and magenta pigments would leave blue, which is to say, not a ton. It's just that our eyes will still accept the mixed blue as blue despite the loss in intensity, but they won't accept the mixed yellow as yellow.

But as for mixing cyan, this is difficult only if one thinks of cyan as this magical color that exists only in theory and reflects blue and green regions of the spectrum equally. In the real world, cyan is simply a wavelength on the color wheel somewhere between blue and green, and a color with this same basic wavelength can be mixed from blue and green as easily as any other color can be mixed from its surrounding colors. The reusltant color might not be the magic theoretical cyan in terms of its exact reflectance pattern, but it will look much the same - as good as a blue or violet mixed from cyan and magenta pigments.

Richard Saylor
01-28-2004, 08:12 PM
Well, Marc, I don't think we can get anywhere with this right brain vs. left brain discussion. In practice, it really doesn't matter, as long as one is able to get the desired results.

Patrick1
01-28-2004, 11:55 PM
Marc, you're not really Einion, are you? :p

...and since nothing is lighter than yellow among full intensity colors, we are definitely up against it trying to mix a yellow.

FWIW, I tried lemon yellow + a slightly orangy yellow (like PY 74) and got a very acceptable 'middle' or 'primary' yellow. But as you said, as your two starting yellows get farther apart, the resulting yellow gets dull very quickly.


As for mixing a cyan from the colors you mention, ultramarine and phthalo green produce as good a cyan in terms of how we actually perceive it as any actual cyan and magenta pigments will produce a blue or violet.

I just tried this. I mixed Quin. Magenta (PR 122) + Phthalo Blue GS to get a violet-blue, and I also mixed Ultramarine Blue + Phthalo Green BS to get a 'cyan'.

The results? Judging only by my eyes, the mixed violet-blue was noticably duller than Dioxazine Purple, and way duller than Ultramarine Blue. The mixed cyan got the cyan hue right, but it looked as though some black was in the mix...it didn't have any of Phathalo Blue GS's brilliant cyan undertone. Both mixes were dull, but to my eyes, the mixed cyan was at least slightly duller, comparatively, to the mixed violet-blues. If anyone's interested, I'll post the swatches in a few days when they dry.


True, the mixed cyan won't behave as the theoretical cyan that reflects the blue and green regions of the spectrum equally - but nor does any existing cyan pigment.

You can get the hue of the mixed cyan right...so that it reflects equal amounts in the blue and green thirds of the spectrum...but it's the amount of blue and green light that it reflects...the lack of it...that makes it dull. On Handprint somewhere it says that even the best single-pigment cyan we have, Phthalo Cyan (PB 17) only reflects something like 40% (forget the exact number) of the blue and green light that the ideal cyan would.


But in the real world, cyan and magenta *don't* mix a wider range of colors than any other two colors equally spaced on the color wheel.

I'm sure they do. If that were true, two secondaries would mix up a primary as clean as a secondary mixed from two primaries. Try mixing a clean magenta from an orange-red + a violet-blue :). That's the whole reason printers, after experimentation, chose CMY; it's far from perfect, but it's the best gamut you can get from only three colours. But which magenta and cyan pigments? That is the question.


Suppose we have perfect red, green, blue, cyan, and magenta pigments. It is impossible, even in theory, to mix a yellow hue from them, since they combine according to the colors they absorb, not the colors they reflect. The same goes for cyan and magenta as for yellow. To see this, one can either use common sense or (if so inclined) algebra.

If you're dealing with the non-existant primaries and secondaries with the 'ideal' reflectance (100% reflectance in the proper thirds of the spectrum and 0% in the other thirds), then this would be true. So for example, mixing red + blue would result in black, not magenta. Green + red would also result in black...etc.

One caveat about using algebra to calculate the resultant relectance curve from the reflectance curves of the two starting pigments: Handprint says that this will only give a rough estimate, because it doesn't take into account things such as scattering power (which I have no idea what that is), the pigmnet vehicle, and several other important factors. He said that a pigment chemist told him that the only way to know for sure what you get when you mix two pigments is to go out and mix 'em. :cool:

Bad Dobby
01-29-2004, 08:20 PM
It is fashionably postmodern (and gives the impression of profundity) to redefine established terms in such a way that they become meaningless, but it becomes annoying after awhile. :(

Hahahaha... well I did go into a bit of detailed explanation beyond the few words that you quoted me on (rather than leave anyone hanging out in leftist LaLa land, as you suggest). As usual, there is rarely one answer. Having the ability to think more flexibly in any body of knowledge yields more variety of application for an array of unique needs. Remember that we are talking about artistic painting (not process inks) which is expressive and subjective. Having a broader understanding of "What are the 3 primary pigments?" is helpful along with a reductionist approach. I like having more choice and options rather than less. But I certainly understand that sometimes information needs to be distilled down to more simplified and practical terms as well. That is why I like this forum, multiple approaches (theoretical and applied) get aired ... and we learn more from all contributions, including yours and mine and theirs. ;)

The purpose of a forum is to solicit a diverse contribution. What is meaningless to one person may be a revelation to the next person. As usual, take what you like and discard the rest. That has always been my selection process in life's offerings. Sometimes I find myself returning to that which I have previoulsy discarded and picking it up ... and also discarding that which I had previously taken in. Learning is an iterative process. Knowledge is rarely static .. it evolves.

Marc Sabatella
01-30-2004, 12:36 PM
Marc, you're not really Einion, are you? :p


No, but the name sounds familiar. Another poster here, I assume?


I just tried this. I mixed Quin. Magenta (PR 122) + Phthalo Blue GS to get a violet-blue, and I also mixed Ultramarine Blue + Phthalo Green BS to get a 'cyan'.

The results? Judging only by my eyes, the mixed violet-blue was noticably duller than Dioxazine Purple, and way duller than Ultramarine Blue. The mixed cyan got the cyan hue right, but it looked as though some black was in the mix...it didn't have any of Phathalo Blue GS's brilliant cyan undertone. Both mixes were dull, but to my eyes, the mixed cyan was at least slightly duller, comparatively, to the mixed violet-blues. If anyone's interested, I'll post the swatches in a few days when they dry.


I'd be curious to see, although certainly, I've done the same thing, and I'd just say I was comparatively more disappointed in the mixed blue-violet than in the mixed cyan. Neither, obviously, were perfect. But I happen to be especially sensitive to violet, or so it seems.


If that were true, two secondaries would mix up a primary as clean as a secondary mixed from two primaries. Try mixing a clean magenta from an orange-red + a violet-blue :).


I've tried it with naphthol and ultramarine, and didn't think the results were any more disappointing than the mixed violet I got from quinacridone and phthalo blue.


That's the whole reason printers, after experimentation, chose CMY; it's far from perfect, but it's the best gamut you can get from only three colours.


Note this much I do agree with - it's just that I am not convinced at all about *why* it's the best. I think it's because they are reasonably equally spaced on the color wheel, and by including yellow, it allows the cleanest yellow mixtures. I'm still not convinced that, in practice, either cyan or magenta pigments have any properties that make them harder to mix matches for than any other color, which is the usual definition of "primary". That is, four or five equally spaced colors on the color wheel would seem to produce a wider gamut (assuming yellow was included) than CMY, as far as I can tell, even though C or M would be missing from the four or five color palette.

Richard Saylor
01-30-2004, 04:49 PM
Note this much I do agree with - it's just that I am not convinced at all about *why* it's the best. I think it's because they are reasonably equally spaced on the color wheel, and by including yellow, it allows the cleanest yellow mixtures.
I read somewhere else about three equally spaced colors on "the" color wheel being suitable primaries. (It sounds so cool and slick. Almost as neat as Bad Dobby's "any three colors." :)) However, one wonders which color wheel, since there isn't a "the" color wheel.

Using as a criterion for primaries that they cannot be mixed from other hues is antiquated. That's what they told us in elementary school about yellow, red, and blue, and that's clearly absurd. In theory, cyan, magenta, and yellow are perfect because they can be mixed to give the light primaries (red, green, and blue) and any hue which can be produced by the additive mixture of the RGB light primaries. In practice, the problem is finding CMY pigments which are sufficiently close matches to the ideal CMY.

It's certainly possible that one could achieve a wider gamut using five or six colors (not including C or M) rather than any three CMY colors now available. Many people have excellent results using a six color system: two reds, two yellows, two blues (one warm and one cool of each), but I don't think of this as having much, if anything, to do with a "primary" three-color system.

By the way, when I speak of "theoretical" CMY, I'm thinking of the colors produced on a computer monitor having hexadecimal representations 00FFFF (cyan), FF00FF (magenta), FFFF00 (yellow).

Marc Sabatella
01-30-2004, 06:12 PM
I read somewhere else about three equally spaced colors on "the" color wheel being suitable primaries. (It sounds so cool and slick. Almost as neat as Bad Dobby's "any three colors." :)) However, one wonders which color wheel, since there isn't a "the" color wheel.


True. And I suspect the answer to this question is related to the question of how we actually calculate the gamut of colors that can be generated from the three colors we chose, and more particularly, how we evaluate that gamut in practice. For instance, once we determine that a given set of colors has a gamut that excludes a certain set of colors, our perception of how much of an issue that is is going to depend in part on how important we see the omitted set to be, and that is in turn going to be in fluenced by how large it appears on the color wheel we have chosen. The same set of colors will look bigger or smaller on different wheels. See below.


Using as a criterion for primaries that they cannot be mixed from other hues is antiquated. That's what they told us in elementary school about yellow, red, and blue, and that's clearly absurd.


Well, we agree on this, but it is interesting how many artists still buy into the notion of yellow, red, and blue as the primaries.

As I see it, these colors do indeed split the color wheel into thirds, if we use the traditional color wheel that goes along with this model (yellow, orange, red, violet, blue, green all equally important). But yellow, cyan, and magenta do better if we use something more like the Munsell color wheel, in which the orange range is subsumed into the yellow and red, leaving only five roughly equal regions identified by the remaining colors - supposedly more in tune with how we actually perceive hue changes.

Now, one easily measurable downside of RBY as primaries is that the gamut won't include as good a range of bright greens as CMY. It can't - blue is further from yellow on any color wheel than cyan is. However, CMY won't give the same range of oranges, for a similar reason (magenta is further from yellow then red is). But the traditional color wheel will show the deficiency in oranges as a corresponding larger problem compared to the deficiency in greens, compared to how the Munsell wheel shows it.


It's certainly possible that one could achieve a wider gamut using five or six colors (not including C or M) rather than any three CMY colors now available. Many people have excellent results using a six color system: two reds, two yellows, two blues (one warm and one cool of each), but I don't think of this as having much, if anything, to do with a "primary" three-color system.


Then part of the debate is simply regarding terminology - what exactly does it mean to say a color is "primary"? I have been assuming the traditional definition - cannot be mixed, can be used to mix everything. The futility of this definition was pretty much my starting point. There is hardly a color worth having as a pigment on a limited palette that can be mixed *exactly* from other pigments, and no finite set of pigments can be used to mix *everything*. Mixtures of full intensity colors will always be less than full intensity. Again, this doesn't make we want to throw all the theory out - it makes me want to look deep enough to understand what is really necessary to have on one's palette in order to get the mixtures one desires.

Richard Saylor
01-30-2004, 07:02 PM
I feel that if three ideal colors don't work in theory as primaries, then three approximate colors stand little chance of working well in practice.

If red, blue, and yellow are to work, then it doesn't seem unreasonable to expect the perfect, middle-of-the-road theoretical colors to work: red (FF0000), blue (0000FF), yellow (FFFF00). However, as is well-known by now, these fail miserably.

Thus, a major difference between RBY and CMY is that CMY does indeed work theoretically (and it works pretty good in my inkjet printer too).

Patrick1
01-31-2004, 06:41 AM
No, but the name sounds familiar. Another poster here, I assume?

Yes...Einion is also one of the guides of this forum. But he hasn't posted in months now. Too bad, because he's a wealth of colour theory info.

...I was comparatively more disappointed in the mixed blue-violet than in the mixed cyan.

I re-did my mixing test, this time more carefully, and I'll say the mixed cyan turned out better this time. When the retarder medium and the oil dries, I'll scan it and post it here.


I've tried it with naphthol and ultramarine, and didn't think the results were any more disappointing than the mixed violet I got from quinacridone and phthalo blue.

Any middle red I've tried can't make anything close to magenta when mixed with ultra blue...it gives a clolour closer to a maroon. Adding white helps, but that's cheating. I'll bet your naphthol is one of the bluer ones and/or your quinacridone is not PR 122.


I'm still not convinced that, in practice, either cyan or magenta pigments have any properties that make them harder to mix matches for than any other color, which is the usual definition of "primary".

My understanding is this (anyone correct me if I'm wrong):
Their property that makes primaries more difficult to match than secondaries is that they maximally stimulate two rod types in the eye, and minimally stimulate the third.

A mixture of other two pigments might get the hue dead-on (equally stimulate those two thrids), but not stimulate them maximally, because subtractive mixing results in loss of light reflectance. And the key: all else being equal, lower brightness results in lower apparent saturation.

So you can't fully match an ideal primary (in theory and in practise), but in theory, you could match a secondary perfectly...if you had the ideal primaries. The only way to come close to matching a primary would be to choose starting pigments that bracket it as closely as possible (to minimize light loss), but ideally, a secondary could be matched by the two primaries that bracket it, which aren't that close to it.


