PDA

View Full Version : Mixing paint and the true nature of the color


couturej
04-12-2011, 11:05 AM
Well this is one I've been contemplating for a while now. What I'm seeing when I mix is that the true nature of a color can only be determined by looking at the colors undertone or even better when white is added to the color. For example some of my yellow oranges in masstone when white is added move so far over toward mid yellow that I could consider them to have somewhat of a green shade as pigment manufacturers put it.

This pigment manufacturer is one that clearly states the shade in the groups:

http://guerrapaint.com/

Take a look at the Group drop down menu and you'll see in the Yellow Green shade category pigments like Cadmium Yellow Medium.

I understand that if you were using a color straight no mixing then the masstone would be extremely important.

I actually startied to wonder if the HSB color wheel can be used both for mixing complements and visual complements. It's just a matter of determining the proper nature of the color. I'm not sure if nature is the right word. So forgive me if it isn't.

When I talk about mixing complements. I'm talking about mixing to a true grey.

This thread posted by Bill started me thinking more along these lines:

http://www.wetcanvas.com/forums/showthread.php?t=583719&highlight=mixing+complements

Here's the example he offered that Yellow and Blue yield grey:

http://www.wetcanvas.com/Community/images/12-Apr-2011/84697-13079-Yellow_Blue_Yield_Gray.jpg

http://www.wetcanvas.com/Community/images/12-Apr-2011/84697-swatches_yellow_blue.gif

The mixing complement of a Yellow at a hue angle of 65 is 245. This is the same Blue that people mean when they say blue is the complement to yellow for visual complements. At least close. Here would be a middle yellow (hue angle 60) and it's complement of middle blue at the same saturation and brilliance as the above swatches.

http://www.wetcanvas.com/Community/images/12-Apr-2011/84697-mid_yellow_plus_mid_blue.jpg

So what do you think? I'm not posting this to say this is the way it is. Just exploring the idea. Thank you Bill for opening my mind to the possibilities.

Goldeelocks
04-12-2011, 11:18 AM
Well, it's "sort of" blue I feel. It's a very light violet blue, not blue-blue.

To replicate that colour blue you used, using my 3 primaries, I had to add quina (magenta) to my blue, so you're already moving down the hue space away from blue..getting away from the green, which explains why you're not getting green with ""blue"" and yellow. (Ignore the small red blob, I picked the wrong tube at first)

http://www.wetcanvas.com/Community/images/12-Apr-2011/204698-blueviolet.JPG

left is what I got trying to replicate your color, the right is my "pure" blue lightened to match it in value. The left blue is violet-blue, not blue-blue.

http://www.wetcanvas.com/Community/images/12-Apr-2011/204698-violet_blue.JPG

I don't think visual complements are the same as mixing complements, because paints are "weird", they don't do what the light spectrum does in all it's additive perfection. But I also think most color wheels are wrong, because when they say the complement of yellow is pure blue, it isn't, not even in nature, it's closer to violet, what you are using.


Another thing I wanted to add, your colors aren't that opaque or fat, you can get away with much more if you're using less opaque stuff I feel. With acrylics, put on heavily, you get the heavy shifts, with watercolor I can get away with slight errors, not when I use my acrylics.

sidbledsoe
04-12-2011, 12:55 PM
Here is Grumbacher cad yellow pale and ultramarine violet.
Even this combo yeilds a green.
http://www.wetcanvas.com/Community/images/12-Apr-2011/112587-P1000186.JPG

Einion
04-12-2011, 01:27 PM
Well this is one I've been contemplating for a while now. What I'm seeing when I mix is that the true nature of a color can only be determined by looking at the colors undertone or even better when white is added to the color. For example some of my yellow oranges in masstone when white is added move so far over toward mid yellow that I could consider them to have somewhat of a green shade as pigment manufacturers put it.
Rather than 'true nature', which could come with a lot of baggage, a better term might be mixing character or something like that.

A paint's masstone gives very little idea about the way it might mix (in detail) and the undercolour can indeed offer useful additional info but at the end of the day it's just a pointer since even a simple tint can be a little different to what you'd guess from looking at the undercolour. Bottom line is that any colour needs to be mixed with whatever else to know exactly what the results will be like; which comes back to the old advice to learn your chosen palette inside and out. Which in turn supports the idea of smaller palette over larger palette.

I understand that if you were using a color straight no mixing then the masstone would be extremely important.
Yes, because that could be all you'd see, unless the paint is used thinly enough to show something of what's underneath.

I actually startied to wonder if the HSB color wheel can be used both for mixing complements and visual complements.
From numerous previous threads I would have thought that there would be no doubt left about whether it's possible to predict complementary mixing from a paint's position on a colour wheel. You can plot mixing on any wheel, but no matter how accurate the data used to lay it out there's essentially no predictive value when it comes to complementary mixing behaviour. That is unless you specifically pick every colour for that purpose... and even that falls down in certain instances, where NO paints of opposing hue exist that will mix to neutral - Yellow and Blue being the prime example.

The mixing complement of a Yellow at a hue angle of 65 is 245.
What I'd suggest is trying a similar mix yourself and seeing what kind of results you can get :cool:

Technical note: incidentally if you're posting a picture used previously can you not upload it afresh? (I see I clarified this in post #3 of that very thread!)The pics are fairly small of course but it's better not to have the same image uploaded to the servers two, three or multiple times. Just right-click the image and enclose it in the IMG tags, or quote the relevant post and copy the code in full.

Einion

couturej
04-12-2011, 04:06 PM
Goldeelocks thank you for that but I think your blue is more of RYB primary blue and not at all what would be considered a mid blue in HSB.

Sid, Thank you for posting your example! You might have something with your blue. Remember that with this thread "Desaturated Yellow... Brown or Olive Green": http://www.wetcanvas.com/forums/showthread.php?s=&threadid=928132

If your mixing complementary colors the color would go to olive first before it went to gray because of the hue angle of the yellow you're using. That color looks olive to me. I don't have that color and would be immensely interested in what you get when you add more of the ultramarine violet. If it isn't shifting to another hue we may have a winner or maybe not. Just want to explore the idea.

Enion, Thank you! I'm currently doing mixes. Yes I have a lot of colors which is probably not a good idea. I keep changing my mind regarding my palette and one part of doing the mixes is to narrow down my palette. I think I will need to mix some of my complements but in order to do that it would be nice to have a ballpark idea as to where to seek out these mixing complements. Sorry about the image thing. Didn't realize that.

Goldeelocks
04-12-2011, 04:42 PM
Goldeelocks thank you for that but I think your blue is more of RYB primary blue and not at all what would be considered a mid blue in HSB.


