PDA

View Full Version : Color Spill


Randize
12-28-2010, 01:40 AM
Hi,

Any body knows about the technique of putting "random" color that artists usually use in oil painting? I find it beautiful and hope to learn more about it.

The problem is I don't know what the technique is called and where to begin.

Below are some references I got from art renewal.

http://www.artrenewal.org/pages/viewartwork.php?filename=/artwork/2009-2010%20Salon/TheBallad-large.jpg
http://www.artrenewal.org/pages/viewartwork.php?filename=/artwork/2009-2010%20Salon/NorthMeridan-large.jpg
http://www.artrenewal.org/pages/viewartwork.php?filename=/artwork/2009-2010%20Salon/BarcelonaSideStreet-large.jpg

Lunatique
12-29-2010, 07:14 AM
It's not really random, because you do have to take into consideration how the various colors either contrast or harmonize against each other, and how they form the overall color palette of your painting as a whole. There's definitely a lot of thought that goes into these seemingly random color embellishments. For example, if you find that the image is overall a bit too cool in color temperature, yet the inherent local colors and the light situation does not logically allow warmer colors (such as an overcast day, a gray concrete jungle, and people wearing blue and gray and black, and not a warm color in sight). In situations like that, you can embellish some areas with warmer colors so the image isn't too cold and sterile. This is in very simplistic terms, but you can do this kind of harmonizing and balancing not just considering the entire image, but various areas of your image. For example, one particular corner of your image is too warm and sticks out too much, you could then embellish it a bit with cooler colors just to help it fit in easier in the painting. Of course, it's not just for balancing and harmonizing--you can use these embellishments to make your image appear more dynamic and lively. You should check out Pino's paintings--he's a master of that approach.

Where and how you place these color embellishments is a matter of personal taste and experience--you really have to experiment and try different approaches to find one that works for you.

Quadart
12-30-2010, 11:52 AM
Randy, I think you may be referring to color mingling, and as Robert mentioned, it is not random, though it may be spontaneous there is purpose in the color choices. The other important thing to remember when mingling in colors that are a compliment, off compliment or even a purer hue of the main surrounding color is to be aware of their grayscale value. The color needs to work with, not against, the grayscale composition of the work. You will notice that a lot of these mingled colors share similar, if not the same, grayscale value of the surrounding color in many cases. This can create vibrant, shimmering color field effects.
Color mingling works with loose impressionistic painting styles. I assume this is what you mean by color spill, opposed to tightly render paintings, which don’t normally use bold mingled color accents.

Robert Heindel (oil) & Nathan Fowkes (watercolor)
http://www.decards.co.uk/images/art_cards/large/rh011.jpg http://4.bp.blogspot.com/_pbZw476tEsE/Re42GOETZrI/AAAAAAAAABc/FLeAHB-Qg3s/s400/blue-flowers-low.jpg

Some good digitally and traditionally painted concept work examples showing color mingling can be found in Nathan Fowkes' blog:
http://nathanfowkes.blogspot.com/

Lunatique
12-30-2010, 02:20 PM
That's a good point about matching the values. This is a complicated subject though, since it seems no one can agree on just what the inherent values various colors are. Digitally, when converting color to B/W, different algorithms will give you different results, which means a set of colors in one algorithm might convert to all the same value, while another algorithm will make them look different. You can google this topic--especially on how Photoshop's B/W conversion--it's really a can of worms.

While we can sort of visually gauge the values of colors, such as by squinting our eyes and try to homogenize them visually so estimate their values, it's not an exact science, and also, because of the anatomy of our eyes, we are more sensitive to certain colors than others, which further skews our ability to match values with colors. I think the most important thing is to simply keep your eyes on the overall big picture and make sure that every decision you make, you are making them to benefit the overall image, not just that one section you're working on. Whenever you find yourself working too closely on one area, always spend a moment and fly back up to the 10,000 feet macro view and survey the entire image as a whole and see whether the area you were working works well with the rest of the image.

Quadart
12-30-2010, 06:55 PM
Yes, I too have spent my fair share of hours scouring the net on the topic of color in the digital age, along with a lot of time spent in books on the subject prior to the digital age.

I don’t think determining the inherent value of a color is an unresolved issue. The inherent grayscale value of a color or a pure hue is always consistent in relation to the values of other colors or pure hues in the same image. The values of colors in an image change as a whole when the white point value of a scene changes, but this does not change there value ratios to one another. In the same way that chroma has no intensity limit. Red at 100% chroma can appear very bright or virtually black.