That is, four or five equally spaced colors on the color wheel would seem to produce a wider gamut (assuming yellow was included) than CMY, as far as I can tell, even though C or M would be missing from the four or five color palette.

The only way to tell for sure would be to use a spectrophotometer. The reason is because, in practise, you can't predict the resultant mixtures of two pigments by simply drawing a line between them on a colour wheel; this only gives a very rough estimate.

If we're dealing with artists quality lightfast pigments (and not fugitive printing pigments) I'll venture a guess that a good Y R B G palette will give an overall gamut roughly equal to a good CMY. I'm sure a good 5-colour palette would beat CMY easily. Lascaux acrylics has such a system...click on Sirius Primary System at left:

http://www.lascaux.ch/english/farben/index.html

JamieWG
01-31-2004, 11:52 AM
Yes...Einion is also one of the guides of this forum. But he hasn't posted in months now. Too bad, because he's a wealth of colour theory info.

Not to worry; Einion is still around. :D He's been having some problems with his browsers lately and logging on to the site without his system crashing. But he has not deserted us! :angel:

Jamie

Marc Sabatella
01-31-2004, 02:52 PM
Any middle red I've tried can't make anything close to magenta when mixed with ultra blue...it gives a clolour closer to a maroon. Adding white helps, but that's cheating. I'll bet your naphthol is one of the bluer ones and/or your quinacridone is not PR 122.


Adding white is really cheating? I figure that's just par for the course because we are starting with relatively dark pigments - a trait I *like*, because it is much easier for me to lighten a color without affecting hue too much than to darken (I don't use black). All that matters to me in practice is whether I can mix the color I want - not whether I have to add white to get it.

Anyhow, I can't compare my naphthol to others so I can't how blue it is relatively speaking. I can say it is less orange than most cadmiums, and definitely makes for better magentas. Although I attribute this at least in part to its transparency relative to cadmiums, and its generally cleaner mixing properties. Also, it is true that I hadn't been using PR122 but PV19 as my quinacridone, and I'm aware this isn't the truest magenta, so of course that could affect my results. In any case, I'm not saying I mixed a truly great magenta - just that my perception of how much I missed by was not especially more than my perception of how much cyan & magenta missed mixing a blue violet.


My understanding is this (anyone correct me if I'm wrong):
Their property that makes primaries more difficult to match than secondaries is that they maximally stimulate two rod types in the eye, and minimally stimulate the third.

A mixture of other two pigments might get the hue dead-on (equally stimulate those two thrids), but not stimulate them maximally, because subtractive mixing results in loss of light reflectance.


That's my general impression of how it would work in theory as well, if there were actually pigments that behaved this way.


ideally, a secondary could be matched by the two primaries that bracket it, which aren't that close to it.


I am not so sure about this, however, even in theory, although I'm certainly willing to be convinced. When you reduce everything to the simple level of cone response, it appears this could be the case. But looking at the actual reflectance curves & the idea of subtractive mixing yielding a geometric mean, I'm not sure you could really match the reflectance curve of any given color as well as you might think. I could be wrong, though - I haven't done the math.

Patrick1
02-01-2004, 07:43 AM
Jamie, that's good news...I assumed he got fed up with people like me who don't know what the heck they're talking about :D. Marc, heh heh...I consider adding white cheating because it makes some mixtures shift hue towards blue (which can be a useful thing). Instead of adding white, I add water (or some medium) and pull the mixture down thinly to see the undertone...like watercolour.

Regarding why your naphthol red makes better magentas than cadmium reds, I'll be willing to bet it's mainly because it simply has more blue & violet reflectance. To my knowledge, transparency itself does not result in a cleaner mixing pigment...it's just that many of the cleanest-mixing pigments happen to be transparent.

Okay here's my understanding of why I believe, in theory, two ideal primaries could make a secondary as high in chroma as a single-pigment 'ideal' secondary. The diagram below shows the visible spectrum simplified by breaking it up into three 'thirds'; red, green, and blue/violet. Of course in reality, it's a smooth continuum: from near infra red-red-orange-yellow-green-cyan-blue-violet-near ultraviolet...and everything in between, but just for the sake of simplicty here, it's simplified as three thirds.

As you know, when two pigments combine to form a new mixture, the wavelenghts that they both originally reflected (before mixing) will be reflected by the new mix, and those wavelengths that they don't both reflect will be 'destroyed' (absorbed and turned into heat).

The ideal primaries will reflect some wavelengths 100% and the others 0% (unlike actual pigments which never completely absorb or completely reflect any visible wavelength). Here, the ideal magenta is mixed with the ideal yellow. You can see that the only wavelenghts that are commonly relected among them both is the reds. Thus, in the resultant mix, reds are the only group of wavelengths that are reflected. And since this magenta and the yellow both reflect red 100% before being mixed together, doing the math...square root of 100x100...you get 100% reflectance; the new mixture reflects 100% of the group of wavelengths that make up the 'red third'. And those wavelengths that are not reflected by both magenta and yellow get completely 'destroyed' because the square root of 0 times anything equals 0. No single-pigment red could do a better job of maximally stimulating the red cones, and minimally stimulating the green and blue cones. Mixing a green or violet-blue would work the same way. Again, this is my understanding...I stand to be corrected if wrong.

In real pigments, there isn't 100% or 0% reflectance at any visible wavelength...it's somewhere in between 0 and 100%...so if you do the math for every wavelength seperately, you'll end up with a more undulating curve, which means lower saturation.

Also, I don't think the ideal pigments would be able to replicate some characteristics of real pigments, like dual-toned nature, and you'd also have to take into account the effects of the binder, and other stuff, so this 'ideal primary pigment' stuff is still (and will probably always be) just a pipe dream. I'll post my mixed cyan swatch in a few days (the green is oil...I don't want it to smudge my scanner). It made me realize that, in practice, cyan is not as unmixable as I once thought.

Patrick1
02-01-2004, 04:43 PM
Marc, I'd like to clarify: when I said I consider adding white to be cheating and that I only use water or a medium to lighten (by helping to pull out the undertone), I meant when doing a swatch mixing test to see what colours those two colours can mix by themselves...I didn't mean it's cheating to add white to such a mixture in an actual painting. About the only way I can think of to cheat in an actual painting is plagiarism/copyright infringment, etc.

JamieWG
02-01-2004, 05:34 PM
Jamie, that's good news...I assumed he got fed up with people like me who don't know what the heck they're talking about :D.

LOL, Patrick...not at all. :) If you ever have a question for him, just drop him a PM with a link and I'm sure he'd be happy to respond.

Jamie

kbuckland
02-01-2004, 07:35 PM
Wow! I know more now about the "three primary pigments" than I ever dreamed of! So here is the final results of all my palette trials: (I can make the orange perfectly with CadRed and CadYellow, but I can not work with just one red and one yellow)

DioxinePurple RoseMadder CadRedDeep CadYellowMed

UtramarineBlue

Viridian

CadLemon

Plus Flake White
This is working out great! The painting on my homepage of my website, www.kylebucklandfineart.com was painted with this palette. Thank you all so much for the input!
-Kyle

Marc Sabatella
02-01-2004, 11:28 PM
Okay here's my understanding of why I believe, in theory, two ideal primaries could make a secondary as high in chroma as a single-pigment 'ideal' secondary. The diagram below shows the visible spectrum simplified by breaking it up into three 'thirds'; red, green, and blue/violet. Of course in reality, it's a smooth continuum: from near infra red-red-orange-yellow-green-cyan-blue-violet-near ultraviolet...and everything in between, but just for the sake of simplicty here, it's simplified as three thirds.


I've thought about this too. This is what I mean about things working OK when we deliberately simplify to thinking about only disjoint three wavelength regions - red, green, and blue. However, again, in the real world, you've got to come up with an actual reflectance diagram to represent your pure primaries - meaning you've got to pick actual wavelengths where the cutoffs occur. And unfortunately, there is no wavelength you can pick for the boundary between blue and green such that the "blue" wavelengths only trigger the B cones and the G wavelengths only trigger the G cones (the same problem occurs at the other boundaries).

What this means in practice is that while you can indeed mix two of these ideal primaries to get *something*, whether or not that something is really what we'd recognize as what we want is another matter. A blue that resulted from a mixture of magenta and cyan would indeed have 100% reflectance at a given range of wavelengths and 0% at others - yet it would still be likely firing off G cones in our eyes as well, and perhaps some R also, the specifc percentages depending on where we drew the line between B and G wavelengths. Meaning we may or may not actually see the resultant hue as precisely the blue we want. If we're forced to accept thius as the one and only full intensity color to be mixed from these two primaries, I guess that's worth something. But trying to mix any other colors in the cintinuum will prove problematic. Attempts to shift the hue by altering the ratio between the two primaries would result in 100% reflectance in the blue region and a low, flat reflectance across either the entire R or the entire G region - again, probably not what we really wanted.

This is what I mean about not being so sure that C, M, and Y do the job entirely even in theory - although it seems clear they will work better in theory than any other set of three colors.

Patrick1
02-02-2004, 08:36 AM
oops - ignore.

Patrick1
02-02-2004, 08:40 AM
Kyle...I checked your website and there's lots of good work there. Just keep experimenting to see which colours work best for you, and ignore all this theory junk :D.

Marc, you're absolutely correct that for virtually all of the visbible spectrum, any one single wavelength of light will stimulate more than one cone type....so we can't get a truly 'pure' colour sensation from pigments (real or ideal)...regardless of their reflectance curve or the combined reflectance curve.

Making the reflected light 'purer' by restricting its wavelengths to minimize unwanted stimulation of the other cone(s) will make it 'purer' but will also make it darker, resulting in lower apparent chroma.

But if we did want a very 'pure' (saturated, dark) mixed blue for example...one which reflected 100% in a narrow band in the blue region, and 0% elsewhere, would the 'ideal' CM or CMY be able to mix that colour?...and I think this is your point.

Combining of 'ideal' C and M reflectance curves, you couldn't match that narrow blue reflectance curve. Using CM and Y (see diagram), you could make a colour that would stimulate the cones in almost the same proportions. It could match the hue and value. But this mixed blue would be stimulating the green (and even red) cones a bit more than that single-pigment narrow-band blue, since it would would have more blue wavelengths closer to the greens; if you look closely at the diagram, you'll see that the mixed blue fires the green and red cones a little bit more. So this mixed blue would be less saturated and slightly less chromatic than that narrow wavelength single-pigment blue.

Would there be enough of a difference for our eyes to notice a difference between the two? Playing with the following program, it's hard to tell if there's a difference:

http://www.cs.brown.edu/exploratories/freeSoftware/repository/edu/brown/cs/exploratories/applets/spectrum/metamers_java_browser.html

Also you must consider the limitations of computer monitors when using that program.

So I think you're right that in practise, and even in theory, 'ideal' CMY can't match every reflectance curve, or exact RGB cone stimulation (perceived colour) of any imaginable single pigment. Under normal lighting it might look close...maybe even virtually identical for some colours, but not a 100% match.

Trying to mix a colour in the continuum in between a primary and secondary (let's say a hue halfway between magenta and red...roughly a rose colour) will work the same way. Rose would be similar to magenta, only with some yellow added to cancel out some blue reflectance, but not enough yellow added to cancel out all the blue reflectance (and end up with red).

If you had a highly saturated, narrow-band rose pigment, trying to match it with 'ideal' CMY would turn out the same way as trying match that highly saturated, narrow-band blue...you couldn't match it 100%, but possibly very close...maybe even virtually identical to our eyes.

There will be certain reflectance curves which in theory will combine to exactly match a target reflectance curve (mathematically), but these starting colours wouldn't be primaries...so if you want to do it all, three 'ideal' primaries aren't enough. Thanks for pointing this out.

Michael24
02-02-2004, 09:29 AM
Dear Primary Color Folks:

I am not sure why this topic of primary colors is such an obsession with some people. This has all been worked out by color scientist to a high degree. I think artists have taken bits and pieces of this supposed puzzle and twisted it a bit. Here are some of the stumbling blocks that many are pondering.

First, letís stop mixing color theories. Additive and subtractive color theory is being melded into a single unified theory. This is both wrong and confusing to the argument. Leave the rods and cones and green and red creating yellow as a separate entity. This is the additive realm of color. Please donít read any mysticism into Red, Green and Blue creating Cyan, Yellow and Magenta as their secondary colors. These are phenomena of light NOT subtractive mixing of colors.

The true primary colors are not a Holy Grail question. Sorry to disappoint anyone. CMYK was developed by the printing industry because it created the widest color gamut to reproduce color images. It has lots of shortfalls regarding color fidelity and accurate reproduction, but it fits technically and economically in the printing industry.

CMY, specifically Cyan and Magenta are variations on the true primary colors of red, blue and yellow. Cyan is a greenish blue and Magenta is a purple biased red. Again, this shift allows printed material to have a wider gamut in the greens and the violets. As we know, if printers, or artists use cad red medium, cobalt blue and cad yellow light as primaries, the violets and greens suffer from lack of chroma. This is easily explained when you look at a spectral reflectance curve of these colors. (The problem lies in the cobalt blue Ė it has so much red reflectance it dulls any violet made with red and dulls the greens as well when mixed with yellow)

Specific to the questions posed in this thread- No pigment reflects only a narrow band of the electromagnetic spectrum. Pigments reflect a definable percentage of all of the visible wavelengths. Only lasers can broadcast light at narrow wavelengths. So the idea that PURE primary and secondary colors with pigments is unattainable. All pigment mixtures are a product of the reflectance and absorbance of their respective spectra. Pigment mixtures will always be less intense than single pigments because the combined biases of each color will decrease the chroma. (Look at the chroma of a cad orange next to a cad orange made with cad yellow light and cad red medium. Ė The slight green bias in the cad yellow light will dull the mixture slightly)

So, go ahead and use CMY if you wish. It does have a broader gamut because the cyan has very little red reflectivity and the magenta contains purple reflectivity so that mixes of magenta and cyan make strong chroma violets, cyan and yellow mixes make strong greens, and magenta and yellow mixes make good clean oranges.

