Painting Over to Isolate Design

In this post I’d like to explain why I think it is useful to paint on top of master paintings. Hopefully it is something that you will start using as a tool to better understand pictures and design. It’s a bit of a controversial topic from what I have seen (I've had some very negative reactions when I suggest it to people), and I think it stems from people not understanding the point of painting on top and why I do it.

To some it may be sacrilegious to “ruin” paintings like this (which is a bit silly when we are just painting on digital files), or may seem like an act of hubris as though the original painter was not masterful enough. But that’s not what this is. The thing with painting and design is that there are so many layers to it and so many things happening at the same time, so when analyzing things it can be helpful to isolate one aspect to make it easier to see. Much in the same way you may turn a painting into a greyscale version to see the values without the trouble of the hues and saturation, painting on top is used to isolate one aspect of a painting and see exactly what effect it has on the whole and hopefully understand why the original artist made the decisions they did (which you may agree or disagree with).

I also think it is useful because it helps with the problem of putting certain artists on pedestals and treating them as gods which make perfect images every time. It is good to be able to think rationally and trust your own tastes and if you disagree with the choices a master painter made then that’s okay—in fact it is good to have your own tastes and opinions and not blindly treat an image as perfect just because of who painted it. We have to remember that even the greatest artists were humans too and had their flaws and mistakes.

I like to paint over when I see a painting that has a choice in it that I would not have personally made. Either it was a very bold or unusual move, or was a choice that doesn’t coincide with my personal tastes. I don’t normally save these paintovers, but for the sake of this article I went into my files and picked out a couple paintings to use as a demonstration. All I do is remove the decision that the artist made and see how the image would look without it.

I’ll start off with this painting by Howard Pyle, which happens to be one of my favourite paintings of his. I could talk for hours about the various decisions he made in it that I love, but for now I just want to focus on that red shawl. Pyle often liked to use strong saturated reds in his paintings, especially at focal points (I suspect it was his favourite colour). But the strength of that red in this painting is so bold and strong that I would never have the courage to put it in like that. Why did he make it just so strong? In lighting situation no red would ever look like that so it is clearly a decisive move on his part for a very good reason.

So I tried lowering the saturation on it to match the rest of the clothing worn by the other people.  The first thing I noticed was how dark in value the shawl actually is…in fact it is the exact same value as the trees behind her. This is a great example of saturation affecting the perceived value of something (which will be another topic for an article probably).  The painted over image is more cohesive without the red, but it suddenly seems boring as everything falls into each other. By painting the shawl such a strong red Pyle made the girl pop out from the mass of trees and people, and gives her a lot of attention as well as making her appear slightly out of place. Sure, she is standing in a large value grouping, but she doesn’t really belong there. She is out of her element here in the wintery outdoors. The strong red makes her the target and sets up a relationship between her and the wolf (especially since the entire image is very gray, and the only other strong colour note is behind the wolf). Without the red the image is simply a wolf standing there and a group of frightened people. It lacks storytelling and interest. With the red the wolf has a target and the relationship between them becomes clearer as the townspeople stand by her for protection. The other thing to consider is the meaning of red. Why did Pyle choose red instead of any other colour? The red he used is the colour of fresh blood, it is a very tense and bold colour, and hints at the potential violence of the situation. It makes the image have much more impact as a scene. By painting out the red and seeing how much the image begins to fall apart I began to ask myself more questions about it and gain deeper insight into the choices he made. If I had simply looked at the image I would not necessarily have come to the same realizations.

The second image I would like to paint over and discuss is a landscape by Kenyon Cox. I really love his artwork, especially his figures and many of his drawings, but in this particular landscape I always found that front bush to be an odd choice. It’s very large, isolated, and very high contrast. My eye goes right to it, and I am unsure why he would give it so much attention when there is nothing special about the bush and there are more interesting things in the rest of the painting such as the white building on the righthand side (which is the intended focal point). I notice he connected the bush and the house with a line of smaller bushes, and my eye keeps going back and forth along this line, split between the big bush and the house. So of course I try painting it out.

Without the bush we are left with a wide empty expanse of grass.  While I no longer have issue with an out of place bush, the image now is very static and boring. I find my eye is getting caught up in the right hand side of the image and as a whole I am not really looking around the image much. The balance is thrown off slightly, though I notice the heavy treeline and second white building on the left near the horizon does keep it from becoming entirely unbalanced. Still though, it feels like the image is missing something. So rather than remove the bush entirely, I experiment with reducing how attention grabbing it is by lowering the contrast of it. Also by putting this lighter value on the top planes of the bush I feel it has more form now and feels less like a flat splotch on the canvas. It also sets the perspective more clearly as we can see more of the top of this bush than the ones further in the distance where the side of them is more apparent. Now that the bush is sitting more comfortably in the composition, and now that I have seen the dark treeline on the left with the building there, I realize that Cox has crafted a triangular composition between three points on the painting. This is why without that front bush the image became very static and the eye did not wander throughout the whole image.

So while I disagree with how Cox handled the front bush, I now realize the importance of it in the painting. I missed out on the design behind the image before I took the time to paint on top of it.

I hope that this post gives some insight into what a powerful learning tool it can be when you paint on top of images to figure out how they work. Ultimately it is this understanding and ability to break things down that allows us to apply similar concepts in our own artwork. Too many people are spoonfed information and never gain the ability to learn on their own. This type of analyzation is great for discovering your own ideas and theories in imagemaking, and for discovering some of the things others before us have come up with to make successful imagery.