Ah, it's pthalo blue, which is pretty much blue blue pigment. I'm just confused sometimes what you mean with HSB. It means something to you outside of Hue, Saturation, Brightness but I don't know why because they're just terms to describe colors for me. RYB and HSB are not in any similar category for me, so idk what you mean with it in that context. Did you mean that my blue is saturated?

couturej
04-12-2011, 05:09 PM
I understand what you're saying Goldeelocks. In HSB Mid Blue is at an angle 240 degrees. This is from Mid Cyan 180 degrees to Violet 270 degrees. This is in 15 degree steps so you can see the difference from one step to the next. Your Phthalo Blue would be about 210 degrees.

http://www.wetcanvas.com/Community/images/12-Apr-2011/84697-Blues_to_Violet.jpg

couturej
04-12-2011, 05:22 PM
Here's the color wheel for HSB. 60 degrees at the top for Mid Yellow. 15 degree steps. So the color next to mid yellow going toward green would be at 75 degrees next 90 degrees and so on and so on.

http://www.wetcanvas.com/Community/images/12-Apr-2011/84697-HSB_Color_Wheel.jpg

sidbledsoe
04-12-2011, 05:44 PM
Janet, I fired that off really quickly this morning but since then I ran the mixes on out and also put it next to the same yellow and french ultramarine blue mixes. Yes the ultramarine violet looks like a close mixing complement (at the third mix down it looks gray) but the blue does not mix anything close to a gray before it goes bluish green.
Violet is on the left column of mixes and blue is on the right column of mixes.
I sequentially mixed in more of the violet and more of the blue respectively.
And french ultramarine itself is considered to be a violet blue.
http://www.wetcanvas.com/Community/images/12-Apr-2011/112587-P1000187.JPG
Here is my ultra violet tint next to your first image which you say is a blue:
http://www.wetcanvas.com/Community/images/12-Apr-2011/112587-112587-P1000186.JPG http://www.wetcanvas.com/Community/images/12-Apr-2011/112587-84697-13079-Yellow_Blue_Yield_Gray.jpg
They aren't the same value but I must say that the example mix that you first posted looks very violet to me and not what I would think of as a true blue. These color names are getting to be wide ranging in the scope of understanding amongst different people!

Goldeelocks
04-12-2011, 06:12 PM
Ah thanks couturj. Well, it looks too violet for me to be in the pure blue range, and even if I mix it, I get a khaki green, like Sid.

Would be helpful to know what pigment that "blue" is, or even if it's a pure pigment, otherwise we're just eyeballing it. All our screens show different colors too heh.

couturej
04-12-2011, 06:45 PM
Sid, Thank you so much for posting the extended mix. Pretty interesting that it works so well as a complement. Well I do have a color wheel that's obviously based on HSB and 240 degrees is called mid Blue. I agree it's a little more violet then what we're use to. But you can call it what you want but the end result is what's important. The hue angle on Bill's mix is at 246 degrees which is 6 degrees toward violet so yes a little more violet then even what would be considered mid blue in HSB. His yellow though is at 65 degrees which is 5 degrees toward green which would compensate. Your version with the Ultramarine Violet when white is added is about 241 degrees close enough for me.

I don't think Ultramarine Blue is a complement to yellow unless when you added white it looked a lot like the Ultramarine Violet which I don't think is possible.

HSB also called HSV is so much easier then any other color space that I've studied... Munsell, LCH and RYB. The spacing is wonderful. Also almost any graphics program you can get has HSB. Corel Draw and PhotoPaint, Adobe PhotoShop and Illustrator and a free graphics program called Gimp. I also have ColorImpact which the wheel and swatches are generated from. Their are many more I'm sure but these are the ones I own.

Goldeelocks, remember that olive is a shade of yellow so keep adding your "violet blue". Bill used a mix to get to that blue not a single pigment but Sid used Unltramarine Violet to get to similar results. If you're mixing colors to get to that mid blue or "violet blue" you'll need to use your eyes and not think I'm mixing a blue. Just mix the color you see.

couturej
04-12-2011, 07:05 PM
This is from Bill regarding the mix he used for the Blue: "I began with Grumbacher French Ultramarine, Blue, as I recall, and commenced to mix with it, Winsor & Newton Permanent Rose 502. I would mix and test, mix, and test, etc., until I had created a "blue" that, when mixed with this Yellow would produce neutral, or as near neutral as I could determine without the use of a color measuring instrument."

In this thread: http://www.wetcanvas.com/forums/showthread.php?p=7306339#post7306339

sidbledsoe
04-12-2011, 07:39 PM
But I think we do agree that the first green I mixed is indeed somewhat close to a true olive green :D :
http://www.wetcanvas.com/Community/images/12-Apr-2011/112587-P1000188.JPG
now that is accurate color identification!

couturej
04-12-2011, 08:45 PM
:lol: Yes I agree it certainly is a perfect olive green. Now I have my alternative to using black which my heavy handedness usually doesn't give good results. BTW the Ultramarine Violet you used, which manufacturer is it. I need to get some of that.

sidbledsoe
04-12-2011, 09:06 PM
Hey Janet, these were all three Grumbacher brand, Cadmium Barium Yellow Pale, Ultramarine Violet, and French Ultramarine Blue:
http://www.wetcanvas.com/Community/images/12-Apr-2011/112587-P1000189.JPG
I have not used the violet much yet but I am going to now because I like having mixing complements for my main colors, I also am liking that olive green (and of course we know black is :evil:) so thanks to you also!

couturej
04-12-2011, 09:08 PM
You're welcome Sid! Thank you for the info! I love using complementary colors. I've already seen enough threads here that resulted in a battle over black... so mums the word.:angel:

Einion
04-13-2011, 03:00 AM
If your mixing complementary colors the color would go to olive first before it went to gray because of the hue angle of the yellow you're using. That color looks olive to me. I don't have that color and would be immensely interested in what you get when you add more of the ultramarine violet. If it isn't shifting to another hue we may have a winner or maybe not.
Maybe not. In addition to the usual problem of pigment variation - it doesn't work equally with every yellow, can't even be guaranteed to work with two versions of Cad Yellow Light - it's a weak pigment, very transparent and of poor tinting strength. I bought it to experiment with for exactly this purpose but in acrylics at least I don't consider it much use practically (should be a lot stronger in oils of course).

For anyone who isn't in a position to check this kind of thing, in Sid's pic the hue has changed very little in the mix (only around 3) and the hue is still firmly yellow, so it provides a perfect example of a dull yellow that looks green.

Goldeelocks, remember that olive is a shade of yellow so keep adding your "violet blue". Bill used a mix to get to that blue not a single pigment but Sid used Unltramarine Violet to get to similar results. If you're mixing colors to get to that mid blue or "violet blue" you'll need to use your eyes and not think I'm mixing a blue. Just mix the color you see.
http://www.wetcanvas.com/Community/images/20-Aug-2003/3842-thumbsup.gif


Ah, it's pthalo blue, which is pretty much blue blue pigment.
Much greener in undercolour though. Plus mixing is about pigments anyway, not colours in the abstract - two mixes of the same colour created from different pigments will often exhibit different mixing behaviour, sometimes wildly different.

I'm just confused sometimes what you mean with HSB. It means something to you outside of Hue, Saturation, Brightness but I don't know why because they're just terms to describe colors for me. RYB and HSB are not in any similar category for me, so idk what you mean with it in that context.
RYB primaries can be positioned on any colour wheel. In essential structure an HSB wheel just has CMY and RGB equally spaced, i.e. like this:

http://www.wetcanvas.com/Community/images/18-May-2008/3842-Colour_Wheel_Example.JPG

Would be helpful to know what pigment that "blue" is, or even if it's a pure pigment, otherwise we're just eyeballing it. All our screens show different colors too heh.
Monitor variation is an unfortunate problem with this kind of thing. But if you sample the colours in an image-editing software it sidesteps this issue.


They aren't the same value but I must say that the example mix that you first posted looks very violet to me and not what I would think of as a true blue.
It does indeed look distinctly violet to me too (most observers would see it this way). This is due to the Abney effect.

BTW, the tint of your Ultramarine Violet is blue in hue, in the photo at least. The hue varies a little but areas of it are actually the same hue as the tinted blue in Bill's pic.

