PDA

View Full Version : color mixing experiment


Patrick1
08-26-2001, 03:57 PM
I don't know if I'd call this an experiment, but a big color mixing exercise to see the results when mixing between different primaries. These are all acrylics. I have 4 colors for red/magenta, 3 yellows, and 3 blues. My goal was to see if I could find 3 primaries that seem to be the best overall compromise. There are also comparison swatches below the mixing steps for comparison to see how well my mixtures compared to a color straight from the tube (some were single pigments, others were mixtures).

I cleaned my brushes very thoroughly when changing to a new color(s). In most color swatches, the paint was applied heavily at the top, then scrubbed down lower to see the undertone. However,
when I did the comparison samples, I had more time and was able to get the undertone using a thin wash using lots of water. I suppose that this will give clearer undertones than scrubbing, but it would've taken forever to do this with all the swatches.

The photo was taken outdoors under direct noon sunlight, but it still looks dark. Any grey blurs are fine black labeling that wasn't resolved in the photo.

http://www.wetcanvas.com/Community/images/26-Aug-2001/colormixsteps-compressed.jpg


In the top middle, are the primaries I chose to use:

-Liquitex high-visc. cad red med. hue (mixture)
-Liquitex h.v. Acra red (quinacridone red PV 209)
-Golden heavy body quinacridone red (PV 19)
-Tri Art high viscosity quinacridone magenta (PR 122)

-Wallack's Student grade 'lemon yellow' (don't know the pigment, but it's even slightly more greenish than PY 3...hansa yellow light)
-Liquitex h.v. cad yellow light hue (bismuth vanadate...PY 184)
-Wallack's Student grade 'bright yellow' (doesn't say the pigment, but looks exactly like PY 73...hansa yellow medium)

-Tri Art h.v phthalo blue GS...PB 15:3
-Golden h.b phthalo blue RS...PB 15:1
-Liquitex h.v. ultramarine blue PB 29

And the comparison colors are:

-Grumbacher Finest hansa orange
-Liquitex Basics cad. orange hue
-Golden h.b. phthalo green YS...PG 36
-Golden h.b. phathalo turquiose...PB 15:4 + PG 7
-Wallack's Artist grade dioxadine violet...PV 23
-Wallack's Student grade 'purple'...doesn't list the pigments, but looks exactly like a mixture of anthraquinone blue and quinacridone magenta by Golden, under the name of 'permanent violet dark'...more reddish than diox. purple

Note that the only red which didn't have a pronounced bluish undertone is cad red hue. Also, the bismuth vanadate yellow I used is approx middle yellow (looks about halfway between hansa yellow med. and light) yet my information is that bismuth yellow is usually one of the most lemony yellows available. Don't know why mine is not lemony.


Overall, the results were pretty much as I expected, but there are a some very interesting things I found:

1)oranges were the 'easiest' to make with the primaries I used, while greens and purples were
considerably more difficult. But despite what I've read, oranges werent easy to make..many of the combos don't even come close to making a clean orange

2)the only oranges that were virtually identical to my comparison samples were a mixture between
PR 209 and hansa yellow medium. But note that my comparison oranges (hansa orange and cad orange hue)are not as saturated as real cad. orange, which I don't have a tube of.

3)I was surprised that the cad. red med. hue didn't make great oranges...worse than PR 209...this must be becasue it's opaque...can't think of any other reason.

4)all of the mixtures with bismuth vanadate yellow were surprisingly dull...in fact, lemon yellow made cleaner oranges than bismuth yellow! Again, the only explanation I can think of is that it's
opaque.

5)quinacridone magenta PR 122 generally makes poor oranges and reds

6)none of the greens and turquioses were as clean as thalo green YS and phthalo turquoise. Even the best mix, between lemon yellow and phthalo blue GS, were still noticably less saturated.

7)all purples made with cad red med. hue were very blackish or grayish

8)mixing phthalo blue GS plus quin. magenta, I couldn't even come close to an ultramarine blue...despite what I've read elsewhere, nor could phathlo blue GS mix clean purples

9)ultramarine blue plus quin. magenta came close to diox. purple, but still a noticable difference
More reddish purples were generally eaiser to mix cleanly.

10)biggest surprise: not much difference in oranges mixed with lemon yellow and hansa yellow medium, (which has a very orangish masstone). But there was a big differnce in greens mixed with these yellows.

11)it was impossible to mix a clean lemon yellow
with hansa yellow medium, but I could get a very decent imitation of hansa yellow medium with lemon yellow plus red. This seems to suggest that lemon yellow is closer to the theoretical primary yellow.


So if I had to choose only 3 colors as primaries, I probably wouldn't choose the oft-recommended phthalo blue GS and quinacridone magenta. That just doesn't give clean oranges. And I can't understand how a cyan would give clean reddish blues to purples. I'd go for something more like:

-a somewhat lemony yellow (something like hansa yellow light)
-for red/magenta, probably quin. red PV 19
-phthalo blue RS

Einion
09-04-2001, 07:21 PM
Hi Patrick, thanks for posting the results of your mixing experiment. Sorry for not commenting earlier but I was occupied with an ongoing discussion elsewhere. Okay, some interesting results as you say. I was particularly struck by what you said about PY184 which, as you pointed out, should be a distinct lemony shade as all my references point out - the hue-angle is precisely the same as some Cadmium Lemons. Assuming the pigment is correctly stated it's possible I suppose that their version is slightly different from others (as with W&N's Chromium Oxide Green which is distinctly bluish!) Any extenders should not account for a hue shift. Very curious.

What did you mean by "while greens and purples were considerably more difficult"? Difficult to achieve clean mixes or tricky to mix the incremental steps?

Oranges
Not surprised you couldn't mix particularly clean oranges with the reds you had - colour bias theory would predict pretty much the results you got. PR209 and what you think is PY73 would mix the best as you indicate.

Liquitex Cadmium Red Medium Hue is a mix of PR170, Naphthol Carbamide, and PY97, Arylide or Hansa Yellow. While the choice of yellow is superb (the most saturated yellow pigment available) the red has quite a significant violet reflectance (I use it as my violet-red) so the resulting mix is 'dirty', resulting in lower-chroma mixes than with the genuine colour. Cadmium reds and yellows mix particularly good oranges - Cad Red Light and Cad Yellow Medium mix to a virtually indistinguishable result to Cadmium Orange in masstone and undercolour - which is they are valued so highly.