CMY is still based on the foundation of red, blue and yellow, it just shifts the red and blue to increase the gamut of secondary mixes.

I donít intend this to be argumentative but this topic comes up again and again and we seem to share the same good and bad interpretations of color theory. These topics are not issues with the scientists, paint chemists and paint manufacturers that I have contacted. It just seems to be an irresolvable one with artists in these forums.

Let me know if anything is unclear. This topic is fascinating as an exploration of the physics of paint meeting the needs of artists. I think that if we approach it using the proper color theory, we can advance the discussion in a more meaningful way.


Michael Skalka
Conservation, National Gallery of Art. Washington, DC

JamieWG
02-02-2004, 10:12 AM
Dear Primary Color Folks:
As we know, if printers, or artists use cad red medium, cobalt blue and cad yellow light as primaries, the violets and greens suffer from lack of chroma. This is easily explained when you look at a spectral reflectance curve of these colors. (The problem lies in the cobalt blue Ė it has so much red reflectance it dulls any violet made with red and dulls the greens as well when mixed with yellow)
......................
Let me know if anything is unclear.
Michael Skalka
Conservation, National Gallery of Art. Washington, DC

Hi Michael. Thank you for chiming in. I was hoping you'd poke your head in here!

The only part of what you said that is confusing to me is the part of the paragraph that I quoted above. When I mixed my violet charts many years ago, I mixed every red I had with every blue I had. There were many things that surprised me in the outcomes at that time, but one of the most striking was that the violet mixes seemed far more dependent on the particular reds I used than on the blues. In other words, a magenta or even an alizarin mixed far better violets than some of the other reds, regardless of which blue was used. And the cads mixed the *worst* violets of them all, regardless of which blue was used. What am I missing here?

Jamie

Marc Sabatella
02-02-2004, 12:38 PM
So I think you're right that in practise, and even in theory, 'ideal' CMY can't match every reflectance curve, or exact RGB cone stimulation (perceived colour) of any imaginable single pigment. Under normal lighting it might look close...maybe even virtually identical for some colours, but not a 100% match.


I think this summarizes it well. I would agree that an "ideal" CMY system could mix a specific set of "ideal" other colors - all of which would have the characteristic of flat reflectance across each of the three regions, in varying amounts that could be controlled according to the ratios of the source colors. But I don't see how this system could mix any colors with any other type of reflectance curve. Unless, of course, the basic geometric mean model of substractive color mixing doesn't tell the whole story, which I wouldn't doubt for a second. But any deviation from this might also result in loss of ability to mix perfect reds, greens, and blues.

Truth be told, it's not like I had thought all this through before - I've been pretty much thinking out loud for the last couple of days here. But I think it's been quite interesting. I'm gaining a new appreciation for what is and is not reasonable to expect in color mixing. And after all this, I'm all the more impressed that a relative hack of a model like the color wheel and the idea of mixtures between two color lying along a line between them works *at all* :-)

Richard Saylor
02-02-2004, 12:38 PM
CMY, specifically Cyan and Magenta are variations on the true primary colors of red, blue and yellow.
Why are red, blue, and yellow the true primary colors? Popular opinion? History? Usage?

Michael24
02-02-2004, 12:38 PM
Jamie:

Great Question: I did not focus on the magenta hues so I took a good look at magenta and alizarin crimson and compared it with cad red. I have included the spectra just to show you where I am coming from on my reading of it.

Cadmium red has very little reflectance (absorbtion) - from 380 to 590 nanometers. In other words, it is absorbing the violet, blue and green portions of the spectrum and anything mixed with it will subdue this colorant range. Alizarin and magenta have a small but noteworthy spike in the blue range of the spectrum. This blue spike will aid in mixing with blue to make violet. This is especially true of magenta. Magenta also has less reflectance in the yellow portion of the spectrum than cad red. Magenta starts at 620 nanometers and cad red starts at around 600. Cad red reflects more yellow than magenta. Yellow being the complement of violet, the cad red tends to dull any violet made with it and blue. Magenta and alizarin do not. They mix well with blue reflecting their combined blue characteristics and the clean non-yellowed red they possess.

Thanks for the question. Real life inquiries are nice practical problems to solve. Any other inquiries like this one?

I feel another Powerpoint lecture coming on. This is great food for thought. Let me know if you have more similar questions or need clarification on this one. Sorry for the graphs.

Michael Skalka
Conservation, National Gallery of Art, Washington. DC

Marc Sabatella
02-02-2004, 12:47 PM
When I mixed my violet charts many years ago, I mixed every red I had with every blue I had. There were many things that surprised me in the outcomes at that time, but one of the most striking was that the violet mixes seemed far more dependent on the particular reds I used than on the blues. In other words, a magenta or even an alizarin mixed far better violets than some of the other reds, regardless of which blue was used. And the cads mixed the *worst* violets of them all, regardless of which blue was used. What am I missing here?
Jamie

I think the explanation is simply that there is a larger variety in the actual hue of the reds you were using versus the blues. That is, cadmium red light is much more different than magenta than ultramarine is from phthalo blue. At least, it is in how we perceive it, if not in actual wavelength ratio. Since you were shifting one of the variables by a larger amount than the other, it would make sense to me that the results would be affected more as well.

Michael24
02-02-2004, 02:17 PM
Why are red, blue, and yellow the true primary colors? Popular opinion? History? Usage?


The quick and easy answer: PHYSICS. This is elementary color science. Billmeyer and Saltzman 3rd Edition, edited by Roy Berns, defines subtractive color as red, blue and yellow, or cyan, magenta and yellow. Bear in mind that cyan is still a blue colorant. Magenta is still a red colorant. They both have characteristics that allow them to make cleaner chroma mixes in the secondary color group, especially violets and greens.

Michael Skalka
Conservation, National Gallery of Art, Wash. DC

Patrick1
02-02-2004, 05:47 PM
Leave the rods and cones and green and red creating yellow as a separate entity. This is the additive realm of color. Please don?t read any mysticism into Red, Green and Blue creating Cyan, Yellow and Magenta as their secondary colors. These are phenomena of light NOT subtractive mixing of colors.
You cannot discount the way our eyes and mind perceive light to form the sensation of colour. As Bruce MacEvoy (the Handprint author) says:

"the subtractive "primaries" are only an indirect way to specify the R, G and B cone responses"


The true primary colors are not a Holy Grail question. Sorry to disappoint anyone. CMYK was developed by the printing industry because it created the widest color gamut to reproduce color images. It has lots of shortfalls regarding color fidelity and accurate reproduction, but it fits technically and economically in the printing industry.
The limitations of actual, lightfast 'primary' pigments has been discussed on this forum for years and is no news to any of the regulars on this forum....I did a small mixing exercise several years ago just to get a rough idea of the gamut of combinations of a very few 'primary' pigments. I'm well aware that this 'ideal primary' stuff is a pipe dream, but there's no harm in asking questions like "what if we had a pigment like this...how cleanly could it mix greens?".


CMY, specifically Cyan and Magenta are variations on the true primary colors of red, blue and yellow
Since subtractive primaries are the colours that maximally stimulate two cone types and minimally stimulate the third, red and blue don't meet this criteria. Quoting Mr. MacEvoy again:

"...subtractive color mixing proposes its own three "primary" colors: cyan, magenta and yellow (CMY), as shown in the figure below. (The commonly named subtractive primaries blue, red and yellow are inaccurate and should not be taught.)"

Patrick1
02-02-2004, 06:08 PM
Marc, what you said seems to be right. Yes, the geometric mean doesn't take into account things that will also affect the final colour; like the glossyness of the surface, the type of medium and its refractive index, the paint film's surface scattering, the lighting, and lots of other things. I haven't found any other websites that talk about this.

Michael24
02-03-2004, 08:38 AM
You cannot discount the way our eyes and mind perceive light to form the sensation of colour. As Bruce MacEvoy (the Handprint author) says:

"the subtractive "primaries" are only an indirect way to specify the R, G and B cone responses"

Since subtractive primaries are the colours that maximally stimulate two cone types and minimally stimulate the third, red and blue don't meet this criteria. Quoting Mr. MacEvoy again:

"...subtractive color mixing proposes its own three "primary" colors: cyan, magenta and yellow (CMY), as shown in the figure below. (The commonly named subtractive primaries blue, red and yellow are inaccurate and should not be taught.)"


Dear Patrick

I am not sure if you are debating or supporting my argument.

Regarding the rods and cones and RGB related to the eye: Too many discussion in this forum use the RGB stimulus and the eye to lay down a faulty argument that certain color combinations cannot be made. In fact, a whole 7-8 pages was devoted to an argument that blue and yellow cannot make green based on the presence or absence of RGB stimulation in the eye. So when you say you cannot discount what happens in the eye, I disagree to an extent when the argument is used to refute a subtractive mix like yellow + blue = green. Bottom line, you donít need to know the mechanics of sight to explain or manipulate subtractive color to get the results you desire. (Short of having a vision defect)

I think we are saying the same thing with regards to cyan, yellow and magenta. While the Handprint website proclaims that red, blue and yellow should not be taught as primaries, MacEvoy seems to be denying that cyan is a blue hue and magenta is a red hue.

It seems to make some folks happy to refute red, blue and yellow and replace it with the names magenta, cyan and yellow and think that they have walked away from blue and red leaving it behind as an antiquated color mixing construct.

Finally, sorry, my colleagues and I donít take much stock in the Handprint site. From a scholarly viewpoint, it contains a lot of voids, mixes additive and subtractive color models Ė especially in the use of CIE Lab as a framework for subtractive pigment relationships. CIE Lab is an additive light model. Its great for what is was made for but leaves something to be desired in mapping artists pigments. Its unequal spacing makes it faulty for a lot of color plotting in the subtractive realm. The site contains a lot of very useful information for the watercolorist, but one has to subtract the facts from the personal opinion or questionable material that uses inappropriate color theory or models.

Michael Skalka
Conservation, National Gallery of Art, Washington, DC

Marc Sabatella
02-03-2004, 04:56 PM
In fact, a whole 7-8 pages was devoted to an argument that blue and yellow cannot make green based on the presence or absence of RGB stimulation in the eye. So when you say you cannot discount what happens in the eye, I disagree to an extent when the argument is used to refute a subtractive mix like yellow + blue = green.


There is nothing faulty about this. Pick the right blue and the right yellow - ones with 100% reflectance in the corresponding regions and 0% in the other regions - and you *won't* get green. You'll get black. It's just that in practice, there are no pigments like this. Still, understanding how this works can be quite relevant in understanding why some pigments behave different in mixtures than others.


I think we are saying the same thing with regards to cyan, yellow and magenta. While the Handprint website proclaims that red, blue and yellow should not be taught as primaries, MacEvoy seems to be denying that cyan is a blue hue and magenta is a red hue.


I also consider this a relevant question. *Why* should we take red, blue, and yellow to be the primaries, and consider cyan and magenta to be just different shades of these? because that's how we're used to describing colors? If we are going to discuss primary colors at all, we should start by agreeing on what attributes a primary color should have. Right now, it is painfully clear there is no such agreement.

Patrick1
02-03-2004, 05:32 PM
Michael, thanks for your input on this discussion. It's nice some folks like to talk about the 'nuts & bolts' of colour.

On this forum and in many other places, it has been said that according to theory, blue + yellow pigments make black. I'd often suggest that this is not the case with real, available paints...for at least two or three reasons I can think of. The problem is not with theory per se, but the vast oversimplification (actually worse than oversimplification) when applying RGB theory in order to make subtractive mixing look cleaner and simpler than it really is.

To my knowledge, a 'true' cyan would be no more blue than green, a 'true' magenta no more red than blue. It's just that the 'best' lightfast 'cyans' we have, like PB 17, PB 15:3, PB 15:4 happen to be displaced in hue towards blue, and the magentas toward red, and we can take advantage of those characteristcs (to lessen the saturation costs of mixed reds for example). But if we could somehow develop C and M pigments of any desired reflectance curve for the sake of maximizing overall gamut from only 3 pigments, it would be very close to CMY, not RYB. So for now, we live with what we've got and make best use of it. Of course, the largest gamut is not always an artist's priority, so a more artist-friendly RYB is often preferred...especially if reds & oranges are a priority...this is one reason RYB lives on.

Handprint points out that it is impossible to make a subtractive colour model or wheel to accurately show mixing relationships among real pigments. That's why Bruce doesn't even try to have a true subtractive mixing wheel/model...he admits his Artists' Color Wheel is primarily a visual (additive) wheel. (though I will admit that it blurs the lines between additive and subtractive to an extent). For those who need more accurate mixing complement info, it's there seperately in chart form.