Einion

couturej
04-13-2011, 08:11 AM
Thank you Einion! It's interesting that tinting strength and transparency is a factor. Good to know. I did notice a difference from acrylics to oils. I no longer work in acrylics but when I did I had picked out all my mixing complements by making swatches with all by colors and was really surprised when I switched to Water Mixable Oils that I wasn't getting the same results with the same pigments. It was very much a hit and miss situation.

BTW I noticed that Weber wOil has the Ultramarine Violet. I think I'll give that one a try before resorting to adding a traditional oil to my Water Mixable Oils palette.

couturej
04-13-2011, 08:03 PM
Just one more thing to add regarding HSB/HSV:

"Their model was based more upon how colors are organized and conceptualized in human vision in terms of other color-making attributes, such as hue, lightness, and chroma; as well as upon traditional color mixing methods e.g. in painting that involve mixing brightly colored pigments with black or white to achieve lighter, darker, or less colorful colors."

I found that extremely interesting.

Here's a link to read more on HSB if you want details on how it actually works: http://en.wikipedia.org/wiki/HSL_and_HSV

couturej
04-14-2011, 08:34 AM
I noticed that Sid's version of Ultramarine Violet actually has a pigment code of PB29 Ultramarine Blue so it probably is possible to get an Ultramarine Blue that works as a complement to yellow and would depend on the way the pigment was processed. I wouldn't think that a manufacturer would actually label it Ultramarine Blue if it has that much red in it but I guess you never know.

sidbledsoe
04-14-2011, 11:08 AM
I noticed that Sid's version of Ultramarine Violet actually has a pigment code of PB29 Ultramarine Blue so it probably is possible to get an Ultramarine Blue that works as a complement to yellow and would depend on the way the pigment was processed. I wouldn't think that a manufacturer would actually label it Ultramarine Blue if it has that much red in it but I guess you never know.
That would seem to be in error, where did you see that Janet?, the tube is labelled PV15. The chemical components of PB29 and PV15 are similar but processed differently thus making PV15 a decidedly a violet hue.

Sodium aluminium sulfosilicate (blue shade); sodium aluminium sulfur silicate (red shade)
Note that my tube is the red shade and the sulfur compound and not the blue shade;
"The creation of Ultramarine Violet is carried out by heating at 200250C a mixture of ultramarine blue and 2.55% ammonium chloride. It takes 4 days of calcining while being exposed to the air until a purple hue is created.
Ultramarine Red is derived from ultramarine violet by heating it for four hours with gaseous hydrochloric acid at 200C. It's creation can also be accomplished by heating ultramarine violet at a even higher temperatures with gaseous nitric acid "
So you can process it further and even create a red with very similar chemical components. This is my tube of ultramarine violet:
http://www.wetcanvas.com/Community/images/14-Apr-2011/112587-P1000191.JPG

Horsa
04-14-2011, 11:18 AM
The map is not the territory.

"Colour", "Pigment" and "Paint" ae not co-equal terms at all and do not properly refer to the same things. Colour is a property, paint is a mixture of pigment and binder.

"Ultramarine Blue", that is "blue from beyond the seas" is properly ground lapis lazuli. French Ultramarine Blue is an artificial substitute.

Colour theories are just that, "theories" or "models" if you prefer. They are not colour "realities" or colour "actualities"

Colour itself is a mental construct determined by culture and language. It has no intrinsic a priori reality. The apriori external reality which preceive as "colour" is wave lengths of electro-magnetic energy with a band that we term the "visual spectrum" because these are the wavelengths that the rod and cone cells in our eyes are responsive to.

The upshot of this is that there are colours which exist but do not fit on the colour wheel. There are colours which it is impossible to mix with standard artist pigments. There are paint mixtures which do not give the expected result.

As a watercolourist I couldn't help but notice that the original post was talking about the "true nature of colour" while actually discussing the behavior of pigments and paints. Try adding white to transparent watercolour without using white guache which many do not consider to be proper watercolour. Note also the comments that the pigments behaved differently in oils and acrylics.

couturej
04-14-2011, 11:19 AM
Hi Sid! Thank you for clearing that up! I actually found it on the DickBlick site here: http://www.dickblick.com/items/00448-6643/#colorpigments

Thank you for the detailed information on the pigments. It's really good to know.

sidbledsoe
04-14-2011, 11:24 AM
Yes, Blick has been known to have errors here and there, in spite of that it is a very valuable resource, I have seen others spot errors here, they are a partner, so we should notify them of the error.

couturej
04-14-2011, 11:24 AM
Horsa I can't talk about how pigments in watercolor behave as I don't have any and I think handprint would be a good resource for that. I wish their was something out their like handprint that was for oils but sadly that isn't the case.

I guess my main focus is how oil paints behave.

Horsa
04-14-2011, 04:17 PM
Handprint for oils and acrylics would be an absolute godsend. It is sometimes opinionated, but having all the information on pigments, mixing, colourfastness, etc in one place is great.

I have been surpised sometimes by the "hidden" colours in my paints. Also known as "Why won't they mix right?" Yellows that go green are a particular trouble I have.

couturej
04-14-2011, 04:35 PM
It would be nice if all manufacturers would provide the same info as handprint.

You might be interested in this: http://www.brucelindbloom.com/index.html?ColorCalculator.html

hanprint uses LCH so lightness/chroma/hue so you can plug in the value chroma and hue (lightness=value) and then put the Lab value into a graphics program. He gives you how many degrees the hue shifts so you could get an idea as to the "hidden" colors maybe. I don't have watercolors so I have no way of testing it.

Einion
04-15-2011, 09:26 AM
Thank you Einion! It's interesting that tinting strength and transparency is a factor. Good to know.
Welcome! Tinting strength is hugely important, since it directly bears on how any mix tends to go. But when there's enough of whatever ingredient it is then sort of irrelevant. The main relevance of transparency in mixing is arguably the effect on the relatively opacity of mixes rather than being directly tied to how it tends to colour other things.

I did notice a difference from acrylics to oils. I no longer work in acrylics but when I did I had picked out all my mixing complements by making swatches with all by colors and was really surprised when I switched to Water Mixable Oils that I wasn't getting the same results with the same pigments. It was very much a hit and miss situation.
You can experience the same thing within one medium going from brand to brand unfortunately. With certain pigments this is the norm since they commonly vary - numerous makers, lots of versions - and for a single tube name like Cadmium Red Light you can see quite a range in the colour of the paints, from particularly orangey ones to shades you'd more rightly think of as being "medium red" (and in fact would be more towards magenta than Cad Red Medium from another brand!) and needless to say the way they mix is going to be all over the place.

But milling differences alone can account for variation in colouring (even for the same material smaller particles can be a different colour to larger particles). This appears to be the reason that two versions of a single-pigment colour that appear very similar can tint quite differently for example.

BTW I noticed that Weber wOil has the Ultramarine Violet. I think I'll give that one a try before resorting to adding a traditional oil to my Water Mixable Oils palette.
Are you building a palette based around complementary mixing pairs?


"Colour", "Pigment" and "Paint" ae not co-equal terms at all and do not properly refer to the same things. Colour is a property, paint is a mixture of pigment and binder.
Absolutely right, but sometimes informal uses such as colour for paint are inevitable, unless you're comfortable repeating the same noun again and again in a short space.