Greens
PY3 and PB15:3 should make very clean greens and the intermediary steps look good and bright in your picture. Of course it's impossible to match PG36 perfectly because of the specific hue, value and chroma of this colour.

If your lemon is indeed PY3 it should have been possible to mix something nearly identical to the Golden Turquoise, it might be the lower pigment loads of the brands of each colour that is at fault.

Violets
I was surprised at what you said about violets. Your mixes of PR122 and PB29 look pretty good and theory predicts they would be particularly clean, perhaps the best mixed violet possible from a 'red' and a blue.

Not at all surprising that the violets mixed with the Cad Red Hue were grey and dull - the yellow component would account for that. The genuine colour would yield even duller results, because it reflects a lot of yellow and orange light and very little violet.

No surprise at all that PB15:3 would not mix any clean violets, especially in masstone.

Quinacridone Magenta, while a fine colour, falls far short of the perfect reflectance spectrum for a magenta in the CMY group; also there is the issue of masstone v. undercolour of course. Phthalo Blue GS when used thinly is a very good cyan so in watercolours works very well as part of a triple primary group; its very dark masstone, on the other hand, makes it impossible to achieve decent results in oils or acrylics if you use your paint at all thickly.

I have not heard of Tri Art or Wallack previously. Did you come to any personal conclusions about the qualities of the different paint brands you tried? I imagine that the contrast between the Golden HB and the cheaper grades would be particularly marked.

Einion

Patrick1
09-05-2001, 05:15 AM
I was hesitant to change the photo in any way because I wanted the colors to look as they really do, but since it is definitely too dark, I boosted the brightness and contrast by 20%. Ironically, this picture of it looks closer to the real thing than the original unmodified picture.

http://www.wetcanvas.com/Community/images/05-Sep-2001/colormixsteps-adjustedandcompressed.jpg

Einion,

I guess the only explanation for the Liquitex bismuth vanadate yellow (cad. yellow light hue) is that Liquitex's variation (the pigment?) is signifigantly different than the typical PY 184, just like there is sometimes a hue difference beween colors using the same pigment between different manufacturers. But such a non-lemony yellow (as you can clearly see in the picture) is very strange. On the label, it is shown as a "middle yellow"...no leaning to either side, and
considering it looks roughly halfway between
hansa yellow med and light, I agree. But I'm not complaining about it, in fact I like that I can use it as a good, opaque middle yellow.

By "greens and purples were considerably more difficult" I meant more difficult to achieve clean mixes.

I did notice that cad red med. hue is kindof dirty...especially its undertone which looks very drab (maybe 'cause it's a mixture). Did you say you use this as a violet red? As you can see in the picture, it's the ony red without a distinctly purplish masstone or undertone.

I was considering using real cadmiums
but for health reasons I don't use real cadmiums, and I didn't want to buy some just for this because they're far from cheap. But it would be nice to see how close a mixed cad. orange comes to actual cad orange. Do you know of any organic pigments which come close to real cadmiums' ability to mix clean oranges? You said PY73 is great, but any great reds?

With greens, my best green mix (phthalo blue GS + lemon yellow) made a good turquoise (though still discernable), but the middle greens (like phthalo green YS) was less clean...quite appearent in the picture. It could at least partly be explained becasue the lemon yellow is a cheap, student grade paint. I have a tube of Grumbacher Finest hansa yellow light, but in my preliminary tests, the lemon yellow (which is a tiny bit greener looking) gave cleaner greens
so I used it. I will eventually get a tube of Grumbacher 'thalo yellow-green" to put on as a comparison sample onto the green steps. But since it itself is a mixture, I don't know if it'll be of much value as a comparison.

My best mixed violet is good, but still clearly less clean than diox purple. Could it be that I was kindof scrubbing the mixes to get the undertone rather than using a thin wash (as was done with most comparison swatches)? I hope this didn't invalidate my entire chart!

Could you explain the problems with PR122 as a
magenta? I think I know what you mean. Is it because, for example, PB 15:3, it's undertone is clean and very greenish, and its thickly-applied masstone looks very purplish...at least when it's not dry yet...giving PB 15:3 a dual nature. On the other hand PR 122 does not have a reddish/orangy
side to it. It has a clean undertone, but its masstone, which is more reddish, is dark and drab (should be lighter than the undertone to make a red?)...almost like a burgundy. Is this what you mean?

Do you know of any acrylic brand that sells a true cyan? Handprint says that Holbein does, in Peacock Blue, but I cant find that brand here in Canada. Is PB 15:4 a particularily greenish phthalo blue...even more grenish than PB 15:3?
I want a true cyan just for the hell of it.

I'm glad you asked about the particular brands I used. Wallack's is the house brand of Wallack's art stores, a very small chain of art stores in Ontario, Canada. Their student grade acrylics are very inexpensive, but very extended.
They are also very sticky. I thought I would save money with them, but sice you have to use more of it, you probably don't save much. Their artist grade is better, but still short of the quality of Golden, Grumbacher, Winsor & Newton.

Tri-Art is a manufacturer in Kingston Ontario.
They make their own Tri Art acrylics. The Tri Art artist grade seems pretty good...pretty similar feel as Golden and seemingly good pigment load (based on the two colors I have). Tri Art also makes Wallack's acrylics.

Golden h.b. seems to have the best tinting strength and purest colors, but like Tri Art, they feel quite plasticy in the organic colors. But my Golden h.b. burnt sienna seems less stringy...maybe because its an earth pigment.

Liquitex high viscosity does seem to have less pigment load than say Golden, but I like the feel of Liquitex better...shorter, more buttery...Again, based on the few colors I have.
Liquitex Basics seem to have low tinting strength, but they do have a nice consistency to them...reasoably thick, and also quite buttery.

I still don't have a single Winsor & Newton acrylic color. I heard they're very good. Will try them eventually.