If you are working in PhotoShop, an easy way to match the grayscale value of two different colors is to use the grayscale slider in your color window, as I’ve done with the 3 small squares in each of the 6 middle color swatches. The reliable way to convert a color image to it’s grayscale equivalent is to simply convert it to grayscale. Don’t use the Desaturate option. The result is another topic altogether—which I don’t mind getting into either.
Back in the ‘old days’, I used grayscales painted in the medium I was using to help keep colors in check. And speaking of stepping back to see the big picture, I had a reducing glass to zoom out with. :argh:

PS has no trouble pinning the gray equivalent of a color. Take the linked image into PS and convert it to grayscale mode. You should have 18 gray squares as a result.

http://i219.photobucket.com/albums/cc156/bill_77/colorx690.jpg

*If your matching values in RGB space, keep it RGB. Converting to CMYK for offset printing will cause some mismatches, especially noticeable with purer hues.

--Ultimately a good impressionistic painter usually has a heightened sensitivity and sensibility when it comes to color, and shouldn't need to rely, heavily, on devices to cue color value, let alone shape dynamics, proportions etc. A good painter handles colors and marks like a poet handles words, with little overcritical or over-analytical fanfare while working.

Lunatique
12-31-2010, 01:08 AM
I don’t think determining the inherent value of a color is an unresolved issue.
...
The result is another topic altogether—which I don’t mind getting into either.
...
PS has no trouble pinning the gray equivalent of a color. Take the linked image into PS and convert it to grayscale mode. You should have 18 gray squares as a result.
...
--Ultimately a good impressionistic painter usually has a heightened sensitivity and sensibility when it comes to color, and shouldn't need to rely, heavily, on devices to cue color value, let alone shape dynamics, proportions etc. A good painter handles colors and marks like a poet handles words, with little overcritical or over-analytical fanfare while working.

I'd love to read what you have to say about the topic of using Desaturation. This is something I have discussed in-depth with students--how Photoshop treats Desaturation and Grayscale differently.

While mathematically, a machine and software can determine the inherent value of a color a lot more accurately than humans could by sight, the problem is that we don't paint by mathematics--we paint by vision. While artists could become sensitive to how colors translate to values, it is a skill that takes a long time to develop and require a lot of experience. It's way too easy to have our vision fooled by optical illusions because we simply cannot control the way our brain interprets visual data (such as the same color can appear distinctly yellow or blue, depending on the context), so the more trial and error and knowledge we devote to this topic, the easier it will be for us to not be fooled as easily, or at least learn to not take everything too literally and try ways to verify what we are seeing. It also allows us to use these optical effects to our advantage.

I think experience artists do think critically about color--they just do it in a way where they don't break a sweat about it, because they've accumulated enough experience and knowledge to be able to make the decision in a split second, whereas less experienced artists will take longer to make the same decision, or unable to make the decision at all because they lack the knowledge. But when experienced artists make such decisions, it's not really simply going by instinct or not thinking about it critically--they just do the thinking much faster so that it appears effortless--sort of like how a faster computer will compute a problem in a fraction of the time of a slow computer, while the slow computer might even lock up and freeze due to the difficulty of the problem. I would say it's similar to how advanced musicians can compose/arrange/perform in such a way that it almost feels like they are operating solely on instinct, but behind that effortlessness is years and years of studying and training, and if we really breakdown each move and decision they make, there's a whole wealth of information behind them. So much of it becomes muscle memory or "gut instinct," but there's always a decision made at some point--one that they'll be explain if push comes to shove.

Randize
12-31-2010, 04:22 AM
Thank you for the explanation both masters. Hearing there's no exact explanation in science about this subject makes it hard to master the technique. I did a quick google search about color mingling but non that I find helpful.

Experimenting this on digital is quite hard too as I have tried, I came from non traditional background by the way. As oppose to traditional medium, digital medium has millions of permutation in color combination which only makes the palette more harder to control. I tend to get muddy color as I attempting to mingle colors without proper direction.

But I am thankful with your help I can narrow down my choice of colors in mingling by:
- Complementary colors
- Temperature
- Without disrupting the value

I'll get back to you with this study. Thank you, Robert and Bill.

halen
12-31-2010, 09:55 AM
Thank you. This is interesting to follow. Looking forward for those studies.

Quadart
12-31-2010, 03:10 PM
I'd love to read what you have to say about the topic of using Desaturation. This is something I have discussed in-depth with students--how Photoshop treats Desaturation and Grayscale differently.
If you want a full explanation, you’ll have to take my class. ;)
Kidding aside--I’d also like to here what you have to say on the matter.

While mathematically, a machine and software can determine the inherent value of a color a lot more accurately than humans could by sight
This is in some contrast to what you stated earlier "…it seems no one can agree on just what the inherent values various colors are. Digitally, when converting color to B/W, different algorithms will give you different results, which means a set of colors in one algorithm might convert to all the same value, while another algorithm will make them look different."

The algorithm result differences you describe are based on the fact that the systems are not striving to achieve the same goal. Desaturation and grayscale, for example, do two different things. One produces the grayscale component of a color by desaturation (grayscale mode) and the other averages brightness levels of different hues to a gray value, e.g. hues of 100% brightness are all represented by the same gray value.