Colour itself is a mental construct determined by culture and language. It has no intrinsic a priori reality. The apriori external reality which preceive as "colour" is wave lengths of electro-magnetic energy with a band that we term the "visual spectrum" because these are the wavelengths that the rod and cone cells in our eyes are responsive to.
Debate territory - regardless of word labelling, a colour can be thought of as truly being as it is defined (numerically for example) since those specific colour attributes exist outside of a linguistic framework.

The upshot of this is that there are colours which exist but do not fit on the colour wheel.
Depends on the colour wheel :)

Einion

couturej
04-15-2011, 09:49 AM
Thank you Einion! Clarification regarding tinting strength and transparency helps immensely.

You're probably right regarding going from acrylics to Water Mixable Oils. I really didn't know a lot about pigments and color in general at that time so probably didn't pay a lot of attention to the undertone and the way they looked different when white was added.

Yes I'm building a palette around complementary mixing pairs. That's why this has been so important to me. I obviously have a gap in my palette right now between Ultramarine Blue and Dioxazine Violet so that Ultramarine Violet or something similar would plug that gap. I'm still debating on the other colors but yellow is essential. I'm doing mixes of half and half right now to determine which color pairs give the most saturated colors but also seeing which may be candidates as complements. So I'm not just mixing but analyzing my results as I go so that's why I'm coming up with questions. As I mix I'm noticing things that in order to analyze what is going on I want to pin down the "why". I'm trying to understand the underlying reason behind what I'm seeing.

couturej
04-15-2011, 11:28 AM
Oops I may have been mistaken about handprint LCH. He does use it on some of his pages but in general I think he uses CIECAM J,a,b which I have no clue as to this color space and couldn't find much on it. So it's useless to me. If I can't see the progression from one color to the next in a color space then I can't visualize what the hue shift is. Too bad he didn't use a color space that is more readily available.

Einion
04-15-2011, 01:37 PM
You're probably right regarding going from acrylics to Water Mixable Oils. I really didn't know a lot about pigments and color in general at that time so probably didn't pay a lot of attention to the undertone and the way they looked different when white was added.
Well you couldn't have predicted in advance the way they'd mix anyway. Only way to know for sure is to actually mix A with B :)

Yes I'm building a palette around complementary mixing pairs. That's why this has been so important to me. I obviously have a gap in my palette right now between Ultramarine Blue and Dioxazine Violet so that Ultramarine Violet or something similar would plug that gap.
That's not really a gap to most people (in the sense of needing to be plugged). Ignoring complementary mixing for a sec, if you think about the spacing of many areas on the palette the gap between French Ultramarine and Diox Purple is already smaller than between many other two paints in a lot of palettes.

The problem of neutralising yellows is a common one and there's no single solution, but I don't think there's any pressing need to struggle to find a complementary pair here since there are four other routes to achieving low-chroma yellows, all of which use existing paints from the palette. I'm not a fan of Ultramarine Violet anyway, but before you buy it I think you should at least try a light tint of PV23. You could then compare the results to using a mixed blue/violet and neutral grey of the same value.

The crux of the issue with building a palette intended to consist of mixing pairs is that it can sometimes require buying multiple examples while searching for the ideal candidates; certain colour combos are safer bets of course but even if the PV15 you're looking at can mix to neutral with a yellow there's no guarantee it'll do this with your yellow. So if it doesn't will you then try a different yellow?

I'm doing mixes of half and half right now to determine which color pairs give the most saturated colors but also seeing which may be candidates as complements.
What you really want to do is mix to the colour midway between the two starting paints and a 1:1 blend often won't give you this because of differences in tinting strength - it could require 1 part of A and 9 or more parts of B.

So I'm not just mixing but analyzing my results as I go so that's why I'm coming up with questions. As I mix I'm noticing things that in order to analyze what is going on I want to pin down the "why". I'm trying to understand the underlying reason behind what I'm seeing.
Some colours just mix the way they mix (or some mixtures just turn out the way they turn out) as frustrating as that might sound. Rather than trying to figure out the why being able to identify what you are seeing - both visually and in terms of fitting it into a mental framework of colour (at least by hue, but in more detail if you can) - is probably more important to focus on.

Oops I may have been mistaken about handprint LCH. He does use it on some of his pages but in general I think he uses CIECAM J,a,b which I have no clue as to this color space and couldn't find much on it. So it's useless to me.
You can read about them right on Handprint :cool: Check out Color 6 - Modern Colour Models. Wikipedia also has entries for both Lab colour space (L*a*b*) and CIECAM02.

Handprint used to use CIELAB but Bruce switched to CIECAM a few years ago.

If I can't see the progression from one color to the next in a color space then I can't visualize what the hue shift is. Too bad he didn't use a color space that is more readily available.
They're the most accurate way of defining colour, that's why he used them. At the heart of it is the ability to define all colours humans can see. In terms of hue spacing they're actually not hugely different to what you're already used to.

Einion

couturej
04-15-2011, 03:51 PM
I did read through the Modern Colour Models and I'm no further ahead. It comes down to that their isn't any programs out their that do CIECAM J,a,b. The information may as well be in latin. Even though I know some latin their is no way a lay person is getting through that jargon.

I tend to disagree that color is as unpredictable as you state. I won't say another word on the subject because that wasn't my reason for posting. I did get the useful info I wanted from Sid and Bill and yes you gave me some food for thought.

Einion
04-16-2011, 03:47 AM
It comes down to that their isn't any programs out their that do CIECAM J,a,b.
It might not help anyway! Photoshop has a Lab mode (as does GIMP) and I'm not sure that working in them really helps getting a gut feeling for the colourspace.

I did read through the Modern Colour Models and I'm no further ahead. ...The information may as well be in latin. Even though I know some latin their is no way a lay person is getting through that jargon.
Agreed, as I said the other day in the thread asking what colour experts have studied, "having a solid scientific background certainly wouldn't hurt for the high-level stuff!" It's bad enough for CIELAB but it gets even more impenetrable when we get to CIECAM.

I tend to disagree that color is as unpredictable as you state. I won't say another word on the subject because that wasn't my reason for posting.
Mixing is, which is right at the heart of the subject of the thread.

I did get the useful info I wanted from Sid and Bill and yes you gave me some food for thought.
*sigh*

Einion

couturej
04-16-2011, 09:36 AM
I did plot out CIELAB about a month ago. What some people do for fun. lol

Actually I was meaning I'm not here to argue. Just thought we could explore and experiment with the idea of if you add white to a color and see where the hue shifts to then use that as its true position on the HSB color wheel if you could predict the complementary color in HSB. To me it looks possible and wouldn't that be fantastic because everyone would access to it. With the Mid Yellow and Mid Blue it looked fairly good but one color doesn't make it true for all colors. I was thinking along the lines that Color Theory is about experimentation and exploration.

Sorry about the food for thought thing... that was out of frustration but uncalled for. You have helped me a lot over the years and in recent posts, including this one and I thank you for that.

If by unpredictable you mean you would need to find out the characteristics of a color before it would have any predictive qualities I would agree. I think even from looking on handprints site that 99% of colors have a hue shift.