Big surprise: I bought Grumbacher Finest hansa yellow light and hansa orange for this color chart (didn't end up using the yellow), and I can't believe how oil-like Finest acrylics feel...firm and un-stringy...holds its shape and brush marks much like oils. Would seem to be great for smooth blending by scrubbing...well smooth as far as acrylics go. I'd recommend them to anyone to try out.

Einion
09-07-2001, 11:45 PM
Hi Patrick. I have had experience with Liquitex colours not matching the expected hue position for a given colour myself. Their Raw Umber, in addition to being a bit slack and transparent, is much yellower than I prefer for this colour but natural earth pigments vary so much it is really only its transparency that one can hold against them. I will be in London in a few weeks so I'm going to take the chance to compare Liquitex's PY184 with other yellows. Hopefully I can find some Goldens as well.

The undercolour of the Cad Red Hue being so dull as you report is because of the yellow and the violet reflectance cancelling out each other (similarly to some Alizarin Crimson replacements which show a similar dullness for exactly the same reason when they try to match AC's warm undercolour). The red component of the mix, Naphthol Carbamide, is the violet-red I use. It's reasonably light-valued so it is slightly more versatile than more pronounced violet-reds and I don't need to mix especially bright violets. If I did I would need to get something like PV19 or PR122.

I don't expect to sway you about the cadmiums but in normal studio practice there really is no way for it to get into your system. Only airbrushing, pointing your brushes in your mouth and sanding your work prose any real threat.

There are numerous organic combos that mix decent oranges, but they won't have the exact properties of cadmiums, most especially in regard to opacity. Of the orange- to middle-reds PR112, PR149, PR170, PR178, PR214, PR251, PR254 and PR260 are worth looking out for. At best these are only semi-opaque but they are not as transparent as some organics.

"Scrubbing"to show undertone is actually better than washing out - it helps to maintain intensity when the paint dries. When you thin the medium the pigment can settle on the surface which can appear duller (depends on the pigment, the amount of binder and other factors).

The problems with PR122 is mostly to do with its masstone; as you know, used thickly it just does not look like 'magenta'. In thin washes or glazes it is better but its reflectance curve is way off the ideal magenta pigment. In case you are wondering the pigment most commonly used in process printing is closer (I've seen the reflectance spectra in trade magazines and it is still far from the ideal) but not lightfast so while it yields better results they don't last.

Phthalocyanine Blue GS is usually a pretty good match for cyan, but it varies a lot from manufacturer to manufacturer and of course in masstone it won't yield the right results, no matter how close it is. Remember, in printing the layer of ink is fantastically thin, so there is a lot of light reflected through it from the ground - hence the need for high-white papers for top-class reproduction.

If I could offer some advice, shelve the idea of finding the perfect CMY - they don't exist - the full-colour process printing we see every day on the newsstand falls far short of perfect colour reproduction and yet we all accept the results as being bright, colourful and realistic. There's nothing wrong with using a very limited palette like this of course, but it would be better to just pick three good examples and learn to use them well. PY1, PB15:3 and PR122 will give great results, not perfect, but amazingly varied. PY184, PB15:1 and PV19 will give similarly varied results but of course not with the same spectrum. And remember this is dependant on painting thinly; if you want to be able to paint thickly you will need to utilise a variety of colours for their varied properties and a double-primary palette is really the bare minimum that will do the job.

You should give W&N Finity a try, even if only to decide for yourself whether they suit your preferences. Their Permanent Rose is a particularly good example of PV19 for instance and their PB15:3 might be the greenest available in acrylics. They are the best acrylics I have used, as I've said previously (I haven't tried Golden HB yet) but they may not be to your taste.

Happy mixing!

Einion

Patrick1
09-09-2001, 12:38 AM
Einion, don't get me wrong, I don't mean that Liquitex high visc. cad yellow light hue (PY 184) is orangy..it's just that it's slightly less lemony than my Grumbacher Finest PY3 and clearly less lemony than cadmium primrose (Golden) (even though the handprint colorwheel, which I refer to a lot, says it should be about one of the most lemony yellows).

I have a tube of Liquitex h.v. raw umber, and I agree that it's quite yellowish/greenish...a
yucky color. What is raw umber used for in landscapes, without dirtying up the colors? Underpainting? Tree trunks? I ask because in Blue And Yellow Don't Make Green it says it's often considered an essential color. I personally much prefer burnt umber, it's certainly warmer and richer, and it's not at all drab. ALso, my W&N Artisan burnt sienna seems very reddish and firey...(which is a good thing to me) much more so than the acrylic burnt sienna I have.

You should see if Golden sends hand-painted color charts to Ireland ( I believe that's where you are). I asked for a color chart for their h.b. acrylics and they sent a lot of other info too. I framed the color chart and contstantly pluck it off the wall to check out certain colors. I know that Winsor& Newton also makes hand-painted color charts, I'll see if my local art store has them or can get some, I'd even be more than willing to pay for one...hand painted color charts are that useful.

shimo
09-26-2001, 07:26 PM
Thanks Domer for that very valuable info. I believe you are completely correct. I will not go into the details of the science behind color diffuse reflection but here's a summary of the best results you can expect from 1) Pure colors 2) Mixing colors.

1) Pure colors - Viewing a painted object in its purest color in perfect white light can obviously expect, at best, 100% diffuse refraction.

2) Mixing colors - Viewing a painted object that was mixed from 2 pure colors in perfect white light can expect, at best, 50% diffuse refraction.

What does this mean? In short, if you need a very bright color, lets use purple, you'll need to use a tube of purple instead of mixing red and blue. If most cases you probably will not need such a bright purple so a mix will due just fine.

The reason for this, as I understand it, is as follows:

Mixing colors:
Let's mix red and blue. Basically we end up with two particles, one that reflects only red, one that reflects only blue. If we sum this up simplistically we end up with 2 units of light reflection; i.e., 1 red, 1 blue.
Pure color out of the tube:
Now lets take the same with a pure color of purple. If we isolate 2 particles, as we did from above, we get two purple particles. Each particle reflects both red and blue. The net results in 4 units of reflection.

So that's how I understand it. Please, any scientific people here please tell me if I'm wrong.