Honestly, I've yet to find a web site or book that better bridges colour theory with artistic application. Most colour theory books I've seen in book or art stores are lots of fluff and even half-truths, like "primary colours cannot be mixed from any other colours", "warm colours are inviting", "cool are calming"...

Patrick1
02-03-2004, 06:06 PM
Hi Marc...as I've learned recently, even if paints could have 100% and 0% reflectance in the desired places, the blue reflectance could result in stimulting the adjacent G cone (heh heh...that sounds funny). Only if you could get very restrictive bands of reflectance could this improve. But that would make paints dark. I'd think such a demonstration should work better with light transmission filters and a bright light.

Bad Dobby
02-03-2004, 06:25 PM
Of course, the largest gamut is not always an artist's priority, so a more artist-friendly RYB is often preferred...especially if reds & oranges are a priority...this is one reason RYB lives on.


Thank you ... this was my point early on in this thread. There is the scientifically and gamut-rich set of primaries in theory ... and can be another set of primaries prefered in actual practice (they can be the same .. but usually are not). In practice the 3 primaries can be any three colors that the Artists elects.

Patrick1
02-04-2004, 07:31 AM
I wrote:...even if paints could have 100% and 0% reflectance in the desired places, the blue reflectance could result in stimulting the adjacent G cone...
Ooops: I was considering the RGB stimulation of the individual paints before they were mixed...I should've considered it only from the final mixture. My aplologies.

Michael24
02-04-2004, 08:46 AM
Michael, thanks for your input on this discussion. It's nice some folks like to talk about the 'nuts & bolts' of colour.

On this forum and in many other places, it has been said that according to theory, blue + yellow pigments make black. I'd often suggest that this is not the case with real, available paints...for at least two or three reasons I can think of. The problem is not with theory per se, but the vast oversimplification (actually worse than oversimplification) when applying RGB theory in order to make subtractive mixing look cleaner and simpler than it really is.



I agree. I have lots of spectra to indicate that in the subtractive mixing world that blue + yellow will NOT yield black.


To my knowledge, a 'true' cyan would be no more blue than green, a 'true' magenta no more red than blue. It's just that the 'best' lightfast 'cyans' we have, like PB 17, PB 15:3, PB 15:4 happen to be displaced in hue towards blue, and the magentas toward red, and we can take advantage of those characteristcs (to lessen the saturation costs of mixed reds for example). But if we could somehow develop C and M pigments of any desired reflectance curve for the sake of maximizing overall gamut from only 3 pigments, it would be very close to CMY, not RYB. So for now, we live with what we've got and make best use of it. Of course, the largest gamut is not always an artist's priority, so a more artist-friendly RYB is often preferred...especially if reds & oranges are a priority...this is one reason RYB lives on.


The spectra on cyans I have recorded, PB15 - phthalocyanine pigments, indicate a blue green spectra that is predominantly blue. They are much cleaner than cobalt or ultramarine in that the phthalocyanines do not have a red spectral reflection. They mix cleaner violets and greens because of this. Magenta's spectral response curve looks like a violet-blue and red mixture. The reflectance in each make it respond well to purple as well as decent red and orange mixtures. I think the gamut issue is more that they are type 2 (low scattering pigments) small, uniform particles with high chroma.

Yes, CMY does yield a wider gamut. That's why it is so useful to the color printing industry, but harder for example, the landscape artist to use, without having to subdue the high chroma.

You are right. We do need a better color theory and mixing book. Especially one that does not combine additive and subtractive color.

Michael Skalka
Conservation, National Gallery of Art, Wash. DC

Richard Saylor
02-04-2004, 10:12 AM
Is there a good source for the spectral reflectance curves of artist's pigments besides Hilary Page's book on watercolor paints?

Michael24
02-04-2004, 01:38 PM
Is there a good source for the spectral reflectance curves of artist's pigments besides Hilary Page's book on watercolor paints?


I wish I could say that one exists. The spectral curves in the latest versions of Ralph Mayer are lame since they have no X or Y legend markings. I have about 150 spectra for paints so far. As the samples from the materials collection I oversee at the National Gallery grows, I will have thousands of spectra within the next 5-6 years. If they are ever published it will be web or CD based. The audience will be far too small for hardcopy publishing.

Michael Skalka
Conservation, National Gallery of Art. Wash. DC

Michael24
02-04-2004, 01:58 PM
There is nothing faulty about this. Pick the right blue and the right yellow - ones with 100% reflectance in the corresponding regions and 0% in the other regions - and you *won't* get green. You'll get black. It's just that in practice, there are no pigments like this. Still, understanding how this works can be quite relevant in understanding why some pigments behave different in mixtures than others.
painfully clear there is no such agreement.

Curious! Is this some sort of intellectual or philosophical excercise that leads nowhere? * Pick the right blue and the right yellow* The closest one might come is to combine laser light at a blue and yellow frequency. But there again, we are dealing with light and not subtractive pigments. Sorry, I don't agree that this is relevant and leads to a better understanding of pigment behavior. It does yield information on the biology of color perception. Much research has been done on this subject and lots of books are devoted to subjective color perception that has been objectively measured and recorded. I am sure you have read much of this. It is wonderful stuff.

Better to devote ones time to focusing on the actual properties of pigments. My work with spectral reflectance curves of artists pigments has provided extremely practical answers as to why mixtures behave the way they do. The subtle shifts and sometimes dramatic shifts in absorbtion or reflectance when studying single pigments and combined pigments yields very useful data. One does not have to get hung up on cone response to devine the answer to some pigment mixing questions. The reflectance spectra provide most of the information needed. It does have voids. It doesn't account for scatter or chroma. However, one does not mind since the reflectance curve contains so much other useful data.

Thanks for your comments.

Sincerely,

Michael Skalka,
Conservation, National Gallery of Art, Wash. DC

JamieWG
02-04-2004, 04:05 PM
Better to devote ones time to focusing on the actual properties of pigments.......... It doesn't account for scatter or chroma.

I agree about devoting time to actual paints, Michael. What is "scatter"?

Thank you for the reply and charts that you posted to this thread a couple of days ago, in response to my observations about mixing blues and reds. I can see both in the pigments and the charts *why* they would mix different violets. What was baffling to me is why the blues mattered so little. IOW, the violets I mixed with alizarin all looked about the same, regardless of which blue I used. The violets that I mixed with cad red (well, more like mud than violet!) all looked about the same too. Ditto with napthol. Whether a blue was cerulean or ultramarine seemed to matter a heck of a lot less than whether the red was cobalt or alizarin. I don't remember if I did magenta at that time. I'll have to pull a Search and Rescue mission in my studio and see if I can unearth those old charts.

Jamie

Michael24
02-04-2004, 05:15 PM
Jamie:
Scatter is the property of a pigment that dispurses reflected light in a multitude of directions. Titanium White is a classic scatterer. It subdues the chroma and *greys* inoragnic pigments.

Michael Skalka
Conservation, National Gallery of Art, Wash. DC

Richard Saylor
02-04-2004, 06:49 PM
Curious! Is this some sort of intellectual or philosophical excercise that leads nowhere? * Pick the right blue and the right yellow* The closest one might come is to combine laser light at a blue and yellow frequency. But there again, we are dealing with light and not subtractive pigments. Sorry, I don't agree that this is relevant and leads to a better understanding of pigment behavior. It does yield information on the biology of color perception. Much research has been done on this subject and lots of books are devoted to subjective color perception that has been objectively measured and recorded. I am sure you have read much of this. It is wonderful stuff.

Better to devote ones time to focusing on the actual properties of pigments. My work with spectral reflectance curves of artists pigments has provided extremely practical answers as to why mixtures behave the way they do. The subtle shifts and sometimes dramatic shifts in absorbtion or reflectance when studying single pigments and combined pigments yields very useful data. One does not have to get hung up on cone response to devine the answer to some pigment mixing questions. The reflectance spectra provide most of the information needed. It does have voids. It doesn't account for scatter or chroma. However, one does not mind since the reflectance curve contains so much other useful data.

Thanks for your comments.

Sincerely,

Michael Skalka,
Conservation, National Gallery of Art, Wash. DC
All Marc was doing was describing in a non-technical way what happens if you mix certain yellow and blue colors whose reflectance curves do not overlap. It's like Einstein's Gedanken experiments, such as taking an elevator so far out into space that it is free of all gravitational fields and then accelerating it, thereby perfectly simulating the effect of gravity. Such pigments (and such crazy elevators) do not exist, but the exercise does not "lead to nowhere" if it helps our understanding of what's going on.

It almost seems as if you want to toss out color theory, yet the theoretical properties of relatively narrow bands of wavelengths in the visible spectrum are useful, because those bands can be pieced together by the principle of superposition to yield useful information about pigments, provided you know their reflectance curves.

Bad Dobby
02-04-2004, 07:38 PM
Michael24 .... happy to see someone from a Conservation position. Seems that whether we are talking about color, paints, supports, mediums/varnishes or whatever .... conservation is a key concern. Looking forward to following your participatioin here at WC.

Dobby :)

Marc Sabatella
02-04-2004, 10:04 PM
Curious! Is this some sort of intellectual or philosophical excercise that leads nowhere? * Pick the right blue and the right yellow* The closest one might come is to combine laser light at a blue and yellow frequency. But there again, we are dealing with light and not subtractive pigments.


That may be the closest you can come among things we currently know about. But is it really so hard to imagine that somewhere in the world there might be heretofore unknown pigments that would reflect only a range of blue wavelengths (not just one, but a range) with only a negligible amount of green, and a yellow that similarly reflected only a negligible amount of green? And would it not be nice to not be surprised when they failed to make any recognizable green when mixed? I agree, understanding what happens in the real world is ultimately more important, but understanding *why* it happens is pretty darn useful too, I think. And sometimes, to understand the why of things, we have to look at them in a more theoretical and less practical sense. This doesn't bother me. I know full well than no known blue and yellow pigments are so narrow that they would not make green, but understanding the principle that makes it so strikes me as quite interesting and important - and this same principle is the one that predicts the absence of green in a mixture of the as yet unknown blue and yellow pigments I described.

JamieWG
02-05-2004, 07:08 AM
Whether a blue was cerulean or ultramarine seemed to matter a heck of a lot less than whether the red was cobalt or alizarin.
Jamie

Oops...I meant to say, "cadmium or alizarin," of course. I need more sleep!

Thank you, Michael, for your "scatter" explanation. :)

Jamie

Patrick1
02-05-2004, 09:48 AM
Michael, you said: "The spectra on cyans I have recorded, PB15 - phthalocyanine pigments, indicate a blue green spectra that is predominantly blue. They are much cleaner than cobalt or ultramarine in that the phthalocyanines do not have a red spectral reflection. They mix cleaner violets and greens because of this."

So you're saying that phthalo blues mix cleaner violets than ultramarine blue? Even colour bias theory says that ultramarine will mix cleaner violets. This is also why ultramarine is often the violet-blue of choice in a 'split-primary' palette.

Michael24
02-05-2004, 04:52 PM
So you're saying that phthalo blues mix cleaner violets than ultramarine blue? Even colour bias theory says that ultramarine will mix cleaner violets. This is also why ultramarine is often the violet-blue of choice in a 'split-primary' palette.

I guess this is more personal preference than spectral science. Phthalo, with its intensity and lack of a red tail makes what I think is a cooler purple. The ultramarine and red mixture seems to be a purple with a red bias. I like both. It's more dependent on which red you use. Alizarin and ultramarine make great violets. Cad red and ultra blue are less intense. Cad red being opaque makes an opaque violet. Again, personal preference, I like violets that are transparent rather than opaque, so my own preference for intense, transparent violet leans toward the use of phthalo blue. As a color of regular use, I pick ultramarine every time I paint. I am not picking on ultramarine blue, I really like it. I guess what I meant to say was that phthalo has less bias issues because of its low reflectance in the red portion of the spectra than ultramarine blue does. You are right, the red in ultramarine does make great violets. It's just harder to make them without a red bias when using ultramarine.

Ultramarine mixtures with yellow suffer from dullness when the red tail in ultramarine greys the green.

As I collect more and more spectra, I begin to notice that the things that people complain about when mixing colors, are indicated in the spectra of the individual and mixed paints. It is sometimes very subtle, but a logical reason can be derived from a spectral reflectance curve.

Hope this makes sense.

Michael Skalka
National Gallery of Art, Wash. DC

JamieWG
02-05-2004, 05:39 PM
I guess this is more personal preference than spectral science. Phthalo, with its intensity and lack of a red tail makes what I think is a cooler purple. The ultramarine and red mixture seems to be a purple with a red bias. I like both. It's more dependent on which red you use.

Well, this certainly helps me to understand why, when I did those mixes long ago, I thought *all* the blues produced decent violets, but few of the reds. Since violet is my favorite color, and I have a rare free evening tonight, I'm going to play with paint and make some charts. 'Will post them when I can get a good pic with my husband's camera, which unlike mine, adores portraying reds and violets.

Jamie

WFMartin
02-05-2004, 06:54 PM
Color theorist that I am, I am quite interested in discussions such as this. Let me offer my opinion on primaries. Without going into the many technical and scientific ways of proving it (and there are many), the primary colors of pigment are cyan, magenta, and yellow.

Primaries are designated as such because they exhibit certain properties that other secondarys and "convenience colors" don't. One of these properties is that many other colors can be mixed from them, among them red, green, and blue (violet, if you choose to call it that). Secondaries and convenience colors cannot mix the many other colors that primaries can, nor should they be expected to.