I don't have a scanner right now but will be purchasing one shortly so I can look at this more closely. I never thought that HSB could be used to "predict" mixing complements but from what I'm seeing when I'm mixing their may be something to it. At least it would be much better then RYB or the ideal CMYK.

llawrence
04-16-2011, 12:42 PM
Just thought we could explore and experiment with the idea of if you add white to a color and see where the hue shifts to then use that as its true position on the HSB color wheel if you could predict the complementary color in HSB.Sounds intriguing, but which white would you use for the predicting, and in which medium? They all mix differently and provide different quality tints in terms of saturation and brightness. I don't think I've ever done an actual side-by-side comparison, but I'd bet there's also some variation in hue shifts between tints done with different whites.

For that matter, even if you could use this tool to find what should be a complementary mixing hue of another hue, you still wouldn't know if a mix of two particular paints would docilely go through gray, or if they'd veer far and wide through a strange mixing curve. These materials do act mighty peculiar sometimes.

couturej
04-16-2011, 01:28 PM
llawrence, I'm using Titanium White and it works fine. I'm using oils and not sure how it would work in other mediums. I don't have any other whites but on handprint it gives the hue shift for titanium as 0 for most brands except for one, Rowney Artists, which shift by -5. If Titanium White PW6 has no hue shift it shouldn't be affecting the shift in hue when mixed with another color. What some people might be seeing with some whites is it may be more transparent so the hue shift is still their but less gradual. Don't know about that one. That would be a great test as well because we keeping hearing about white that causes hue shift but may be that the person isn't aware that the color that they're mixing the white with is what is shifting the hue.

If they don't mix a straight line to grey and we do have 2 colors that are complementary based on their characteristics then HSB will fall flat and I realize that we may get to a certain point and see it curve. It would be interesting to see at what point it curves or if it does. I'm not so sure it's the materials but it's difficult to say if it's materials or color space.

Painted swatches would be the best way to do it. I can use a color picker on any swatch and determine its HSB value. It isn't a spectrometer but even if our findings show the color going to cool and warm greys and we don't get a drastic shift in hue that would be extremely useful to me and probably other artists as well.

couturej
04-16-2011, 04:45 PM
Sorry meant transparent colors more gradual. Transparent colors seem somehow more delicate. I don't know if that's a good word to describe it.

Einion
04-18-2011, 08:39 AM
Actually I was meaning I'm not here to argue. Just thought we could explore and experiment with the idea of if you add white to a color and see where the hue shifts to then use that as its true position on the HSB color wheel if you could predict the complementary color in HSB.
It sounds plausible but there's no predicting complementary mixing behaviour unfortunately. This is sort of the Holy Grail for complement users, but Bruce MacEvoy's work has confirmed what lots of individual experience has suggested for years - the colours of complements simply don't follow any set rules or patters.

In some cases a paint of roughly X colour will always work well enough to neutralise Y (i.e. you'll get at least a near-neutral) especially for some earth + blue combos, but this doesn't work equally around the wheel and for pigments of all types.

I think even from looking on handprints site that 99% of colors have a hue shift.
Re. tints, it is usual but not as common as that. The cadmiums for example often don't shift in hue when tinted (some do, some don't or the shift is small enough that it's arguably irrelevant - 3 or less is small enough that most people wouldn't be bothered, and for good reason).

Einion

sidbledsoe
04-18-2011, 11:11 AM
earth + blue combos, but this doesn't work equally around the wheel and for pigments of all types.
Einion
What if you construct a wheel that is based on actual pigments and their actual complements, such that all opposites are complementary (mixing)?

This is my problem with wheels, they are invented by people, they can be arranged the way one prefers is convenient to look at.
Where does it say that colors must fit around a circle in any symetrical pattern. Colors exist more in a linear plane of wavelengths spanning different lengths that has nothing to do with a circular wheel don't they?

Einion
04-18-2011, 03:24 PM
What if you construct a wheel that is based on actual pigments and their actual complements, such that all opposites are complementary (mixing)?
There are a couple of wheels like this. There are a few innate problems with them, the most important of which is that in one way or another they don't take into account pigment variation.

This is my problem with wheels, they are invented by people, they can be arranged the way one prefers is convenient to look at.
Yep, you can of course create a wheel to show what you like.

One of the primary implied or stated purposes of a wheel is to position colours accurately, both outright and in relation to each other. You could also choose to create one around the idea of showing complementary mixing behaviour. The issue lies in trying to do both at the same time.

The two goals are in conflict since many pairs of paints that aren't opposite in hue (way off in fact) mix to neutral.

Where does it say that colors must fit around a circle in any symetrical pattern.
Exactly the point - trying to shoehorn paints into a rigid structure that's decided on in advance already sounds bad when you put it that way :D

For me one of the main values of a wheel is to show hue accurately, and chroma ideally - and the positioning is therefore as irregular as necessary. Anything extra, particularly complementary mixing, has to be charted separately unless you want to just display a cherry-picked selection of paints - this is perfectly doable but it means the wheel would generally work for those exact paints, and only those.

Colors exist more in a linear plane of wavelengths spanning different lengths that has nothing to do with a circular wheel don't they?
This is complex but not all hues are shown on a spectrum, since there are multi-spectral hues that only occur when wavelengths are mixed (e.g. magenta).

The wheel structure is in many ways arbitrary but that's not to say it isn't useful. No absolute reason to use one in any way though - all of painting before the idea of colour wheels didn't rely on them obviously.

Einion

couturej
04-18-2011, 04:47 PM
Yes I see your points but to be honest I would have probably agreed with you about a month ago. All of my research on different color spaces and pigments seems to be pointing in the same direction. I've never been one to jump to conclusions unless several things are pointing in the same direction.

1) handprints info on hue shift (first thoughts ya whatever)

2) Bill's swatches showing blue and yellow make grey (ok I recognize that blue it looks like HSB mid blue, color picker shows his blue at 245 degrees and yellow at 65 degrees... weird... why does that work?)

3) Looked into manufacturers for dry pigments that supply some of the paint company's (Cadmium Yellow Medium = Yellow Green shade... just one example but looked at several before I started to see a pattern. The shade is equal to the hue shift)

4) revisited handprint and hue shift and started to see a pattern that coincided to what I was seeing on pigment manufacturers web sites and to what I was seeing when I mixed with white or neutral grey

5) Rethinking why my swatches of yellow mixed with grey go to olive (is desaturated yellow brown or olive... found out it was olive... with still everything in the back of my mind regarding pigments and hue shift I test the theory with one of my colors that looks more yellow orange to me... add white and lo and behold it shifts almost to mid yellow... that explains why it went olive "maybe". I'm still not convinced at this point but getting more curious.

6) I go and find the thread that Bill posted regarding Blue and Yellow make grey and start this thread. (Sid is able to get olive and extend it to grey)

7) I already have all my paints mixed with white in several steps and the tubes are laid out in the color of the spectrum. As I'm mixing 1/2 and 1/2 of each of my paints ex. started with a yellow light mixed with each color in the order of the spectrum I'm noticing where I would expect the mixing complement to be is where I'm seeing drastic changes in chroma compared to the adjacent colors. (With yellow saw some desaturation with dioxazine violet but knew from past explorations with mixing those two that it was just shifting toward brown. I could tell by the way the brown was changing. Suspected before that the color I'm looking for may fall between Ultramarine Blue and Dioxazine Violet)

So that's the way my mind works.

My thoughts on a color wheel. I think I need something I can measure and for me HSB works. If you can't measure then you can't compare. You need something to compare and measure by. I think with any experiment you start with the known and then go to the unknown.

"Complementary colors are defined to mix to grey, either additively or subtractively, and many color models place complements opposite each other in a color wheel."