Einion
09-27-2001, 08:45 PM
Hi Patrick, just back from London where I bought PY184. W&N also supply the colour but in the interests of trying other brands more I got the Liquitex version (medium viscosity). Shame Golden don't offer this pigment, I would have tried theirs instead. BTW I got one of their handpainted colour charts, thanks for recommending them - worth its weight it gold, which makes the 」1 all the more reasonable!

The bismuth yellow, as I am sure you will agree, is very clean and saturated in masstone and the label clearly shows it to have virtually no bias (plus a chroma of 12.5 which is very high, especially for a yellow). I have done a few limited mixing experiments with it and I was amazed at how dull both the oranges and the greens it mixed were, just as you indicated. This is by no means a problem for me as I favour subdued colours so I'm looking forward to using it, plus it also offers significantly greater opacity than some yellows and is ASTM I.

I tried it initially with Quinacridone Rose, Naphthol Carbamide, Cerulean Blue and Phthalo Blue GS and even with PB15:3 the greens were quite subdued. Needless the say the oranges were dull too but then I mixed it with Cadmium Red Light and what a difference - bright, clean oranges almost on a par with those mixed with Cad Yellow - it even mixed a very similar masstone to Perinone Orange. Just goes to show that if you choose one good contender for a given mixed hue it certainly can help.

The opacity isn't the explanation for its surprising mixing characteristics - Cad Yellow Medium mixes better greens with PB15:3 for example. There are two possible explanations I can think of and the one I favour is that it is very clean chromatically, i.e. it does not reflect a lot of orange or green light. Hence the relatively subdued mixes of either.

Shimo, your explanation about diffuse reflection holds for some cases but it is not a general rule. If you take colours such as Phthalocyanine Green or Ultramarine Violet, compared to mixed hues you would get something of the difference you mention. Oranges on the other hand are the best example of an exception as there are single-pigment examples than can be closely matched by mixing.

The units of reflection idea is a bit flawed. If we take violets as per your example, it is the violet light in the red and blue that provides the end result, not an optical mixing of red and blue light. Although it is an extreme case you can for example mix violets of higher saturation and chroma than Manganese Violet, hence there is more violet light being reflected by the mix than by the pigment particles of the violet.

Einion

shimo
10-18-2001, 05:54 PM
Einion,

Thanks for info. I'm sure we both understand color mixing. We're probably saying the same thing.

Hey, to simplify this I found a cool web page at Winsor Newton. Scroll to the bottom of the page at "SUBTRACTIVE COLOUR MIXING"

Color Mixing Tips & Techniques (http://www.winsornewton.com/Main/Sitesections/EncycloSctn/Techniques/TipsTech98/MXNGTrmnlgy3.html)

SUBTRACTIVE COLOUR MIXING is what happens when you mix paint; i.e., colors get less pure. The 3 primary colors are Red, Yellow and Blue.

ADDITIVE COLOUR MIXING is the result of mixing light sources together; colors get brighter. The 3 primary colors here is Red, Green, and Blue. Example of ADDITIVE COLOUR MIXING is a television, which has three color elements Red, Green, Blue.

Cool, isn't it.
Shimo

Einion
10-24-2001, 10:22 PM
Originally posted by shimo
...I'm sure we both understand color mixing. We're probably saying the same thing.
Nope, 'fraid not :)

You have to go way deeper than The 3 primary colors are Red, Yellow and Blue as this is such a massive oversimplification of things that it is next to useless in practical terms, especially when you want to mix a wide range of colours across the spectrum from a limited palette. Here's a particularly good example: what do you think the result of a mix of Phthalocyanine Green BS and Quinacridone Violet would be? The result is impossible to predict without a much deeper understanding of colour mechanics (if you have these two colours try it and then add a little white, I知 sure you'll be surprised at the result).

Have a look through the older threads here, there are a number of good exchanges worth checking out that are a good start for further research.

Happy hunting,
Einion

Patrick1
10-25-2001, 01:17 AM
Einion, nice to hear you talking color theory again.

[QUOTE]Originally posted by Einion

[what do you think the result of a mix of Phthalocyanine Green BS and Quinacridone Violet would be? The result is impossible to predict without a much deeper understanding of colour mechanics (if you have these two colours try it and then add a little white, I知 sure you'll be surprised at the result)."

Incidentally, I tried something very similar about a month ago. I was doing some color patches
to see if it is possible to mix a black virtually indistinguishable from a tube of black.

One of the color pairs I tried is quinacridone magenta (PR122) + phthalo green BS, which is very similar to what you suggested. Here it is:

http://www.wetcanvas.com/Community/images/24-Oct-2001/complementariescloseup.jpg

It's the two patches in the middle square. The
left patch is quin. magenta + phthalo green BS, and the patch just to its right is the same mixture plus titanium white.

My first guess would be that it would give something close to black. I couldn't believe the results...a dark blue. Kind of like prussian blue. Not as saturated, but surprising nonetheless. Adding the titanium white made it even bluer. In the square to the left, I used phthalo green YS instead of BS, and although less saturated, it still gives a very noticable blue.

Since I've read Blue And Yellow Don't Make Green, I now know why this is: both the green and magenta reflect some blue light, and when mixed, the only color that escapes the paint film here is blue.

You seem to be saying that the 3-primary system/way of thinking is next to useless in practical terms. If you're saying that two primaries don't "make" a secondary (the way we were all taught),I agree. This common "three primaries is all you need" way of thinking prevents understanding what really happens when paints mix.

If you're saying that using only 3 primaries (be it RYB or CYM or whatever)is next to useless at mixing colors in practice, I don't agree. It has
major limitations, but it definitely does have its place.

bruin70
10-25-2001, 03:45 AM
of course,,,this all depends on one's taste in color or HOW he sees it.

try to get people to agree on what is orange.....{M}

Einion
10-25-2001, 08:22 PM
Thanks Patrick, nice to see more of your colour-mixing experiments. Yep, spot-on, it mixes a very low-value dark with a just-discernible blue bias in masstone under most lighting. In actual fact left with a semi-matt finish it was distinctly darker than two of my single-pigment blacks but when wetted Carbon Black and Bone Black both achieved their true depth and showed how much darker they truly are (Mars Black was not as dark but much more neutral BTW).