One cannot simply decide on your own to rotate points uniformly around a color wheel, and, as a result, simply designate other colors as primaries. Well, I suppose you CAN call other colors primaries, but they cannot ever be expected to produce the myriad colors that primaries, and primaries alone can produce.

The result of assigning to colors which aren't primaries a "primary" status, is that they won't perform as primaries, and colors mixed from them will not produce secondaries. Those colors that they DO produce will be somewhat grayed. The farther you rotate this assignment of "arbitrary primaries" around the wheel, the closer you come to assigning the secondaries a "primary" status, and the grayer will be the resulting mix from these new colors you decided to call "primaries". For example, let's consider a color wheel, composed of cyan, magenta, yellow, red, green, and blue. On that color wheel as well as many others, the color yellow is midway between red and green. However, if one attempts to mix red and green to produce yellow the results are a dirty gray. Red and green are secondaries and cannot be expected to mix anything other than gray. One cannot consider secondaries to be primaries. They don't behave like primaries, and never will.

As I always say, one may call any color a primary that one may wish to, but that doesn't make it a primary. Let me offer as an analogy that one can call a horse a pig, but don't expect it to yield pork chops. ;)

Bill :)

Marc Sabatella
02-06-2004, 01:05 PM
Color theorist that I am, I am quite interested in discussions such as this.


I'd be really curiouis to hear your responses to the many excellent points made on this thread, then. We really got pretty in depth into looking at some of the issues, trying to get more specific about what colors an ideal C, M, and Y can and cannot mix, the effect of the fact that no pigments actually reflect the way the ideal C, M, and Y colors would, and so forth. It's been quite an illuminating discussion, and most of your points here have already been discussed at some length.
[/QUOTE]

JamieWG
02-06-2004, 01:10 PM
As promised....Sorry about the slight blurriness; I'll get another photo on a better day. It's heavily overcast outside and we're in the middle of an ice storm, so this was take by artificial light...but I think you can get the idea nevertheless!

http://www.wetcanvas.com/Community/images/06-Feb-2004/13766-Purple_Chart_500.jpg

There may be one or two violets in the colums that look very different from the others; that's due to my hurried mixing and I'm sure those could be tweaked redder/bluer to match more closely with the others in their column. But you can clearly see that the violet "families" are identifiable by the reds, and not by the blues!

Jamie

JamieWG
02-06-2004, 03:58 PM
Oh---I forgot to say that for consistency's sake, I used all WN Artist Oils for the tests, except:

Gamblin Magenta (it was the only PR122 I had)
Gamblin Indanthrone Blue
Gamblin Ultramarine
Classic Artist Oils Phthalo Blue

Napthol PR170 was a big surprise to me. Marc, did you say that you always use this pigment instead of cad red? Which brand and PR# are you using, and why did you choose it over cadmium (besides price)? It is a much cleaner-mixer than cad red or cad red light.

Jamie

Richard Saylor
02-06-2004, 04:57 PM
Without going into the many technical and scientific ways of proving it (and there are many), the primary colors of pigment are cyan, magenta, and yellow.
This could get interesting, since we have been informed that "CMY, specifically Cyan and Magenta are variations on the true primary colors of red, blue and yellow." Moreover, "It seems to make some folks happy to refute red, blue and yellow and replace it with the names magenta, cyan and yellow and think that they have walked away from blue and red leaving it behind as an antiquated color mixing construct."

WFMartin
02-06-2004, 08:28 PM
This could get interesting, since we have been informed that "CMY, specifically Cyan and Magenta are variations on the true primary colors of red, blue and yellow." Moreover, "It seems to make some folks happy to refute red, blue and yellow and replace it with the names magenta, cyan and yellow and think that they have walked away from blue and red leaving it behind as an antiquated color mixing construct."

cmyguy,

Ah, yes, but in this case we must realize that we are actually dealing with different colors! Ideal cyan is not blue (at least not any more than it is green, as it reflects blue and green light equally); it is a different color than blue, not just a different name.

Ideal magenta is not red (at least not any more red than it is blue, as it reflects red and blue light equally); it is a different color than red, not just another way of naming "red".

Again, you may wish to call magenta "red" (or some sort of a variation of red), but it is magenta, a color with distinctly different reflectance than red. And, once again calling magenta "red" doesn't make it red--nor does calling it "red" magically force it to behave like red.

Yellow is the third primary of pigment. If some artists wish to think of cyan as being a form of "blue" or magenta as being a variation of "red", then why don't these same artists think of yellow as being a variation of "green", or a variation of "red"? The ideal primary of yellow reflects as much green light as it does red light, so does that make it a variation of one or the other? No, I contend, it does not. It is a unique color unto itself, and that's one trait that makes it a PRIMARY.

Cyan is not a variation of blue; it's a primary color, "cyan".
Magenta is not a variation of red; it's a primary color, "magenta".
Yellow is not a variation of red (or green, either); it's a primary color, "yellow".

Bill ;)

JamieWG
02-07-2004, 08:25 AM
Hey, isn't anybody going to drool over my beautiful color charts? :D

I can easily see both sides of the Primary Colors argument. Although CYM works for printing purposes, it is impractical for painting purposes. I know because I spent seven months with it last year. If you can imagine the hell of trying to mix colors from CYM while painting plein air landscapes almost every single day, for nearly three seasons, you know what my life was like back then. Never again!!! I did it to learn from it and in that regard it was very successful. I ended up supplementing a bit with an orange and a violet (for convenience), but there's no longer any doubt in my mind about why artists choose RBY "primaries" over CYM.

Jamie

Eugene Veszely
02-07-2004, 09:11 AM
...Information overload ;)

Marc Sabatella
02-07-2004, 02:40 PM
Napthol PR170 was a big surprise to me. Marc, did you say that you always use this pigment instead of cad red? Which brand and PR# are you using, and why did you choose it over cadmium (besides price)? It is a much cleaner-mixer than cad red or cad red light.


I started using it in watercolor because I hated the opacity of the cadmiums - red passages just looked completely out of place in my paintings. I had heard someone mention naphthol as a good alternative, so I gave it a shot, and liked it well enough - it was much easier to get a consistent look within a painting. When I started oil painting, I was using borrowed paints, including a cadmium red, but when I starting buying my own, and saw how much cadmiums cost, and knew I was happy with naphthol from watercolor, so I tried a tube in oils - W&N's "Bright Red", which at the time I bought it was PR170 (they've used a couple of different pigments for this color recently, apparently). For a while, I would occasionally switch back to the cadmium (which I hadn't run out of), just to see if I was missing something, but never found that I was.

When I switched to MGraham paints, I started out thinking I didn't even need this color - I was going to use something resembling a CMY palette - but when I finally got frustrated to my ability to mix the kind of oranges I could get with easily with cadmiums or naphthol, I picked up a tube of that - PR112, in MGraham's case. As it is, I would prefer something slightly more orange, but not enough to actually buy a tube of cadmium orange.

JamieWG
02-07-2004, 03:26 PM
........ I was happy with naphthol from watercolor, so I tried a tube in oils - W&N's "Bright Red", which at the time I bought it was PR170

That's exactly what I used in my test---WN Bright Red, PR170.

Thank you for your explanation of why you use that instead of cad red. I find that cad red goes a very long way, so I don't need much of it. I had that tube of WN Bright Red for ages and ages, but never really gave it a chance except on some charts...'just didn't need it. I think I'll be more inclined to use it now, following these more recent tests.

Jamie

Patrick1
02-07-2004, 03:54 PM
Hey, isn't anybody going to drool over my beautiful color charts? :D

I love it...it's a lot neater than my colour mixing swatches & charts. It's striking how the mixtures with cad red are so distinctly greyer than any other. Mixed greys like this can be very beautiful in the context of an actual painting.

Richard Saylor
02-07-2004, 11:20 PM
If you can imagine the hell of trying to mix colors from CYM while painting plein air landscapes almost every single day, for nearly three seasons, you know what my life was like back then. Never again!!!
I can certainly imagine the stress and inconvenience of taking the time to mix colors from CMY, because with plein air one is essentially in a race with the ever-changing light. I wouldn't even attempt it myself. Fortunately with watercolor the most commonly used mixtures can be made up ahead of time, because if they dry out they can be quickly rehydrated.

WFMartin
02-08-2004, 12:14 AM
Jamie,

Knowing how primary colors behave and work with each other doesn't mean you are somehow obligated to paint with them.

I use cyan, magenta, and yellow primary paints very seldom, but that hasn't prevented me from attempting to gain knowledge regarding just how they behave as ingredients in all those other secondary and convenience colors. The knowledge of primary behavior helps in mixing other convenience colors with each other, or even in selecting a new tube of paint color for what its color properties are, based upon the primaries that it contains. For example, if you can recognize a burnt umber as containing more magenta than, say, raw umber, you will select it for mixes which require a bit more magenta content than raw umber. Although anyone can mix the color, burnt umber, from cyan, magenta, and yellow, why bother? Reach for a tube of burnt umber, instead. The "primary police" surely won't be watching you. ;)

They call convenience colors by that term because they are just that--convenient. Although I can mix an orange so close to cadmium orange that you couldn't tell them apart, I surely would not waste my time doing that every time I need the color, orange. But, it's surely nice to know in advance just what choice of cyan, magenta, or yellow I'd need to employ to achieve it if I had to.

If I were a plein aire painter as you are, I certainly would not waste my time proving to the world that three colors is all one needs to create every color required. I'd buy every practical convenience color available, and use 'em all.

Bill :)

JamieWG
02-08-2004, 07:20 AM
Jamie, The knowledge of primary behavior helps in mixing other convenience colors with each other, or even in selecting a new tube of paint color for what its color properties are, based upon the primaries that it contains.

Exactly. This is why I tortured myself for three seasons. :)

If I were a plein aire painter as you are, I certainly would not waste my time proving to the world that three colors is all one needs to create every color required.

I agree, Bill, but I needed to try. :D Actually, many plein air painters use only three or four colors. However, the ones they choose are generally alizarin, ultramarine, and cad yellow---much easier "primaries" to work with, regardless of what you call them. I know you don't like those being called primaries, and technically you are right.

Jamie

Jamie

WFMartin
02-08-2004, 11:32 AM
Exactly. This is why I tortured myself for three seasons. :)



I agree, Bill, but I needed to try. :D Actually, many plein air painters use only three or four colors. However, the ones they choose are generally alizarin, ultramarine, and cad yellow---much easier "primaries" to work with, regardless of what you call them. I know you don't like those being called primaries, and technically you are right.

Jamie

Jamie

Jamie,

Good points. I hope you would also agree that the "torturing" you did to yourself when you were forcing yourself to use colors that were very close to true primaries really helped you to gain knowledge of just how the primaries work. Would you agree that the knowledge gained by that exercise helped you in your selection of tube colors, as well as selecting which colors on your palette to mix together to achieve the intermediate colors for which you may be aiming?

I remember the painting I did for the "primary" project for which I believe you got recommendations from W & N regarding the specific colors they considered closest to cmy. I had the most fun I ever had in painting a subject. I purposely selected one of the reference photos that contained many colors just to purposely make it difficult on myself. As I was going into that project I was nearly temped to believe what others had contended that it was impossible to produce bright reds, oranges, greens, and violets with colors that were primaries.

Well any fears I may have had in that regard were put to rest, as I began to witness some of the most vivid, bright, clean colors emerging on my painting from those three primaries you recommended. Being as steeped in color theory as I am, I had believed it would work, but I actually never believed that it would work SO WELL. All the claims that one cannot mix clean oranges or reds by using primary yellow and magenta proved to be completely false, in my opinion.

Once again, I will have to qualify my statements in that we as artists do not have our hands tied by "color theory", but only by the deficiencies of pigment in meeting the demands of what characterizes primary colors.

Bill :)

Marc Sabatella
02-08-2004, 10:22 PM
I can certainly imagine the stress and inconvenience of taking the time to mix colors from CMY, because with plein air one is essentially in a race with the ever-changing light. I wouldn't even attempt it myself. Fortunately with watercolor the most commonly used mixtures can be made up ahead of time, because if they dry out they can be quickly rehydrated.

FWIW, although I'm not currently using CMY, I am still using a very simple color wheel type of palette for plein air work, and like it for probably the same reason others *don't* like it - you get pretty high intensity colors by default. This matches how I tend to see and want to paint. In pastels, I have and use way more high intensity colors than earth tones. In oil, I have to struggle to avoid making my colors duller than I intend. And once I expanded my palette from 3 to 5 colors, I felt I could get pretty much any hue I wanted more than quickly enough.

If, on the other hand, you were going after more realistic - which is to say, mostly dull - color, then I can see this type of palette as being more work than you'd want.

Marc Sabatella
02-08-2004, 10:33 PM
Good points. I hope you would also agree that the "torturing" you did to yourself when you were forcing yourself to use colors that were very close to true primaries really helped you to gain knowledge of just how the primaries work. Would you agree that the knowledge gained by that exercise helped you in your selection of tube colors, as well as selecting which colors on your palette to mix together to achieve the intermediate colors for which you may be aiming?


I can't speak for Jamie, but I would definitely agree with this. Understanding more about color theory - from reading, thinking, and experimenting - has definitely enabled me to settle on a palette that I think I can pretty safely say I would not have hit upon otherwise.