From here: http://en.wikipedia.org/wiki/Grey

I see a color wheel as the spectrum wrapped around a circle and if the spectrum is spaced correctly then the result should be a color wheel that can be used to predict mixing complements. Wrong spacing = wrong mixing complements.

The way I think is that nothing exists in a vacuum or bubble. Everything can be explained with science. Yes I don't know the formulas and a lot of the jargon but deductive reasoning can sometimes work wonders.

I realize that at this point nothing is conclusive and I'm the last person to say yes this is the way it is but "Nothing ventured nothing gained".

I think poking around looking for complements without some sort of direction is akin to looking for a needle in a dark room. Color is much more complicated then those 12 section color wheels you can purchase in stores. I think we already went through something similar. You say yellow and everyone has an idea of yellow already preformed in their mind and it will be different for everyone and same thing goes for any other color you name and we know from this thread that the same thing goes for Blue.

I'm posting the 24 section HSB color wheel with the labels for the color names and hue angles. (keep in mind these labels are not me making up the names but going by a color wheel I have that is based on HSB) I posted this in another thread so will just use the image tags as requested earlier in this thread.

http://www.wetcanvas.com/Community/images/17-Apr-2011/84697-Labeled_Color_Wheel.jpg

Goldeelocks
04-18-2011, 06:03 PM
Grey + yellow goes to green because there's blue pigment in grey. Especially in payne's. The only greys I haven't tested yellow with is greys made from true black pigments like carbon or ivory. I would love for someone to do that test with yellow + ivory / carbon + tit white and show what happens until I get my tube.

The problems I have aren't the complements anymore, it's titanium whites pulling the chroma down too much or going too much off the hue. Finding the complements has been fairly easy compared to those.

sidbledsoe
04-18-2011, 06:47 PM
Grey + yellow goes to green because there's blue pigment in grey. Especially in payne's. The only greys I haven't tested yellow with is greys made from true black pigments like carbon or ivory. I would love for someone to do that test with yellow + ivory / carbon + tit white and show what happens until I get my tube.

The problems I have aren't the complements anymore, it's titanium whites pulling the chroma down too much or going too much off the hue. Finding the complements has been fairly easy compared to those.
I thought that it was my blacks (http://www.wetcanvas.com/forums/showpost.php?p=9546811&postcount=7)making for such greens but I think much of it is coming from the white too. Poor yellow has everything making it go green in those mixes, reduced saturation, cold bluish whites, and cool bluish blacks.

couturej
04-18-2011, 06:51 PM
Hi Goldeelocks! Actually the greys I was using was Golden Acrylics Neutral Grey. I can mix acrylics with my Water Mixable Oils. Can't say I'd do it on a painting. I'd wait about 10 years and see how other peoples painting that mixed Water Mixable Oils with Acrylics fared before I ventured into that realm.

The results are not Green, they're olive green.

I realize that Ivory and especially Payne's Grey have blue in them so pull the color toward Green. Keep in mind though if you don't see the difference between Olive Green (desaturated yellow) and other greens it's very difficult to determine whether you're just getting a desaturated yellow or it's actually pulling the color toward green.

couturej
04-18-2011, 07:18 PM
Hi Sid!

The Rembrandt ivory black when tinted is at 215 degrees HSB, resulting green is at 75 degrees

The W&N Ivory Black when tinted is at 215 degrees HSB, resulting green is at 70 degrees.

The Holbein Ivory Black when tinted is at 215 degrees HSB, resulting green is at 70 degrees.

Interesting all the Ivory Blacks are at the same Hue Angle and the only difference is the value and chroma. As I was saying a way to measure.

If people want to post swatches with some colors with white added I can plot them on a blank color wheel. If we have enough of your colors plotted if their is anything to what I'm saying it should result in determining the mixing complement.

Mix your color with white in several steps and if you can make your swatch an even mix like Bill's that would give better results.

If you want to try in Acrylics then no problem I'll do separate wheels for different mediums and manufacturers. I will need the pigments code or codes (doesn't matter if they're a hue), the manufacturer and color name.

My daughter gave me her scanner so I'll start doing mine soon.

Einion
04-19-2011, 03:55 AM
I see a color wheel as the spectrum wrapped around a circle...
Just to reiterate something I refer to above: no magenta in the spectrum. A full wheel shows all possible hues but a spectrum doesn't.

...and if the spectrum is spaced correctly then the result should be a color wheel that can be used to predict mixing complements. Wrong spacing = wrong mixing complements.
But masstone colour doesn't predict complementary mixing behaviour... I assure you, it's not getting any less true each time I say it but I can keep typing it if it'll help :)

Surely all you need is one example of a pair of paints that aren't opposite in hue that mix to neutral to shoot the theory down once and for all?

I'm sure you can think of one yourself if you go through all the common complements you know and then visualise or manually plot the positions of the paints on the wheel above. There are literally dozens of examples to pick from.

I think poking around looking for complements without some sort of direction is akin to looking for a needle in a dark room.
It can be like this. The only real shortcut is to pick safe/safer bets in the pairings, although it's still not a perfect solution.

It's just occurred to me that I haven't mentioned this recently - even if you created a 24-colour palette where each of your dozen pairs of paints mixed perfect neutrals you wouldn't have solved the practical issue for painting, since a lot of the colours you'd need to dull down as you paint will already be mixtures, hence won't have a single-paint complement ready to pick up and use...


Grey + yellow goes to green because there's blue pigment in grey.
This also happens with neutral greys made without any blue (commercial neutral greys are white, black and some umber commonly).

The greening of yellows in mixtures with paints that are neutral (white, black, neutral grey) is one of those oddball pigment-interaction things that there's no layman's explanation for. We just need to keep in mind that it's the norm.

Einion

couturej
04-19-2011, 07:31 AM
Maybe color space is a better word then spectrum.

No masstone can not predict mixing complements. I never said it did. Actually I've always been saying that where the hue shifts to when white is added would be its position on the wheel. The only thing I'm interested in is the hue angle and hue shift. I know that Ultramarine Blue is about at 225 when white is added but not sure about Burnt Umber yet but it looks like a fair bet that it would be about 45 so yes I'm seeing a correlation between the wheel and known mixing complements. But I've never taken hue shift into consideration so can't say at this point what the results will be.

I don't mind having hues (paint mixtures) on my palette as long as they get the job done. I wouldn't want a palette that consists of 24 colours.

I'm looking for a functional palette. One that will have everything I need to get the results I want. Colors not too close together but not too far that I'm not able get saturated colors. Mixing complements are important and would work out great because they would be part of that spacing because they lie directly across from the colors I select. Any additional colors that may produce mixing complements would be on standby maybe for certain paintings.