Originally posted by Domer
Incidentally, I tried something very similar about a month ago. I was doing some color patches to see if it is possible to mix a black virtually indistinguishable from a tube of black.
Have you tried a mix of Phthalo Green BS and Quinacridone Carmine? These are a near-perfect complementary pair. Unfortunately I have not been able to find PR N/A in acrylics but some versions of Quinacridone Rose and Anthraquinoid Red are nearly identical in hue.

BTW, which mix you tried did you think was closest to black in masstone and undercolour?

Since I've read Blue And Yellow Don't Make Green, I now know why this is: both the green and magenta reflect some blue light, and when mixed, the only color that escapes the paint film here is blue.
Yep, that's exactly what I was getting at. It is only with this kind of knowledge of what actually happens inside paint films that you begin to get a true feel for accurate colour mixing.

You seem to be saying that the 3-primary system/way of thinking is next to useless in practical terms... If you're saying that using only 3 primaries (be it RYB or CYM or whatever)is next to useless at mixing colors in practice, I don't agree. It has major limitations, but it definitely does have its place.
Well the major limitations are what make it next to useless ;) but I was speaking particularly in terms of thinking that red, yellow and blue are the true subtractive primaries, which of course they are not.

Originally posted by bruin70
try to get people to agree on what is orange.....{M}
Good point Milt. Cultural differences, for example, make for funny shifts in perception of colour; one doesn't think of them in quite the same way when one names them differently. Germans for instance don't distinguish between yellows, oranges and earth colours quite the way most English-speakers do and I'm sure there are other similar differences from country to country (or language to language) for similar reasons.

Einion

shimo
10-26-2001, 10:30 PM
Einion :-) Are we just going to keep going in circles. I think you missed the concept of my last post AND the web page I pointed out. "Mixing Colors is Subtractive Color Mixing." If you read that web page you would understand Subtractive Color Mixing makes color less pure, darker each time you mix.

:-) Again we are saying the same thing... Additive color mixing makes colors brighter.

Again, enjoy the article.
http://www.winsornewton.com/Main/Sitesections/EncycloSctn/Techniques/TipsTech98/MXNGTrmnlgy3.html

shimo
10-26-2001, 10:50 PM
Originally posted by Einion
You seem to be saying that the 3-primary system/way of thinking is next to useless in practical terms. If you're saying that two primaries don't "make" a secondary (the way we were all taught),I agree. This common "three primaries is all you need" way of thinking prevents understanding what really happens when paints mix.

If you're saying that using only 3 primaries (be it RYB or CYM or whatever)is next to useless at mixing colors in practice, I don't agree. It has
major limitations, but it definitely does have its place. [/B]

Hi Domer,

As far as "the way we were all thaught" I think many of us get confused between additive and subtractive color theories! You can make ANY color with Additive mixing but NOT with subtractive mixing. Mixing colors is Subtractive mixing.... So it's a good idea to keep many of those vibrant tubes of paint :-)

Keep up the experiments. They're very educational.

Einion
10-27-2001, 02:05 AM
Shimo, no offence but since you have been here such a short time I would respectfully suggest again that you take the time to look through the previous exchanges here and in other threads. It would make the point unequivocally that the very basic information on colour theory on that page is far from comprehensive, plus it would acquaint you more fully with the other members here. It would for example show you that I am not in the least confused about the difference between additive and subtractive colour.

Einion ;)

Patrick1
10-27-2001, 06:06 AM
Shimo, the reason I and Einion often talk about light in subtractive mixing is not because we're confusing additive and subtractive mixing, but because light is a HUGE part of subtractive mixing, even though the vast majority of explanations of subtractive mixing ignore light altogether. Why? Because subtractive mixing is very complex and the '3-primary' explanation is
a way of putting it all into a convenient, simple little package that is easily taught...
"blue + yellow make green, red + blue make purple, and yellow + red make orange". Even the Winsor & Newton page you recommend starts to get into the limitations of the 3-color system and says that a 6-color split primary system will give cleaner mixes. But, like most other explanatations, it doesn't say WHY.

To understand why, you must look at which colors of light the paint film absorbs and which ones it reflects. This is where you start to really understand paint mixing. I myself am just learning about paint mixing beyond the 'old' way of thinking. If you'd like to what I'm talking about, check out:

http://www.schoolofcolor.com

This is the website of Michael Wilcox, the author of Blue And Yellow Don't Make Green. It has a few tidbits from his book. The book explains, in simple terms, what actually happens when paints mix. If you don't want to order the book, I'd be happy tell you the crux of it all in a couple of paragraphs...it's not too complicated...it's just different from what we were all taught...be prepared to have some of what you know about color mixing turned upside down!

Then there's the Handprint website's section on Watercolors. Much of the color theory info applies to other types of paint also:

http://www.handprint.com/HP/WCL/wcolor.html

This is by far, the largest, most complete
explanation of color theory I've seen anywhere.
It's unbelievable that this was written by just one author. The wealth of information here is staggering and fascinating. And it's frequently revised. This guy REALLY knows his stuff. But it's so much stuff, I don't think I'll ever absorb it all. It fries my brain trying to read and understand it all. But it's so interesting I cant help trying.


Einion, nope, don't have quinacridone carmine.
In fact, I really don't have that many colors, mostly just warm and cool variations of the 3 primaries, the 3 secondaries, white, and a few others I use a lot for landscapes. But in my experience too, some of the best blacks are made by mixing a middle green like phthalo green YS with a color about halfway between red and magenta.

As to my trying to mix black...My 'control sample' black was mars black (the only black I have). In masstone, it's hard to tell, all the blues + browns look very close (most look actually darker than mars black in masstone). In undertone (scrubbed) the closest was probably ultra. blue + burnt sienna. In undertone achieved by adding titanium white the closest was clearly ultra. blue + burnt umber.

But I could've made many of the combinations better (closer to mars black) by varying the proportions a little differently. A lot of the swatches were halfway painted when I realised that I should've used slightly different proprtions, but by then it was too late. Doesn't matter. I should be spending more time painting pictures than doing this stuff :)

That's interesting about how Germans see oranges, yellows and earth colors. You mean they
have just a few words for all of them? It's like the opposite of the Inuit; I read that they have
twenty-something different words for different kinds of snow.

shimo
10-27-2001, 02:44 PM
Again, we are saying the same thing.
The page was only to reiterate my point -- WE AGREE ON THE SAME THING. I never said the web page covered all color theory in detail. It clearly demonstrates three primary colors getting darker when subtractive mixing.