All the claims that one cannot mix clean oranges or reds by using primary yellow and magenta proved to be completely false, in my opinion.


In my experience, it isn't so much that oranges *can't* be mixed from magenta and yellow - although I would argue that the oranges that result *are* noticeably duller than cadmium, to my eyes. And again, I like my colors somewhat more intense than nature, so I would consider decrease in intensity in my paints to be problematic. But the oranges I got were indeed good enough for most purposes. Instead, this was the one place where the time factor really came into play - orange seems especially sensitive to getting just the right balance of yellow and magenta, and when I needed it lighter, adding white would shift the hue again. I spent more time trying to mix a good orange than any other color.

WFMartin
02-08-2004, 11:49 PM
I can't speak for Jamie, but I would definitely agree with this. Understanding more about color theory - from reading, thinking, and experimenting - has definitely enabled me to settle on a palette that I think I can pretty safely say I would not have hit upon otherwise.



In my experience, it isn't so much that oranges *can't* be mixed from magenta and yellow - although I would argue that the oranges that result *are* noticeably duller than cadmium, to my eyes. And again, I like my colors somewhat more intense than nature, so I would consider decrease in intensity in my paints to be problematic. But the oranges I got were indeed good enough for most purposes. Instead, this was the one place where the time factor really came into play - orange seems especially sensitive to getting just the right balance of yellow and magenta, and when I needed it lighter, adding white would shift the hue again. I spent more time trying to mix a good orange than any other color.

Marc,

I believe you are absolutely correct in your assessment of the mixing of orange. Getting the correct balance of yellow and magenta to approximate cadmium orange is time consuming.

Also (and this is important and quite interesting), certain tube colors exhibit an "overtone" which actually shifts them toward adjacent hues when mixed with white. Tube Cadmium orange, for example, shifts toward red when mixed with Flake White. In other words, you won't get a light orange by mixing white with orange; you'll get pink! Now, the mix of yellow and magenta that produces the masstone color approximating cadmium orange won't do that. In fact, if I recall, it shifts decidedly toward yellow! I tried it, and it certainly doesn't exhibit the same overtone.

Unusual things happen when you mix certain colors with white. Instead of simply "graying" the color, as others believe (and I used to believe), many times the addition of white actually shifts the hue of the color in question toward an adjacent hue.

So, those who dislike mixing secondaries from primaries definitely have a point in the fact that those "mixed" secondaries do not exhibit the same results when mixed with white that a tube color of that same secondary would. Not to say that you can't eventually create the color you want, but it often might not be a simple two-color mix, and will require a degree of adjusting to get the color you want after mixing white with it.

Bill :)

Michael24
02-09-2004, 09:26 AM
Bill and Marc:

First, my apologies if my debate on additive and subtractive theory or pure primary colors seemed too heavy or argumentative. Both of you have a wonderful knowledge base and command of color. I get a bit frustrated that so much theory gets misinterpreted and sends a group of uninitiated artist out spreading a faulty theory to others.

Marc: I see your point on blue + yellow = black. Again, purely in an abstract way. I don't think one will ever find examples in the real world of pigments that contain a mono-spectral signature. I have lectured to a lot of artists and a lot of the abstract and scientific material on color is less interesting to them than practical mixing applications.

In addressing the issue of a single pigment orange verses a mixed orange, I have an example using cad orange and a mixed orange made of cad yellow light and cad red medium. Since I know you love this stuff, I have even attached a graph. If you could see cad yellow light by itself, it reflects a fairly healthy portion of the green part of the spectrum. (The organic yellows for the most part, reflect even more green than cad yellow lt.) The mixture of cad yellow light and cad red, will of course, not be as high in chroma as the single pigment cad orange. The green dulls the red in the mixture and lowers its chroma. That small hump in the mixture is the green reflectance. Using CMY, you see this phenomenon since the green reflectance in many organic yellows reacts with the redness in the magenta to take some of the chroma down a bit in the mixture. Granted, relative to cad orange it is still a fairly high key orange. Need more graphs, I have hundreds of them.

Michael Skalka
Conservation, National Gallery of Art. Wash. DC

Marc Sabatella
02-09-2004, 05:53 PM
I have lectured to a lot of artists and a lot of the abstract and scientific material on color is less interesting to them than practical mixing applications.


Presumably you've figured out by now that those of us still tuned in to this thread have more of an interest in the theory than is typical :-)

Thanks for the graph of the mixed orange. This definitely fits with my assumptions about would be happening. One thing I wish is that it were possible to draw a graph, plug it into some sort of machine, and out would pop a paint swatch that generates that reflectance...


If you could see cad yellow light by itself, it reflects a fairly healthy portion of the green part of the spectrum.


Of course, that's the whole point in CMY theory - you want your yellow to reflect the green and red regions fully, with no blue reflectance. As I see it, the green reflectance in the yellow pigment wouldn't be an issue if there weren't also some green reflectance in the red pigment. If the red truly had no green in it (as with the idealized magenta), it would absorb all the green in the yellow without affecting the red. As it is, I'll bet there is just enough green in the red pigment to allow all the green in the yellow to cause noticeable trouble. For something like cadmium red, most of the green is probably at the yellow end. For most magenta pigments, there is probably a bit of green at the blue end.

Bad Dobby
02-09-2004, 06:34 PM
Presumably you've figured out by now that those of us still tuned in to this thread have more of an interest in the theory than is typical :-)


Only because those of us that acknowledged the importance of theory but shifted the discussion to practical mixing applications and alternatives were chided out of the discussion early on. Hehehe, I don't mind, however ... I am one of those that believe that conflict is necessary in the furtherance of knowledge, and have enjoyed the directions of the thread! Theory is great ... and I think most of us have a sufficient working knowledge in this area ... but when the pigments are on the palette or the canvas, that is where the rubber hits the road. When it comes to painting, I indeed am one of those that is more interested in the nature and properties of pigments and their interactive relationships. I already know the scientific theory and all the competing color models, gamuts, etc.

OK, go back to the graphs.

Richard Saylor
02-09-2004, 09:36 PM
Only because those of us that acknowledged the importance of theory but shifted the discussion to practical mixing applications and alternatives were chided out of the discussion early on.
It wasn't me that did it. Anyhow, I've decided that my personal primaries are scarlet, ultramarine, and violet. Maybe I'll change my login to SUVguy. :evil:

Michael24
02-10-2004, 08:36 AM
Of course, that's the whole point in CMY theory - you want your yellow to reflect the green and red regions fully, with no blue reflectance. As I see it, the green reflectance in the yellow pigment wouldn't be an issue if there weren't also some green reflectance in the red pigment. If the red truly had no green in it (as with the idealized magenta), it would absorb all the green in the yellow without affecting the red. As it is, I'll bet there is just enough green in the red pigment to allow all the green in the yellow to cause noticeable trouble. For something like cadmium red, most of the green is probably at the yellow end. For most magenta pigments, there is probably a bit of green at the blue end.

Marc:
Cad. yellow light is tame by comparison to the organic yellows that have a perceptible green bias. Most of the reds that I tested spectrally had low reflectance up to 580-600 nm, and then sharply climbed in reflectance. At 600 nm I would not consider them to be green reflective. Only a few has a trace of yellow-green reflectance. Looking at magenta, most colors within that pigment type have far more green reflectance (480-510 nm) making them far more blue-green than most of the reds. (cadmium, quinacridone, even alizarin)

This is the trouble with mixing colorants. Nearly every pigment has a shortcoming in some area that interferes with obtaining clean mixes. So you just have to choose and learn what you need to get by and study the strengths and weakness of your choices.

We (as artists) are at the mercy of the large consumer marketplace. The style of car colors and plastic bottles has more to do with what we get to paint with than we may want to admit. Unfortunately, other than W&N making their own alizarin crimson genuine, no manufacturer makes pigment for the artists market.

Have a colorful day.


Michael Skalka
National Gallery of Art, Wash. DC

JamieWG
02-10-2004, 09:38 AM
FWIW, although I'm not currently using CMY, I am still using a very simple color wheel type of palette for plein air work, and like it for probably the same reason others *don't* like it - you get pretty high intensity colors by default.

Hi Marc. There is a huge difference between using CMY vs. other "primaries" when restricting to just a few colors. IIRC, you use ultramarine rather than phthalo. Don't get me wrong here; I understand why completely, but ultramarine already goes a huge step in the direction of toning down the palette and mixing less intense colors, especially less intense greens than a cyan-type phthalo (GS). That is why ultramarine is so valuable in a landscape palette.

Michael says all pigments have shortcomings in terms of creating clean mixes. Although that proved true in tests I did as well, most of the mixes that resulted from CMY, especially the greens, were far too intense for even a painter like me (and Marc) who like boosted chroma. (I felt orange suffered the most in terms of being dirty, so I started to carry a tube of cadmium orange, though mostly it was used in failed attempts to gain some control over the Evil Beast phthalo. :evil: )

With 20/20 hindsight, of course, I now know that CMY enables me to mix nearly all the colors I'd ever need, but most easily mixes the colors I don't need. I'm better off using whatever tubes will most easily mix the colors I need, and having others along "just in case", that will mix the colors I only sometimes need. Painting this way can require a huge number of tubes of paint in order to choose the three or four colors that will give you exactly the mixes you want for a given subject. (We're not talking about painting only landscapes, but also still life, portraits, figures, interior scenes, etc.) I think it will be a long time for me before those three tubes for a specific painting turn out to by CYM again. ;)

In oil, I have to struggle to avoid making my colors duller than I intend.
Marc, one of the main reasons I went to a CMY palette in oils for awhile was because I felt my paintings were getting muddy. The CYM palette cleared that up right away and led me to understand how to keep my mixes and paintings cleaner, even when adding the other colors back in. Overuse of earth pigments was my main culprit.

Michael, I found your comment about industry and pigments very interesting. I didn't realize that WN's alizarin was the only exception in the artist pigment world. On the phone with Triangle Coatings a few weeks ago (makers of Classic Artist Oils), they told me that they get all of their pigments from industrial suppliers. When they said that, I assumed it was unusual. 'Guess not!

Jamie

Michael24
02-10-2004, 12:12 PM
Jamie -

Your comments are really hinting at the true heart of the matter. Colorants are like the human voice or instruments in an orchestra. Instruments have a range of notes they can achieve. Amateurs get a smaller range, while experts can achieve a bigger one. However, a high and low limit does exist for each instrument or voice.

By their nature, pigments have spectral signatures that allow them to excel in certain areas and achieve poorly in others. So when people talk about one true set of primaries or palettes, in some ways they are talking about the maximium number of hues they can achieve with that limited set of materials. They might define the primaries as the smallest set that achieves the highest chroma mixes.

As you said, if you are constantly trying to lower the chroma of CMY colors when landscape painting, then you may not be using the right tools for the job. However, if you like high chroma landscapes, then you might be thinking, What's the problem with CMY? I don't want to beat down the chroma. I guess if the paints annoy you to a great degree, its time to switch to another set of colors that do the job more efficiently for you. As you know, you can always lower the chroma of a color but can never raise it higher than the maximum of highest chroma of any single pigment you are using.

I enjoy when the discussion turns to what little mixing tricks work for landscape painters. It is this kind of problem solving that provides insight for artists into the working properties of colors. Also the questions that ask why something does not work provide insight into the strengths or weaknesses of artists colors.

I do enjoy the CMY discissions because they span both additive and subtractive domains. Color printing that uses the CMYK method is really an additive system because the benday dots work in a similar fashion to what Seurat sought to achieve, a visual mixing of pure colors to achieve blends. CMYK also works in printing as an overlay of dots to make colors similar to glazing- the subtractive side.

Jamie - I looked at your color chart further. Very sweet. However, you complained about how some hues did not appear right. What is right when it comes to your test. Should violet be redder or bluer? Does a violet without bias exist? A good purple for you might be gastly for me. More to ponder.

Anyway, just a few thoughts.

Michael Skalka
Conservation, National Gallery of Art, Wash. DC

Marc Sabatella
02-10-2004, 05:51 PM
Hi Marc. There is a huge difference between using CMY vs. other "primaries" when restricting to just a few colors. IIRC, you use ultramarine rather than phthalo. Don't get me wrong here; I understand why completely, but ultramarine already goes a huge step in the direction of toning down the palette and mixing less intense colors, especially less intense greens than a cyan-type phthalo (GS).


True, but note I do have phthalo green on my palette as well, so I *can* get those really intense greens when I want them - which I'd say is about one in every three paintings. The changes I made in my palette weren't so much to help me get duller colors, but to increase the range of really intense ones I could easily get. Ultramarine got added as much because I love its color straight from the tube as for the fact that it makes somewhat dull greens. It also allows me to mix better violets. OK, that and the fact that I couldn't stand having phthalo as my only blue, and I'm not crazy about cobalt or cerulean either.

But you are right - having ultramarine does make mixing dull greens easier, and in that respect, it does solve one of the problems I too would have had with a CMY palette, and that isn't a problem with too much intensity. Since this was the first change I made, I had forgotten about that aspect of it.

Richard Saylor
02-13-2004, 01:43 AM
FWIW there are some small, somewhat crude spectral reflectance curves at Bruce MacEvoy's site http://www.handprint.com/HP/WCL/water.html
Go to paints > guide to watercolor pigments, and there is a color menu at the top of the page. Click on a color, and there are descriptions of various pigments. Some of the pigments have a tiny blue-green-red rectangle beside the description. Click on this and you get a reflectance curve for the pigment.