Off to do some scanning.

couturej
04-19-2011, 11:30 AM
The scanner didn't work out. This scanner distorts the colors so much that it's useless. Taking photographs works better but distorts the lighter colors so I think a paint pile mixed loosely with white works better and just taking an average on that pile or using a color mixed with white that the final mix isn't too light. Sorry I know it's not an exact science doing it this way but it would be something that most people have access to.

couturej
04-19-2011, 01:16 PM
I'm sorry I still can't get accurate colors some look good others are not close. I guess this method can't work unless we could get accurate colors with a scanner or a camera. I know I can't. I did learn a lot about hue shift and color in general. Thank you everyone for your help!

wingedbear
04-19-2011, 02:07 PM
if you will forgive, i think you're going about this the wrong way - working from a perfect, idealized theory of color and trying to make pigments fit into that theory - trying to make the facts fit the theory as it were.

what is often overlooked in these talks about color theories and color wheels is that these are just concepts to help the artist mentally organize their thoughts on color, and it is generally assumed that these color wheels are a tool for the artist, a means of creating finished paintings and not an end in themselves.

you might do well to choose a palette of colors and experiment with varied mixes rather than looking for the perfectly crafted color wheel. and unless you're working with digital art, computer colorscapes are rather poor substitutes for using actual pigments.

i, myself use a palette of only 5 colors, mimicking the CMYK printing process that has successfully printed millions and millions of colorful books, including many, if not all, color theory art books.

granted, there are some colors i cannot achieve, but, for example, if i'm trying to paint a vivid orange, i can achieve that by shifting the hue, value and particularly the chroma of the other colors in my painting to make that orange stand out.

in the end, while trying to figure out the various merits of RGB over HSB, or arguing the benefits of CMY vs RYB, while any of these questions of color may prove to be a fun intellectual exercises, it's the artists understanding of the pigments she uses that's of import, and the most perfect selection of colors on the palette does not make for a beautiful painting.

the true nature of color lies not in the most carefully selected palette of pigments, but in the artist's ability to use whatever pigments she may have to interpret, not copy, and paint what she perceives, regardless of the limitations of those pigments.

the true nature of color lies not in the pigment, but in the artist.

Einion
04-19-2011, 02:14 PM
No masstone can not predict mixing complements. I never said it did.
It did seem to be implied by the bit I quoted, but no biggie :thumbsup:

The only thing I'm interested in is the hue angle and hue shift.
One thing to bear in mind is that there isn't a hue angle for the undercolour or tint since this changes dynamically with the application thickness / wash strength or the amount of white used (as well as which white is used).

I know that Ultramarine Blue is about at 225 when white is added but not sure about Burnt Umber yet but it looks like a fair bet that it would be about 45 so yes I'm seeing a correlation between the wheel and known mixing complements. But I've never taken hue shift into consideration so can't say at this point what the results will be.
There will be correlation for certain colours, just as there is if the masstone were used. But there's no broad pattern to find I can assure you; like I say on the previous page nothing will predict complementary mixing behaviour. Would you like proof or would you prefer to keep on looking?

I wouldn't want a palette that consists of 24 colours.
Ditto. What I was getting at is that even with a palette that large (about double the size that many painters use) it doesn't completely solve the problem of neutralisation.

Although complementary pairs are very nice when they're available a lot of the time you won't be able to use 'em anyway, so for this reason I think it's probably best to put aside the search for them during palette construction as a main goal, learning early the need to rely on other means - mixed complements / split complements, or neutral greys, or any of the three depending on circumstance.

Einion

couturej
04-19-2011, 04:13 PM
I do like using complementary pairs for the simple reason that their is no hue shift. If I mix a red with its complement I can use it either for something in that hue angle or pull it toward lets say orange with a yellow if I need a desaturated orange. At least I know that original desaturated red I used didn't shift. The level of desaturation you can get isn't like any near complement can achieve.

I would love to see an example of why it doesn't work. I'm not interested in chasing my tail. If it doesn't work then time to move on. I think I still prefer HSB compared to anything else available. But I don't want to waste time trying to fit a square peg into a round hole.

couturej
04-19-2011, 04:22 PM
Thank you wingedbear for your thoughts on the process and see what you mean! HSB has never been considered as beneficial to artists to be used as a color space. I see a lot on CMYK and RYB but never HSB. I think if their was more info along these lines it would have been apparent to me whether this space has any benefits compare to the others.

Einion
04-20-2011, 02:48 AM
I do like using complementary pairs for the simple reason that their is no hue shift.
Ditto.

If I mix a red with its complement I can use it either for something in that hue angle or pull it toward lets say orange with a yellow if I need a desaturated orange. At least I know that original desaturated red I used didn't shift.
http://www.wetcanvas.com/Community/images/20-Aug-2003/3842-thumbsup.gif

The level of desaturation you can get isn't like any near complement can achieve.
You can get equal results with near-complements within certain limits, and certainly using split-complement mixing (which is basically what's being used when you mix a pair of paints that aren't perfect complements and then tweak the result back to the original hue).

I would love to see an example of why it doesn't work.
Okay, as I say above, Bruce MacEvoy's work in this area proves unequivocally what a lot of anecdotal evidence already suggested.

Painters are quite familiar with mixtures of Cobalt Blue and Burnt Umber or Cobalt Blue and Burnt Sienna both mixing to good or perfect neutrals. This alone is enough to illustrate the point, it works even better if we look at it from the opposite direction - many people realised that Burnt Sienna will often work as a complement for both Cobalt Blue and French Ultramarine, but painters who have experimented more might also know that it can work well with Prussian Blue too, redder shades of Phthalo Blue and even Cerulean Blue.

Taken together this is absolute proof that masstone, undercolour, the difference in hue between them (if any) as well as pigment type tell us nothing in advance about the complementary mixing behaviour we can expect from our paints.

From Handprint's page, colour 13, Color Wheels:
The first surprise is that the mixing complement will often be very different from the visual complement...
The second surprise is that substance uncertainty creates insurmountable problems for subtractive color mixing. The issue is not just that the "mix to gray" results are dependent on the media used we could just accept watercolor paints as our standard. The problem is that when we use paints, the "mix to gray" results are not consistent across hues different colored paints will happily mix to gray with the same primary color! (The problem is described and illustrated in the section on painting in neutrals.) So we often end up with several different hues as mixing complements for a single paint which means the different hues must all be located at the same point on the color wheel!

From Handprint's page, colour 16, An Artist's Color Wheel:
If we compare the results of subtractive paint mixing, additive light mixing and perceptually defined color models, we find that visual and mixing complements are almost never the same. The visual complement of ultramarine blue is a yellowish green, but the mixing complement of ultramarine blue is a dull deep yellow; mixing ultramarine blue and greenish yellow paints produces a dark bluish green. The visual complement of phthalo green is close to quinacridone rose, but the mixing complement of phthalo green is a middle red; mixing phthalo green and quinacridone rose paints gives a muted dark violet, not a neutral tone. And so on. (These color circle comparisons clarify the different hue relationships that appear in visual and mixing color circles.)

From the page Watercolor Mixing Complements:
Why have I presented the results as a big list rather than as a color wheel? Because it is not possible to make a consistent mixing color wheel...
Ideal complements for French Ultramarine:
Raw Umber, Quinacridone Orange, Benzimidazolone Orange (!)

Ideal complements for Phthalo Blue GS:
Venetian Red, Gold Ochre, Perinone Orange (!!)

Ideal complement for Cobalt Teal Blue include:
Pyrrole Orange, Pyrrole Scarlet, Quinacridone Maroon, Cadmium Red Deep (!!!)

From Painting In Neutrals:
The warm hue mixing complements for most blue or green pigments cover a large hue span. This means that paint mixing cannot identify unique complementary colors for any blue or green.
Some of the mixing complements contradict color wheel logic. Compare, for example, the mixing complements of prussian blue (PB27) and phthalo turquoise (PB16). The perfect mixing complement for prussian blue is venetian red (PR101), and the best complement for phthalo turquoise is perinone orange (PO43). But phthalo turquoise is much closer to green than prussian blue, and perinone orange is also closer to green than venetian red: as the blue hue moves counterclockwise around the color wheel, the complementary color moves clockwise! You just can't get that to make sense on a color wheel.