Sir, I let most of your comments slide but I now feel the need to defend myself.

Your quote:
Shimo, your explanation about diffuse reflection holds for some cases but it is not a general rule. If you take colours such as Phthalocyanine Green or Ultramarine Violet, compared to mixed hues you would get something of the difference you mention. Oranges on the other hand are the best example of an exception as there are single-pigment examples than can be closely matched by mixing.

Sir, you are misquoting! I never said you can make any color with the three primary colors. My own personal color theory clearly states you cannot make all colors such as the bright purples by color mixing. I never gave an example of Phthalocyanine Green and Ultramarine Violet. First I would never make such an assumption of the results unless I saw the color in full spectrum. All manufactures colors usually vary. If you would post a image of your Phthalocyanine Green and Ultramarine Violet I could easily predict the out come. Oranges are not my exception in color mixing prediction. I have never yet miscalculated a color mix.

Your quote:
The units of reflection idea is a bit flawed. If we take violets as per your example, it is the violet light in the red and blue that provides the end result, not an optical mixing of red and blue light. Although it is an extreme case you can for example mix violets of higher saturation and chroma than Manganese Violet, hence there is more violet light being reflected by the mix than by the pigment particles of the violet.

Sir, your statement is not clear enough for a complete response. All my examples are for white light. White light does include red and blue. My example used a mixture of red and blue. A mixture of red and blue in white light will not reflect the pure violet ray, but rather a red and blue ray. If you take a pure color of Single violet pigment AND that pigment is pure in itself (that is, it reflects the violet ray), then it obviously reflects pure violet ray and not red and blue rays.

Your quote:
You have to go way deeper than The 3 primary colors are Red, Yellow and Blue as this is such a massive oversimplification of things that it is next to useless in practical terms, especially when you want to mix a wide range of colours across the spectrum from a limited palette. Here's a particularly good example: what do you think the result of a mix of Phthalocyanine Green BS and Quinacridone Violet would be? The result is impossible to predict without a much deeper understanding of colour mechanics (if you have these two colours try it and then add a little white, I知 sure you'll be surprised at the result).

Sir, I never stated any such thing! Again, my point was clearly to reiterate we agree. Any novice should know color mixing is complex. Not only must one understand the color spectrum, but subtractive theory and an understanding of color transparencies. Again, the web page clearly shows subtractive color mixing makes colors darker. My original post stated that mixing red and blue paint will never result in a near 100% pure purple.

Please feel free to post any paint images and I will give you my prediction if you feel my theory is in error.

I'll try to post more detail on my color theories later when I get some time. Until then, do we agree that mixing different colors makes a darker color? :-)

Ok, peace to you.

shimo
10-27-2001, 03:10 PM
Hello Domer,

I was writing my post during you last post.

Originally posted by Domer
Shimo, the reason I and Einion often talk about light in subtractive mixing is not because we're confusing additive and subtractive mixing, but because light is a HUGE part of subtractive mixing, even though the vast majority of explanations of subtractive mixing ignore light altogether. Why? Because subtractive mixing is very complex and the '3-primary' explanation is
a way of putting it all into a convenient, simple little package that is easily taught...
"blue + yellow make green, red + blue make purple, and yellow + red make orange". Even the Winsor & Newton page you recommend starts to get into the limitations of the 3-color system and says that a 6-color split primary system will give cleaner mixes. But, like most other explanatations, it doesn't say WHY.


Yes, again we agree :-) Let's get beyond this OK? As I believe you pointed out, Winsor & Newton page is starting from the basics. Painters must learn the basics of three primary colors. Additive color theory is very simple, subtractive is not. I'm glad you picked up on the part in Winsor & Newton page that 6 color palette gives cleaner mixes. That's something painters must learn initially.




This is the website of Michael Wilcox, the author of Blue And Yellow Don't Make Green. It has a few tidbits from his book. The book explains, in simple terms, what actually happens when paints mix. If you don't want to order the book, I'd be happy tell you the crux of it all in a couple of paragraphs...it's not too complicated...it's just different from what we were all taught...be prepared to have some of what you know about color mixing turned upside down!


Yep, I agree, we do need green, but a good pure blue and yellow paint does make green. Although it's not pure green with a loss in Saturation and brightness. Yet who wants to paint with pure green. I just wanted to make that clear to other painters. Saying Blue and Yellow do not make green hue will confuse beginners. The reason blue and green paint make a darker color than the original blue and yellow is due to pigment absorption. Light shining on the blue pigments absorb at most half the yellow rays in the white light and the yellow pigments absorb at most the blue ray.

Getting into more detail, the transparency of the paint makes a huge difference in the color end result color brightness and saturation. Transparent colors mixes are less pure than Opaque color mixes. But we need our transparent parents so don't throw them away :-)

Einion
10-27-2001, 06:14 PM
Hi Patrick, good post. Ditto on the Handprint site, stunning the amount of work it represents. And you are completely right, it is not enough to know that yellow and blue give green when mixed, because then how do you explain what you get when you mix black and yellow? It is the why that is fundamental in practical colour mixing.

Ultramarine mixes with Burnt Sienna and Burnt Umber are some of my favourite darks too. I'm assuming that by and large your swatches have a semi-matt to finish? If you have the time wet down the panel and have a look at the shifts in value and hue, this is a good simulation of the increased depth you get if you gloss varnish, something worth considering depending on the finish you like.

You will be amazed how something you did as an exercise sticks in your mind for years, only to pop up when you need it down the road. Technical exercises and diligent practice are the cornerstone of art education, and provide a sound basis for future expression so rest assured, the time you are spending will stand you in good stead. Remember in 19th c. atelier they did study after study of the same statue or cast - sometimes for weeks at a time - learning to see, to draw and to work hard all at once.