I think experience artists do think critically about color--
I did not state, let alone imply anything to the contrary. Sorry for the confusion. I should write longer posts to cover all bases.
**It should go without saying that, yes, an experienced/accomplished painter knows his craft on many levels, from the purely technical, and to some degree, the relevant psychological aspects of the science and psychology of perception involved. We know it takes quite a bit of serious thinking, problem solving and trail-and-error work over quite a bit of time to master one’s art, to be able to employ the accumulated knowledge and working experience in a spontaneous, tacit and seemingly effortless manner, especially in the eye of the non-artist observer. I most certainly was not implying the contrary. I guess you may have missed the “while working” part of the last bit you quoted.

It's way too easy to have our vision fooled by optical illusions because we simply cannot control the way our brain interprets visual data
Good site on the subject of color based illusions. You've probably seen it.

http://www.lottolab.org/index.asp

The blue yellow thing you mention is in here:

http://www.lottolab.org/articles/illusionsoflight.asp

Lunatique
01-01-2011, 07:44 AM
Yeah, I've seen those. I first learned about the blue and yellow illusion about ten years ago at Sijun, and I think it was none other than Craig Mullins who was teaching us young pup about all this stuff--color relativity, values vs. colors, lighting, shapes vs. textures...etc. Fond memories.

What I meant about machines being more reliable is that there's no human error involved--it's always consistent. Some of us wake up and can't even see clearly until our eyes have totally woken up too. There are days when it's just a blurry mess for the first 10 minutes or so. Then there's the ambient lighting in the room where you are working, or whether you've been exposed to strong lighting/colors just prior to making color decisions, or even what you've eaten/drank might alter your perception...etc. And of course, there are those crazy optical illusions that will totally influence your decision making ability. With machines and software, once you agree on an algorithm and decide you can trust it, it'll always perform the same. The trick is to first research enough to know which conversion algorithm you agree with the most.

The parts of my workshop where we discuss all this stuff in detail mainly stems from my advice to students to keep an eye on the overall dynamic range of their values while working on a color image, and never to mistaken a change of hue/saturation as a proper substitute for value change. The other aspect is if they ever have their works reproduced in B/W (which is something I hate about some art books), if they weren't paying attention to the dynamic range of the values, their colored work when converted to B/W can end up as a rude surprise (I mocked up a hilarious example of a crude painting where the values are so similar, but distinctly shows a house, water, sky..etc all with saturated colors, but when converted to B/W, is almost unreadable). I advise them to keep a adjustment layer on top of their layers so they can check their values in B/W whenever they want, and the easiest/fast way is to put a desaturate adjustment layer on top, but I also tell them they should use the grayscale too just to see how an alternate algorithm converts their work.

Another thing we discuss related to this issue is value range and why colorizing a B/W piece is much more complicated than they think it is. We get into saturation vs. chroma, and what the ideal situations is in order for chroma to create the illuison of optimal intensity, and so on.

All of this stuff are inter-related, and very fascinating--especially for students who are completely unaware of such issues. The end goal, is always to give the students an arsenal of tools that ranges from technique, knowledge, creative thinking, and critical judgment so that they can scrutinize their own work and try not to make egregious mistakes that they could've avoided as they create, and also so they could effectively deconstruct the work of others and really understand what makes them great and how the students can absorb and adapt the qualities they admire into their own works.

dbclemons
01-08-2011, 09:48 PM
...The problem is I don't know what the technique is called and where to begin...

I would identify this as a "broken color" technique. The intent is to acheive the optical mixing of color in loose "unfnished" patches that has it's roots in the 19th century. You can find the earliest examples in the work of Constable, Courbet, or Turner; although, later artists like Manet and the Impressionists are the ones who get the most credit for it.

Quadart
01-11-2011, 12:20 PM
I would identify this as a "broken color" technique.

I forgot about that description. :thumbsup:

The optical mixing of broken colors (colors that are broken into their more primary components) is one specific aspect of the various results you can achieve by loosely mingling or juxtaposing small daubs of different colors together. Using broken colors to achieve a secondary or tertiary color is used more extensively as a traditional painting technique to try to avoid the ‘muddying’ effect (caused by subtractive color mixing and unfavorable physical/chemical properties unique to specific pigments) that occurs when thoroughly mixing pigments together. That problem can be avoided when working digitally in the additive color mixing environment of RGB color space, for example. The effect is still enriching.

An interesting tidbit about the term ‘broken’ in broken-color was the contradiction of the term I recently found in my copy of Ralph Mayer’s ‘The Artist’s Handbook of Materials and Techniques (p. 163). A broken color is referred to as the inferior secondary/tertiary ‘muddy’ color resulting from the mixing of two or more pure out of the tube or off the pastel stick pigments. Again, this is in contrast to 'breaking' a secondary or tertiary color into it's primary color components to mix optically in the viewers brain.

CGTalk Moderation
01-11-2011, 12:20 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.