The truth is that you can't show mixing complements as a color wheel. No matter how you tug and pluck, you can't unravel the mess created by the substance uncertainty of real paints. The only way to cut through the knot is to work with the color relationships defined by additive color mixing... and learn the mixing complementary relationships by rote for the paints actually on your palette. Paints you don't use you don't have to know about.

Also from that last page:
With few exceptions (yellow ochre and other dull deep yellows), yellow has no role as a mixing complement. This is because yellow reflects both "red" and "green" light, and it is extremely difficult to get a reddish blue or blue violet color that can exactly cancel out both parts of the spectrum equally.
Despite this, a mixed complement can work for yellows as we've seen in the example posted above. This is because paints are not about 'their colour', mixtures even more so, and a two-colour blend of the same basic colour as a commercial paint doesn't mix the same way as most of us know, and this of course extends into its complementary behaviour.

...

HSB has never been considered as beneficial to artists to be used as a color space. I see a lot on CMYK and RYB but never HSB. I think if their was more info along these lines it would have been apparent to me whether this space has any benefits compare to the others.
Just a bit on this, there's little talk about CMYK for painters. Recommendations based around CMY primaries aren't really about four-colour process and often are as much, or more, about the full wheel that includes them spaced correctly with RG and B in between.

As I think was touched on previously in this thread, the colour wheel I posted has basically the same structure (spacing) as HSB but it was made simply by positioning RGB and CMY and blending between them; they're essentially just two ways of defining the same thing. The computational stuff about how colours are defined aside, many people use HSB for the hue angles because they're easily accessible in Photoshop or similar software. Every time I give a hue angle I'm using HSB definitions and I use the saturation percentages frequently too.

Einion

couturej
04-20-2011, 08:57 AM
Thank you so much Einion for all the info!

One big thing that I'm having issues with is that Water Mixable Oils and the pigments are very different as far as the colors and available colors (Holbein Duo Aqua). Something as reliable as Ultramarine Blue looks different and behaves differently then what I'm use to. Same thing goes for the Burnt Umber, Burnt Sienna, Yellow Ochre etc.. Just weird. Oh well I'll figure it out.

I think I'll pick 3 colors evenly spaced CMY. Focus on finding their complements and go from their. I know I can't get that Ultramarine Violet in WMO so I'll settle for something close. I was debating on a Ultramarine Violet in the brand I mentioned before but it's a mixture of 3 pigments...yikes and it's a brand that isn't well known. (I don't mind mixes but 3 is a little much) Royal Talens Cobra (BTW their customer service is the best) had a contender of Violet Blue but they actually tried the mix for me and it's not a winner. It pulls toward brown. I did see people using Dioxazine Violet plus Ultramarine Blue as a mix to use as a complement to yellow so might give that a try and I'll take a closer look at the mix again with Dioxazine Violet with white added with my different yellows and see how that works out.

Books are non existent when it comes to color mixing in Water Mixable Oils but for all other mediums you can find info in books, online etc.

Einion
04-20-2011, 10:10 AM
Thank you so much Einion for all the info!

One big thing that I'm having issues with is that Water Mixable Oils and the pigments are very different as far as the colors and available colors (Holbein Duo Aqua). Something as reliable as Ultramarine Blue looks different and behaves differently then what I'm use to.
Welcome! Pigments are more variable than consistent anyway, but French Ultramarine is surprisingly varied compared to how we think of the colour, with lighter and "greener" versions (closer to cyan), deeper and more-violet versions. The thing that I found especially prominent with PB29 is how much difference there can be between the tints of versions that look pretty similar in masstone - some are waaay duller than others, and it's not related to the pricetag.

I did see people using Dioxazine Violet plus Ultramarine Blue as a mix to use as a complement to yellow so might give that a try and I'll take a closer look at the mix again with Dioxazine Violet with white added with my different yellows and see how that works out.
If you space the pigments further apart you'll generally have an easier time with neutralisation, since the mixture of the two is already lower in chroma - if you visualise a straight line drawn between two paints, it's closer to the centre of the circle the wider you space them.

Alternatively, just use grey and compensate for any shift in hue my mixing in a smidge of red! At the end of the day the result is what matters, not the route.

Books are non existent when it comes to color mixing in Water Mixable Oils but for all other mediums you can find info in books, online etc.
No need as there should be no colour-mixing issues specific to the medium. It'll come down nearly entirely to being about the pigments and the amount of them in the binder, just as it is between two different brands of regular oil paint.

Einion

couturej
04-20-2011, 03:48 PM
Thank you for the additional tips!

What the books offer is the brand name and color so you know that if your using the same brand and color you're pretty well guaranteed it will work.

Just a few more questions:

You stated that Cadmiums tend to have less of a hue shift but not all. Is their any other colors or group of colors in general that are notorious for having less of a tendency to shift in hue?

Regarding white you mentioned something about one of the factors is which white you use. Which white is better when it comes to trying to control hue shift?

Einion
04-21-2011, 09:14 AM
What the books offer is the brand name and color so you know that if your using the same brand and color you're pretty well guaranteed it will work.
Oh I see what you mean. For a colour-mixing book generally there's the issue of the limitations in reproduction accuracy, but this wouldn't apply to listed complementary pairs (if they included them).

You stated that Cadmiums tend to have less of a hue shift but not all. Is their any other colors or group of colors in general that are notorious for having less of a tendency to shift in hue?
In terms of pigment families the cadmiums may be the prime example. Off the top of my head I can't think of a single other pigment type that has a tendency to shift in hue less.

Regarding white you mentioned something about one of the factors is which white you use. Which white is better when it comes to trying to control hue shift?
I think it'll depend more on the exact example rather than whether you're using titanium, zinc, a lead white or a mixed-pigment white. In terms of colour shift, Titanium White tends to be worst of course because it's the strongest, but I don't know that I've ever seen anything comparing hue shift.

Einion

couturej
04-22-2011, 05:15 PM
Thank you Einion!

couturej
04-24-2011, 07:10 AM
HSB is wrong in a lot of ways. The desaturated colors in HSB are wrong and Munsell is right on the way that colors look when desaturated. Some coincide with HSB but Munsell is more accurate. Sorry about all the fuss with HSB. It did help me nail down some complementary pairs if you're going by the tinted colors but all in all it's a system that is off. I'm getting some neutral grays and some straight line mixing to gray in one part of the wheel but the results can not be compared to the HSB desaturated colors and I need to use Munsell to make sure I'm getting a straight line to gray.

I'll be able to plot my results using Munsell and reorganize the Munsell wheel so the mixing complements fall across from each other. Munsell is not as widely available so that's unfortunate. At least their is a program here that is fairly inexpensive: http://livingstonmanor.net/munsell2/index.htm

I still need to use the tinted color to determine the hue shift and place it on the wheel based on that. So using Munsell is a good way to go but the spacing around the wheel just needs some adjustments.

Just thought I''d give a heads up in case someone else is using HSB and relying on it to see how a desaturated colors look.

couturej
04-27-2011, 08:51 AM
HSB looks fairly accurate to use. It was the RGB values that Pantone provided for the Formula Guide that are wrong.

couturej
05-07-2011, 10:08 AM
Finished testing out using the HSB color space to determine complements and results are here: http://www.wetcanvas.com/forums/showthread.php?t=929396