Einion

Einion
10-27-2001, 09:00 PM
Shimo, no need to jump on us. Patrick and I are only trying to be helpful. We don't know your art background or how much experience you have but your posts so far have not shown any deep understanding of the nature of practical colour mixing otherwise we would not be questioning certain of your statements. Just repeating "mixing colors is Subtractive mixing" and "colors get darker when subtractive mixing" merely reiterates a basic point (that we all know) but provides no useful springboard for practical application, something I try to provide as much as possible.

I don't want to get into a tit-for-tat exchange on these issues, especially if you're not willing to approach the discussion with an open mind, but I really have to take you to task for many of the things you have said. I am NOT saying everything you say is incorrect, but much of your posts are full of mis-statements and outright errors and, at least to me, seems confused on the differentiation between theory and practical application. One needs to be clear and precise when discussing this sort of thing to make a distinction between personal opinion and observable fact, willing to give specific examples to back up your arguments and if an example disproves something you have said accept that it is incorrect.

No, you never said the page covered colour theory in detail, but you did feel the need to post the link to it twice, holding it up as some sort of paragon which implied that you thought of it in this way. Did you really think either Patrick or myself was unfamiliar with anything on that page? If you had taken the time to check past posts you would see that it was actually quite insulting, much as it would be if provided links for you to look at with lessons on rhetoric and grammar.

Going back to your first statement, I did not misquote you. Taking two specific examples to prove the falsity of your assertion is not a misquote. This idea of yours (I have not heard of it before so I'm assuming it's yours) about "color diffuse reflection" is flawed*. I'm sure you think the idea is sound and it probably makes logical sense to you but look back at what you posted and read it from the point of view of someone who can't see inside your head - it does not make sense. You didn't give any specific examples, you stated it as a general principle; I showed two examples where this principle is incorrect, therefore what you said is wrong. Let me give further examples so there's no doubt: take Chromium Oxide Green, Manganese Violet, Indanthrene Blue, Benzimidazolone Maroon and Perinone Orange. All are pure colours by your definition yet each can be matched in hue and exceeded in chroma by mixed colours so... again we are talking about real-world colours and real-world conditions here, not theoretical ideals.

My question about PG7 and PV19 was to make a point that the basic idea of three primary colours just can't cope with mixes of this kind. And no offense but stating that you have never yet miscalculated a colour mix is quite a claim. I don't want to come across as a smart*** but that would make you the first artist in the history of the world to have never made an error of this kind. I know a number of experienced full-time artists who are only too willing to admit they are still learning about the intricacies of colour-interaction after decades of hands-on experience.

The way you stated your second point in your first post was not clear and your explanation just now is in fact worse:
Originally posted by shimo
A mixture of red and blue in white light will not reflect the pure violet ray, but rather a red and blue ray.
You are implying with this statement that a mixture of red and blue pigment, which reflecting red and blue light respectively, combine optically to make violet. I'm sorry this is completely incorrect. If a mix such as this did in fact reflect red and blue light out to the viewer then they would combine to make magenta, not violet, as anyone who looks at the additive colour circles can see.

What actually happens inside a paint film is that the blue light reflected from the blue pigment particles and the red light reflected from the red pigment particles is, by and large, absorbed by the other pigment, leaving the remaining reflected spectra to be seen by the observer. In other words, the violet reflectance of each pigment is what accounts for the violet in the resulting mix - the basis of colour bias theory. This general rule holds true in all media and any colour, which is why it is so useful in practical terms. Subtractive colour is not about optical combinations, you are confusing additive and subtractive theory.

Originally posted by shimo
If you take a pure color of Single violet pigment AND that pigment is pure in itself (that is, it reflects the violet ray), then it obviously reflects pure violet ray and not red and blue rays.
And you accuse me of not being clear enough? What I think you are saying here is based on a single-pigment violet that reflects nothing but violet light, is that right? If so you would be completely correct but the problem is that there is no such pigment! ALL pigments reflect some other wavelengths of light without a single exception - some study of the reflectance spectra for pigments would show this statement to be incontestable. This is why one has to make a clear distinction between a theoretical argument and practical application.

Originally posted by shimo
Sir, I never stated any such thing! ...My original post stated that mixing red and blue paint will never result in a near 100% pure purple.
Stated what? That the three primaries are red, yellow and blue? Check you post, that is exactly what you said "SUBTRACTIVE COLOUR MIXING is what happens when you mix paint; i.e., colors get less pure. The 3 primary colors are Red, Yellow and Blue."

*Your original post also said that, "at best, 50% diffuse refraction". Why diffuse and why refraction? This is reflected light we are talking about here, not refraction. There is little refraction to be considered in typical paint mixtures, especially in opaque colours. It is also not correct in regard to amount as you can get much better violet reflectance than 50% from a mix of certain pigments, as appropriate blues and reds (or magenta and cyan) can each reflect more than 50% of the incident violet light. Again, this is a fact, not an opinion.

Originally posted by shimo
Painters must learn the basics of three primary colors.
Why? If the idea has a fundamental flaw, why should they learn this first? Colour-bias theory is so staggering simple you can learn it in about five minutes (I'm exaggerating but you get my point). Even children can grasp the basic principle in a single day when given practical examples. There are no three primary colours in real terms (i.e. three colours that can be mixed to make ALL other colours) and if there were they would be cyan, magenta and yellow and not blue, red and yellow.

Originally posted by shimo
Saying Blue and Yellow do not make green hue will confuse beginners.
Really? I know from practical experience that if you tell a bunch of students what 'primary' means and that blue, red and yellow are the primaries, then get them to try to mix specific secondaries they are extremely confused as to why it does not work in practice. This same problem has been encountered in school art classes and even art schools for generations and only recently have we begun to see a fuller understanding and better teaching of the issue.

Originally posted by shimo
The reason blue and green paint make a darker color than the original blue and yellow is due to pigment absorption. Light shining on the blue pigments absorb at most half the yellow rays in the white light and the yellow pigments absorb at most the blue ray.
No, it's because of light absorption. And why does the light shining on the blue pigment absorb only half the yellow? (I presume you mean the same in reverse too, but the way you have written it actually says all of the blue light.) Nearly all the incident blue light (over 90%) is absorbed by Cadmium Yellow Medium and almost all of the yellow light is absorbed by Ultramarine for instance.