The curves for cad red PR108 and naphthol red PR170 clearly show why, as Jaimie observed, naphthol + blue makes better violets. PR108 has almost 0% reflectance on the blue-violet end, whereas PR170 reflects a significant percentage of blue-violet. Therefore, as Michael observed, a subtractive mixture of cad red and, say, ultramarine blue is going to reflect very little violet. However, naphthol and ultramarine should give some pretty nice violets, although their chroma is going to be somewhat compromised by the green component of naphthol.

JamieWG
02-13-2004, 07:51 AM
CMYguy, how cool!!! Thanks so much for showing those graphs. It is indeed interesting to see why things play out on the canvas the way they're supposed to. The napthol didn't mix great violets, but hey, the cads didn't mix any violets!

Oh---and here's a slightly better pic of the chart, I think.
http://www.wetcanvas.com/Community/images/13-Feb-2004/13766-Purple_Chart_2_550.jpg

Michael asked:
Jamie - I looked at your color chart further. Very sweet. However, you complained about how some hues did not appear right. What is right when it comes to your test. Should violet be redder or bluer? Does a violet without bias exist? A good purple for you might be gastly for me. More to ponder.

More indeed, Michael. You were supposed to ask me those questions before I made the chart. LOL ;) I love the bluer version of dioxazine violet and if I had to pick just one ideal violet for use in painting, that'd be the one. (Of course, that being said, my three tubes of cobalt violet, which is far redder, are all almost gone.) When I mixed these charts, some of the reds made me happy to get anything at all that resembled any hue in the vicinity of violet. I wasn't picky enough about which violet, how red or how blue, in making the chart and mixing the samples. I suppose I should have selected a pure violet, such as the PV23 BS or PV23 RS, and tried to mix matches for that.

<Edit: Well now I see this version of the chart has wavy lines all over it. If anybody wants a full size version emailed to them, just send along your email address. I don't know why the size reduction resulted in the wavy lines.>

Jamie

bigflea
02-15-2004, 12:16 PM
This is an interesting thread for me because of what is being said about the relevance of color theory to our real life painting efforts. It may be accurate to say that some painter's want more technical knowledge than others, re the "why" and specific performance data of available pigments. I believe it is helpful to have some interest in the technical data, and that as we grow as painters it becomes increasingly helpful to understand the way pigments are formulated, or at least the way they will differ in practical application. For me it is simply an interest in getting a more reliable response from a pigment, rather than having to settle for a color quality that does not faithfully express what I believe is possible based on what I am seeing in front of me.

I believe using the limited palette approach starting with three primaries and white might be helpful to beginning painters who as adults are overwhelmed by the idea of color and believe that the object color is a constant. I am trying this out in a class, with the goal being simply to help beginners get a grasp on how pigments can be used in mixtures to make the colors they believe they see. It is another step altogether when I try to show them how the color we believe we see is light rather than a fixed object color. Nevertheless, the principle of mixing a few colors together to get "seen" color is the same, whether or not the painter believes they are seeing a fixed local object color, or color in a light and atmospheric key which is always changing.

How much does "extreme" color theorizing help the painter get more ideal results? To Marc it seems that it is necessary to imagine the ideal pigments and the ideal system for recreating a coloring, and that this imaginative exercise provides a catalyst for getting results in real painting efforts. For Bill it seems that his understanding of the behavior of pigments provides the key to getting color mixes in paintings that are done in a studio format. Correct me if I am wrong. Jamie has gone to an extreme effort in scrutinizing the behavior of pigment mixtures in order to get more chromatic results in paintings. Michael seems to be saying that all of the efforts need to be focused on the pragmatic problems of pigment mixtures, and finding more direct ways of getting mixtures that perform in a way that recreates the way form appears in light.

In my own efforts I think I have gone in all of these directions at different times and at the same time. At one time I imagined the ideal would be to have some machine mix what I was seeing and have the paint come out of the tubes in the exact quantity needed for the square quarter inch area of the board I was painting. Later I realized that machine is my brain. I have found that all of the study of color in the landscape, using the color modeling talked about in other threads, has been very helpful in mixing coloring for interior design or any sort of coloring, without reference to an image. In other words, practical experience in mixing leads to color knowledge whether or not we know anything about the color formulations, or the theories that determine the behavior of the color mixtures. So as Bill pints out, we do not need to be hamstrung by color theory, even if it is helpful or easy for one individual or difficult for another to grasp.

I doubt if anyone has been more theoretical than Seurat, in the adherance to a theoretical or precision theory on obtaining additive coloring effects with subtractive tools. I recently saw a beautiful painting done by Signac, who was employing the same approach, depicting a sunset over a lake surrounded by trees. It was a stunning painting, and it was done with extreme care to use only those dots of color that would recreate a specific light key. What a maniac. What a beautiful mind to have such discipline and vision. How fortunate for us that someone took the time to try this approach. Pissarro experimented with this approach, using dots of pure pigment to suggest light keys, and complained strongly about the tedium and loss of feeling for the subject due to the demand on time and focus to complete a work. I think that suggests that there is a practical consideration to following any purely theoretical model. Not that practical concerns ought to outweigh the ideal, but that they are present because, perhaps, the ideal is not being reached when we follow a purely theoretical ideal. In other words, the experiements and the experiences of Pissarro and others can indicate that there is more to attaining the ideal than in following a pure theory of it.

When the forms of the pointillist paintings are scrutinized, for example in the images Pissarro painted, one may get a contradictory feeling both of perfection and of unreality or stylization. At least that is how I might describe my own response to them in general. There is a sense of mannerism, by which I mean a kind of repeating shorthand that is non descriptive of the truth of the visual imagery.

In regard to Michael's comment that is a practical concern. Whatever approach one may take in painting, how well does it avoid the appearance of mannerisms, either personal ones, or ones learned by following a particular kind of appoach. For example Monet is often imitated by using particular kinds of markings with the paint, and work having those characteristics is called impressionist in style, even though it may not have any reference to light and atmosphere asthe subject.

In looking at painter's work, I cannot see any substitute for hands on pigment mixing trial and error effort, especially when that effort is made in attempting to show the way light alters the local color of objects. Most painter's can get very good at color mixing by design, and at color arrangement by design, that is, without reference to a light key. The wide range of color design found in different cultures throughout the world ought to show us that pigment relationships, as local object colors, can be easily learned by following a limited palette system, provided one has a sufficient interest in color to do so. However it is a horse of a different color altogether when we attempt to begin recreating additive color effects with subtractive pigments. In that effort, a great deal more can be learned by drawing upon the resources of an unlimited palette, and an approach that recognizes the presence of the massing of light and shade as the fundamental way in which the eye assesses the coloring present in nature. In other words, the eye sees coloring in a harmonic range as determined by the fundamental color difference between light and shade.

So, in regard to Jamie's effort to become more chromatically intense, it is perhaps more to the point to become more aware of the harmonic range of color that is present in the situation, called a key, than it is to increase the chromatic intensity of any or all colors. It is the lack of harmonic range and variety of coloring difference which makes a painting appear dull in any key, rather than the need for more intensity of chroma. Correct me if I am misunderstanding your comments on this Jamie. In the pointillist efforts, the chromatic intensities of dots of colors mixed in the eye by neutralizing each other, producing a "greying" effect that is characteristic of daylight keys. However the opposite effect occurs when the dots are instead large chunks of pure color. When painting in large masses of light and shade, as many do, the coloring of each mass has to be harmonically related to one another in order to attain the "greying" effect of the aeriel perspective. It is important to realize that the term "greying" is misleading, since a daylight color key is not ever a neutral "grey" but is instead a harmonic arrangement that approaches neutral in a specific way.

So in practical terms, a painting has to move from simple primaries to tertiary colors, but in a specific harmonic context, in which the family of color is not obliterated, or turned into a true colorless neutral. How pigment mixtures behave then becomes a pragmatic study which if understood does not require a reliance on pigment analysis but on a familiarity with what each one pigment does in relation to another.
bigflea

Marc Sabatella
02-15-2004, 05:22 PM
I believe using the limited palette approach starting with three primaries and white might be helpful to beginning painters who as adults are overwhelmed by the idea of color and believe that the object color is a constant. I am trying this out in a class, with the goal being simply to help beginners get a grasp on how pigments can be used in mixtures to make the colors they believe they see.


This makes a lot of sense, although of course, you don't really have to resolve the issue of what "the" primaries are in order to get moving here. For the beginner struggling with issues like, "what color do you use to paint blonde hair" (a question I recall seeing recently), it can be an eye-opening experience to realize you don't need to have a "blonde" pigment on your palette, but that you really can mix just about anything you need once you understand the basic principles.


How much does "extreme" color theorizing help the painter get more ideal results? To Marc it seems that it is necessary to imagine the ideal pigments and the ideal system for recreating a coloring, and that this imaginative exercise provides a catalyst for getting results in real painting efforts.


I wouldn't say this is "necessary", but you have summarized pretty well what my interest has been on this thread - understanding what is reasonable to expect in an ideal case, how the real world differs from the ideal one, and how to use your understanding of how how things work to address the discrepancy.

BTW, with a few more paintings using naphthol red under my belt, I am now satisfied that it fits the bill for helping me more easily mix the kind of oranges I want - I am no longer concerned that is still too red. In fact, I'm coming to see this as a blessing, as it is red enough to make perfectly adequate violets and tother dark colors for some situations, meaning I can often get away with using naphthol *instead* of quinacridone in some paintings. I still have both on the palette and am not afraid to use both as necessary, but for many paintings, it suffices to ask myself, am I more concerned with clean oranges or violets, and choose my red accordingly.

Michael24
02-17-2004, 02:26 PM
So, in regard to Jamie's effort to become more chromatically intense, it is perhaps more to the point to become more aware of the harmonic range of color that is present in the situation, called a key, than it is to increase the chromatic intensity of any or all colors. It is the lack of harmonic range and variety of coloring difference which makes a painting appear dull in any key, rather than the need for more intensity of chroma. Correct me if I am misunderstanding your comments on this Jamie. In the pointillist efforts, the chromatic intensities of dots of colors mixed in the eye by neutralizing each other, producing a "greying" effect that is characteristic of daylight keys. However the opposite effect occurs when the dots are instead large chunks of pure color. When painting in large masses of light and shade, as many do, the coloring of each mass has to be harmonically related to one another in order to attain the "greying" effect of the aeriel perspective. It is important to realize that the term "greying" is misleading, since a daylight color key is not ever a neutral "grey" but is instead a harmonic arrangement that approaches neutral in a specific way.

So in practical terms, a painting has to move from simple primaries to tertiary colors, but in a specific harmonic context, in which the family of color is not obliterated, or turned into a true colorless neutral. How pigment mixtures behave then becomes a pragmatic study which if understood does not require a reliance on pigment analysis but on a familiarity with what each one pigment does in relation to another.
bigflea

Very thoughtful comments.

With regards to Seurat. He took to heart the color theories of Michel Eugene Chevreul probably more than Chevreul himself. Chevreul was attempting to understand and avoid the greying of color when making tapestries. He noticed that colors tended to loose chroma in certain blends, leaving the design rather dull and flat. Seurat did employ the greying effect, although sometimes he cheated and used nearly neutral grey backgrounds and placed dots of subdued color in these areas. In other instances, Seurat wished to achieve optical mixing of a secondary color by placing the two primaries side by side. His was a sort of Maxwellian additive color spinning wheel approach to blending paints. It is much the same as we interpret color in 4 color CMYK printing today with dots that are much smaller than could ever be painted.

Harmonic arrangement that approaches neutral in a specific way - I love that phrase. What a beautiful choice of words!!

Yes, in the end it come down to understanding the bias, the nuances of colors and how they interact. Color theory is a great starting point and needed to understand how the basics function, but it boils down to experience and manipulation of the hand, brain, eye, tools and local lighting that do all the work.

Michael Skalka
Conservation, National Gallery of Art, Wash. DC

Bud Ralls
04-01-2004, 02:32 PM
My 2 cents, As always a personal choice i
think in school, most of us learned Red, yellow, blue. many times i substituted black for blue, i can get almost anything i need from ( cyan, Magenta, Yellow, ivory black..... Bud Ralls

Richard Saylor
04-01-2004, 03:06 PM
My 2 cents, As always a personal choice i
think in school, most of us learned Red, yellow, blue. many times i substituted black for blue, i can get almost anything i need from ( cyan, Magenta, Yellow, ivory black..... Bud Ralls
I would really like to use black. What's holding me back is the fact that black pigments aren't really neutral/hueless. (As you say, black can often be substituted for blue.) I get color harmony effortlessly using cyan, magenta, and yellow, and I'm hesitant to add another color which might interfere with that harmony. However, maybe ivory black is worth a try. Thanks for the idea.

Einion
04-03-2004, 02:32 PM
Tube Cadmium orange, for example, shifts toward red when mixed with Flake White.
Bill and I have talked about this before and as I mentioned to someone in a PM just a week or two ago, I don't point this out every time he says it but it's been a while so it's time again :) This is the case for the Grumbacher paint Bill uses, it's not always true for this colour so please don't be surprised if you try it and you don't see it! It's not the case for the one tube of this colour I have in acrylics for example.