Originally posted by shimo
Additive color theory is very simple, subtractive is not.
I would have to disagree but this is a personal opinion. I just don't think subtractive colour mixing is difficult, it's complex yes, but not difficult. Subtractive theory is easy IMB because you can do it right there on the palette, see the results and learn from them. With experience you can do it quite successfully in your head.

Originally posted by shimo
Transparent colors mixes are less pure than Opaque color mixes.
Pardon me? If by purity you mean chroma, name an opaque mix with higher chroma than PY3 plus PG36. Or even PY3 and PB15:3, or PV19 (violet) and PB15:3.

I for one would appreciate the courtesy of singing with your real name as both Patrick and I do.

Einion

P.S. Suggested reading:
Colour Chemistry, Heinrich Zollinger and The Physics & Chemisty Of Colour, K. Nassau.

shimo
10-28-2001, 01:12 AM
I have added this statement after typing this post. This thread is nearly useless and hey I wouldn稚 bother reading it myself. Although I would read the last ~four paragraphs if you want to understand this person know as Einion.
---------------------------------------------------------------------

Einion, Einion, Einion,

I was joyful upon reading your post as I thought, "OK this man is now backing off this ridiculous argument of attacks." But reading further proved disappointing.

Listen, I don't have time for this. I mix paints on the fly often and I never miss. Again, you clearly misquoted my statements. I'd rather be painting than discussing this endlessly with you.

Yes I certainly gave examples.

Second, you cannot exceed in chroma/saturation. Einion, when I read that from your post I knew instantly you have some learning of either definitions or color theory. It is a know fact in science that subtractive color mixing Never exceeds in saturation/chroma. Did you mistype chroma for Brightness?

...

I don't mean to be rude but please forgive me that I read not further than your chroma statement. I知 certain you understand color theory. Perhaps your experiments have slightly confused you. Sometimes it痴 visually difficult to tell what changed-- saturation/chroma, brightness, or hue. Chroma never exceeds the original colors. I appreciate your time and effort. When my patience regains I'm sure we値l be at it again another subject hopefully Einion. :-)



OK, I lied :-) I could not resist so I read some more of your post.

Your quote:
You are implying with this statement that a mixture of red and blue pigment, which reflecting red and blue light respectively, combine optically to make violet. I'm sorry this is completely incorrect. If a mix such as this did in fact reflect red and blue light out to the viewer then they would combine to make magenta, not violet, as anyone who looks at the additive colour circles can see.[/QUOTES]

Einion, no offense but did you study any physics? Do you not understand the human eye? The human eye cannot distinguish between true violet and a mixture of red & blue. Mmmmm, please Einion study more.

Your quote:
[QUOTE] And you accuse me of not being clear enough? What I think you are saying here is based on a single-pigment violet that reflects nothing but violet light, is that right? If so you would be completely correct but the problem is that there is no such pigment! ALL pigments reflect some other wavelengths of light without a single exception - some study of the reflectance spectra for pigments would show this statement to be incontestable. This is why one has to make a clear distinction between a theoretical argument and practical application.
That痴 a blatant attack! Obviously there is not perfect material that reflects 100% violet and I never said such a statement.

Stated what? That the three primaries are red, yellow and blue? Check you post, that is exactly what you said "SUBTRACTIVE COLOUR MIXING is what happens when you mix paint; i.e., colors get less pure. The 3 primary colors are Red, Yellow and Blue."

What are you talking about? If you池e going to quote something, please quote the entire sentence.


My quote:
Originally posted by shimo
Painters must learn the basics of three primary colors.
Why? If the idea has a fundamental flaw, why should they learn this first? Colour-bias theory is so staggering simple you can learn it in about five minutes (I'm exaggerating but you get my point). Even children can grasp the basic principle in a single day when given practical examples. There are no three primary colours in real terms (i.e. three colours that can be mixed to make ALL other colours) and if there were they would be cyan, magenta and yellow and not blue, red and yellow.

The fundamental laws of subtractive color mixing are not flawed. Your brain is what痴 flawed :-)

OK, if anyone studied our post they would know for fact you intentionally misquoting. For example, in your last post you quoted me as follow:

Originally posted by shimo
The reason blue and green paint make a darker color than the original blue and yellow is due to pigment absorption. Light shining on the blue pigments absorb at most half the yellow rays in the white light and the yellow pigments absorb at most the blue ray.

Then below that you follow with this statement:

No, it's because of light absorption. And why does the light shining on the blue pigment absorb only half the yellow? (I presume you mean the same in reverse too, but the way you have written it actually says all of the blue light.) Nearly all the incident blue light (over 90%) is absorbed by Cadmium Yellow Medium and almost all of the yellow light is absorbed by Ultramarine for instance.

Hmmm, I really do not wish to converse with you anymore. You have issues.

shimo
10-28-2001, 01:42 AM
Everyone, please accept my apology for the way this thread took off. Some nice gentleman, not Einion, sent me a very kind instant message. I was feeling terrible upon responding to the last post. Thanks to this gentleman I'm feeling better.

I felt the need to defend myself. I honestly thought what I still perceive as blatant attacks would stop. They merely escalated and I should know better than to feed fire with air. Anyhow, I look forward to Domer's exciting experiments.

SueFletcher
10-28-2001, 01:43 PM
There seem to be problems with semantics in communicating here, also, not clearly defining what each is trying to say(tho lots of effort in that direction!)...perhaps more trying to be "right" than understanding. Sometimes when lots of effort in communicating produce only disagreement, it's better to simply acknowledge this and look elsewhere for answers.;)

llis
10-28-2001, 01:52 PM
Originally posted by shimo
Everyone, please accept my apology for the way this thread took off. ..............Anyhow, I look forward to Domer's exciting experiments.

Shimo, thanks so much for taking another look at this entire thread and offering an apology. Sometimes it is hard to really grasp the true intent of everyone posting to threads because we are not sitting across the table from each other, but instead have to rely on flat words. I think everyone was really just trying to discuss and help each other understand differing view points and it is good to disagree honorably.

I think it is important to understand that we are all students.... no matter what level we attain and there are always things to learn and new ideas to explore. Glad you are looking forward to more if Domer's experiments. I am too.