If Bruce MacEvoy's suggestion of why it happens is correct it couldn't be said with certainty to be generally true either, it might be specific to the manufacturer of the pigment Grumbacher use or possibly be a milling issue. Additionally, some Cadmium Oranges are actually mixes of two types of cadmium sulphoselenide - which might not be properly documented on the tube - this could also account for it as the red particles might be of a different particle size than the orange ones.

certain tube colors exhibit an "overtone" which actually shifts them toward adjacent hues when mixed with white
Sorry to be a pedant Bill but given that thread in the oil forum recently... :) In mixes with white it is specifically the tint, in mixes with other colours or when discussing a wash or glaze it is the undercolour or undertone; it's important to make a distinction in the terminology (and we have a special term for mixes with white anyway so we should use it!) since tint and undercolour are so markedly different in some colours.

Unusual things happen when you mix certain colors with white. Instead of simply "graying" the color, as others believe (and I used to believe)...
Yes indeedy, I have a hard time with anyone who says that tints of phthalo blues are greyed or 'chalky' :D

In addressing the issue of a single pigment orange verses a mixed orange, I have an example using cad orange and a mixed orange made of cad yellow light and cad red medium.
Thanks for the chart, it's good to see. The results are pretty much what one should expect from these two colours. It should come as no surprise that a mix of Cadmium Red Light and Cadmium Yellow Medium make a superior mixed orange. In fact it's so good, in practical painting terms if you already have these colours there is very little reason to also have a Cadmium Orange. The very slight difference in chroma is irrelevant for most of us since you would almost never need this hue at maximum saturation anyway.

With 20/20 hindsight, of course, I now know that CMY enables me to mix nearly all the colors I'd ever need, but most easily mixes the colors I don't need.
Hehehehe, what a great way of putting it!

...it is perhaps more to the point to become more aware of the harmonic range of color that is present in the situation, called a key, than it is to increase the chromatic intensity of any or all colors. It is the lack of harmonic range and variety of coloring difference which makes a painting appear dull in any key,
Colour key isn't an important concern in painting for most of us, a vista or scene that does have a definite key in terms of hue is relatively rare. The general idea of 'colour key' and 'mother colour' are usually just a device to help painters achieve work with harmony; they don't necessarily accurately reflect how things are in the real world, just like notions such as 'cool colours recede' or 'warm light, cool shadows'.

In the pointillist efforts, the chromatic intensities of dots of colors mixed in the eye by neutralizing each other, producing a "greying" effect that is characteristic of daylight keys.
I see Michael has touched on this already but one of the interesting things about viewing Pointillist works in the flesh is that they didn't necessarily use dots of the colours one might expect from what is said about the technique. There are two very good ones in the same room in the National Gallery in London and the better of the two IMO is a small, little-known landscape and up close you can see that Seurat used mixed lower-chroma colours in many cases to help the technique work more successfully (i.e. more realistically).

Einion

LarrySeiler
04-05-2004, 10:22 AM
I believe using the limited palette approach starting with three primaries and white might be helpful to beginning painters who as adults are overwhelmed by the idea of color and believe that the object color is a constant.

This is interesting, because I just did my first effort outdoors yesterday with a three pigment plus white limited palette...as an experiment, and I've been painting since 1977...so, guess I'm going backwards. I was amazed what I was able to accomplish, which might be due to my NOT being a beginning painter. I posted that effort in this forum. I certainly do not feel overwhelmed by color either.

Its a bit like learning to play blues after years of playing appeggios in heavy metal. Lightning blitzkrieg riffs on the fretboard has all sorts of notes, where rhythm and timing is everything...but then to go to that thinking of where less is more can be a real challenge. The result, is that you try and pour and tweak as much emotion out of the fewer notes played. What is more is the observed power such few notes sustained and wailed can have over the listeners.

Emphasis changes, but its not a dumbing down, a playing down, nor a step backward. Its more a step of mastery. Okay...you can do this with eight colors or ten...but, let's see what you can do with three, sorta thing.

In that situation you guide the eye to think they are seeing color that might not be actually there, and you craft adjacent color to help pull that off. Such actually might be over the heads of beginners IMHO....


In that effort, a great deal more can be learned by drawing upon the resources of an unlimited palette, and an approach that recognizes the presence of the massing of light and shade as the fundamental way in which the eye assesses the coloring present in nature. In other words, the eye sees coloring in a harmonic range as determined by the fundamental color difference between light and shade.

I found this true yesterday...that the limited palette put me in a position to focus on rudimentary fundamentals of working to establish sound values. The color nearly produced itself. That is, with the split primary palette that is second nature to me...I focus more on the color because I have the value inherency of color understood at a more intuitive level. That, and I respond more to the character of warm and cool. The limited palette was like throwing a bit of a wrench in the routine...yet, I was amazed at how empowered I still felt to imitate color I was seeing. It nearly seemed too easy to be honest.


So, in regard to Jamie's effort to become more chromatically intense, it is perhaps more to the point to become more aware of the harmonic range of color that is present in the situation, called a key, than it is to increase the chromatic intensity of any or all colors. It is the lack of harmonic range and variety of coloring difference which makes a painting appear dull in any key, rather than the need for more intensity of chroma.

First, I wish to commend you Bigflea...for one, it is obvious you have a great deal of passion for your views, and two...the integrity to live it out. Inspite of a number of us that don't see eye to eye. I'm glad you are around simply because your views having a different take is impetus for many to think, even if the end result is we do not agree.

To me...the proof is in the pudding as they say. I have lived my life actively outdoors since I was wee young. I'll be out trout fishing soon. Summer fishing. Then fall bird hunting with my dogs, exchanging hands to bowhunt deer arduously. My son and I harvested seven deer between us last year and this comprises the content of our freezer. Rarely do I buy store meats, and I eat low carb so that let's you know I require more protein. That means I am outdoors a lot.

My past 25 years or so as an artist, 20 of those has been as a reputable wildlife artist with titles and awards to back that up. I've been charged by whitetails, had unique encounters with bears, you name it. As an artist, I feel I am super charged aesthetically. I close my eyes I hear water, hear leaves, sense wind...feel the outdoors.

Then the past seven years I vacated the studio and went outdoors with my easel. I was admittedly quite surprised to see even with all my outdoor experience that my painting relied on presumptions, but direct observation filled in the holes and made correct some of the wrong presumptions.

I have an eye for the outdoors, a soul for its mood. While most are comfy in their homes in the frigid temps, I am usually knee deep in the white stuff. So, I attempt to paint what is true to my experiences. By the same token, I see in the works of some the pulse of what I feel and experience outdoors.

What has driven me to experiment with the 3 pigment palette to be honest is the excellence of the works of Marc Hanson (a member here at WC) and Scott Christensen...who in my opinion put out work with their limited palette that has more pulse on the realities of the outdoors than the proper "key" paintings you have pointed our attention to.

You have said in the past this is because my eye or others is not as highly developed...and to that I say hogwash, there is hardly a soul that spends as much time outdoors as I...and I can tell whether I am indoors or outdoors and I can tell when a painting emits that sense.

I see nice design in the "key" paintings you share and talk about. They are pretty pictures. Certainly fine art. Certainly worthy of marketing and enjoyment. They touch upon moods and emotion, but again...I just don't get the gist of your emphasis where the suggestion of its being superior especially in touching reality is concerned, or where thereupon you attempt to minimize other approaches as less successful. I just don't get it, and I don't think I'm alone. I don't want to be rude...and that is not my intent. Its not about besting another debater. Where it comes to seeing and imitating...we artists simply wear our emotions on our sleeves.

Chroma...purity of color where imitating the actual sunlight is a necessary concern for artists, for pigments have their limitation. Value as determined by light and dark where white pigment is a main additive reduces not only chroma, but as it lightens it reduces its ability to brighten, and brightening is exactly what the sun does. One can attempt to compensate by lowering values in areas of shadow...but only so far. The other end demands stronger purer color, or chroma. This is why I am now convinced I will require, (if I continue to work with the 3 pigment plus white palette) much better quality pigments and perhaps fork out the bucks for the first time in my life.

Larry

Bud Ralls
04-05-2004, 12:16 PM
Larry.
I must say a very interesting reply, i looked at your most eye catching paintings as you said your always confortable living in your own skin, Amen i must say you have a great backgroung somewhat like mine, Navy, i was in the Marines, Korea era, about your paintings in the section - Plein air Gallery i must ask this, some of your work looks like a great painter that as a preference uses a Palette knife, for really getting intense colors i only wish i could his name is- Howard Behren's, i guess some say it is spontaneous.
Also Larry, do you ever use umber on your canvas or just white, i find umber gives a neutral gray. once again Larry you do great work.


"Success to me is liking my own work, this is very seldom.
Bud Ralls

LarrySeiler
04-05-2004, 10:16 PM
well thanks Bud...appreciate the kind regards...

I went thru some shifting in my thinking reading the writings of a painter we don't often talk about here...but who I think shouldn't be dismissed and ought to be regarded highly, and that is Paul Strisik. He had some wonderful books out on painting landscapes on location...long before it was trendy to be called plein air. He passed away a number of years ago, and I still regard his books special.

He mentioned a good reason for using a warm wooden palette...and going with those reason I went a step farther and often add umber to my grayed gesso.

Here's the bit on the palette...

When painting outdoors....you are limited of course to minerals and pigments attempting to imitate light. There is a chroma or intensity you feel in the color of nature in direct light...and a hint of numerous colors guarded from the light from reflected/bounced atmospheric light.

Mixing on a white palette...color will appear warmer more inherently and with greater immediacy. White is cool in temperature.

Yet...seeing that color having gone from the white palette to the painting's surface, it seems to have lost a bit of its punch. Not nearly so intense or warm.

If you have a white blank canvas...and a white palette...intially you might get a sense of the color maintaining its look from palette to canvas...but as it gets filled in, it will change.

Mixing paint on a warm palette...such as wooden one, it will require purer brighter color mixed to appear warmer by comparison to the palette, thus it goes onto the canvas warmer. This then works in your favor attempting to imitate nature.

Well...as I said, I took that thought a bit further of Strisik's and thought, okay what about the canvas?

Sorolla, it was said painted subjects in light outdoors just above mid value gray, and that which was in shadow below mid value gray.

IF indeed nature sits about a mid value to begin with, a white canvas is unsuitable and you get this hasty rash compulsion to quickly fill it in, but it is hard to judge the color's warmth rightly against it. You will find often that as the painting fills in you need to go back and touch up strokes to tweak and correct them.

By taking what Sorolla taught to heart, start the canvas with about a midvalue gray! Yet...then, to guarantee that your paint strokes and pigment choices are more economical with time restraints of elusive light so as more likely to nail it right away...add some warmth to that gray gessoed canvas.

So, I do add often a bit of umber to my gesso for just that reason....

I often do use a painting knife, well...really every painting, but the painting, the mood...the pressing need determines when, how much and where I will use it.

Its been said to use your gift, don't let the gift use you...and I think some in discovering a neat device or technique will for awhile go nuts with the thing. Often they might be looked at for the technique. I did go thru a period of time where perhaps 70-80% of the painting was done with the knife.

That did give me a bit of expertise and a sense of ease using it...but like many things balance eventually gives a thing its proper place.

I'll use a rag...a finger, the knife...brushes...you name it.

I am also somewhat a slop painter in that I have never let not being able to afford $30 sable brushes keep me from painting, nor from considering in competition it was possible to best other artists that were better equipped.

You'll appreciate this Bud...but when I was in boot camp at Great Lakes Illinois, the base battalion commander stopped my basic training. I had been in a special drill team unit which would have gone four weeks longer than other recruits, but this commander found out I had a past as an artist, and some education.

I was given an entire barracks to myself, and my orders were to take a photo reference he had of a pbr (river patrol boat) under fire in Vietnam, and returning fire...and paint a scene of it on a large oval board. In good ole Navy/bootcamp style...I was given old outdated paints in dented old rusty cans. Prying the lids off...the paint was enamel based...like old dripping honey. The brushes I was given were the Ben Franklin store craft variety, and were used and beaten.

Imagine me standing before the commander...a recruit, and telling him that much to my regret I was not going to be able to paint this work for him due to the poor condition of the paints and brushes.

He told me as only an officer to a recruit can (won't repeat words) that I would indeed do the painting, and that it would be OUTSTANDING!!!, to which of course I replied, "YES SIR!!!!"

Took me nearly seven weeks of boot camp to complete that painting. Some basic I had huh? hahaha....

But...by golly....when I was done with that thing, it was a marvel to behold. All I have is an old photograph today, but I wish I could see it in person again.

The painting was placed between large bronze stantions, and two Marine guards on both sides, at the entree of the main base headquarters.

...well...okay, I admit that like many others sometime my artist's emotions are worn on my sleeves as well, (ie, I'm human)...but it really irks me and rubs me the wrong way when someone insists on what is necessary, and what cannot be done.

I learned something about human ingenuity, human willpower...and I made a statement, and I plan yet to carry that out this spring, but in light of the claim that one must have the best hog bristles and so forth aimed in a thread last year to belittle my experience as a painter...I plan on taking various sapling brushes, hammering the ends into mashed fibers and doing a painting with them. I will post it in the appropriate forum when I've done it.

okay...nuff....

Larry

WFMartin
04-05-2004, 10:45 PM
Larry...................You're cool!!

Bill ;)

Eugene Veszely
04-06-2004, 12:45 PM
Here is an interesting page...

http://www.total.net/~daxx/paint_correctcolour.shtml


There is lots of other interesting stuff there about colour too.