What Is Design?

Again, before going into analyzing things I think it is important to set the groundwork first. In this case, when we are discussing design, it is a good idea to first talk about what we even mean with the word “design”. The problem with the word is it is such a broad term that covers many different things and is used in many different contexts with different meanings. The word has a multitude of uses within painting, but also extends beyond that into other areas of art that aren't visual. Defining it can be as difficult as defining “art”, and then it gets more complicated when we get to discussing good design or bad design. That being said, I will try to set out a loose definition and description of how I see design so we have some basis to understand from what viewpoint we are approaching images.

What is design in painting?

To me, design is PURPOSEFUL arrangement, and is the sum of all the choices made in the process of creating a work of art. In the context of painting this means creating hierarchies (I may write a future article on hierarchies, what they are, how to make them and so on) and controlling where things are placed on the canvas. It is the composition of the contrast, variety and unity of EVERY aspect in art, including but not limited to:

·         Value

·         Hue

·         Saturation

·         Temperature

·         Type of shape

·         Size of shape

·         Texture

·         Detail

·         Materials

·         Form/Flat

·         Space (positive/negative)

·         Edges

·         Planar vs rounded

·         Mass Linking

·         Direction/movement

·         Rhythm

·         Brushwork

·         Line (weight, direction etc)

·         Location on canvas (x, y, and z axes)

·         Pattern

·         Abstraction

·         Transparency vs. Opaqueness

·         Micro and Macro reads on an image

·         Mystery (what is revealed or hidden)

·         Proximity

·         Tangents/Pseudotangents

Design in each of those areas is something which should be considered carefully for maximum effect—what do you emphasize? what do you downplay? How much contrast and variety exists within each aspect will be up to the artist and will change how the image reads and how successful it is, and plays into what defines that artist’s style. Obviously not everything on that list will apply to every image (an ink drawing by its very nature won’t have colour for example), and often certain aspects will be downplayed in order for others to shine (for example the Impressionists would sometimes sacrifice value in order to play up purity of colour). In this article I won’t delve into these aspects but will rather just leave it as a list so people can see just how many things in an image can be designed. Future articles may discuss specifics within a single aspect and how we can design and control it for different purposes.

Ultimately, design is decision, and decisions are art.  We all know of those people who meticulously copy out photos to create photoreal images which are indistinguishable from the photo…they make grids and copy mindlessly as closely as is possible. That is the opposite of design since every decision is taken out of the equation. The only design present is that which the photographer did when taking the photo.

There are of course other uses of the word design, such as functional design in mechanical inventions or clothing or other things, but this blog will mostly be concerned with design in a visual sense within the context of painting and illustration.

What is design for?

We use design to lead the eye where we want it to look, to let it jump over less important areas, to make an image visually appealing, to make an image understandable, and to hold the image together in one unified concept. Design can be in itself an end (art for art’s sake) or, more often, it is a means to an end (such as an illustration which depicts a scene or emotion).

What makes good design?

As I mentioned previously, design must be purposeful. If something happens accidentally it is not good design; if a statement is hesitant or unsure of itself, it is not good design. It must be done for a specific purpose, and the quality of that design depends on how successful it was at achieving that purpose. For any problem in an image there will be multiple solutions—which one is chosen will depend on the situation as well as the individual artist and their style and preferences. Good design solves a problem; GREAT design solves it in an elegant, simple, and novel way. Typically the more economical a solution is, the better it is.

Design is not simply observation, but a conscious analysis and decision about how to show that observation. It is important to remember that design hangs upon reality.

Design vs Painting

Many people mistake good painting for good design. The two are interrelated, but ultimately two separate things. A painting may be technically well done, but if it fails to resolve the problem which it set out on, then it is a failure of design. I suppose this is a rather large topic that deserves its own article. It’s rather difficult too since design is often hard to see until it is pointed out, then it becomes clear as day (note that even when we don’t see the design itself, the effects of the design are still effective). For now I’ll just leave a quote by Harvey Dunn:

“We can learn to paint and draw till the cows come home, and won’t know a thing about pictures. We can actually draw and paint better than most, and still not make pictures. Drawing and painting are merely tools in the hands of him who would make a picture and must be used in making pictures. Most painters are inclined to, when painting from a model, let the model take charge. And they never make a picture. Can I define a picture? No. I can’t very well, because a picture is its own definition. It’s something words can’t quite express.”

I realize that much of this may be confusing or unclear since it is discussing a broad, complex subject in a very abstract way. Future posts will take a look at various paintings and see how they were successful or unsuccessful and why, and hopefully will clear up some of the things I am swiftly breezing over. I particularly am excited to compare different paintings of the same subjects which were handled in different manners.

Finding and Recognizing Good Reproductions of Paintings

This is the first post of what I hope to be many regarding the analyzation and appreciation of master paintings (and drawings). I have over the years saved tens of thousands of paintings to my computer, and I feel it would be a good idea to share some of these images and discuss my ideas regarding them and the things I enjoy about them. I specifically want to focus on the aspect of design—the choices the artists made that make the images successful. I also hope to post up some paintings by artists who are less often talked about.

But before I get into that, I think it is important to go over some considerations that must be taken first regarding the images I will be discussing and sharing. Since these paintings are from all over the world and are not in front of us in real life, but rather are small jpg files on a screen, they obviously will not look the same as they should. As such, it is important to select the images of these paintings that most closely match the original. It is also important to recognize how even the best of jpgs will not replicate the truth of a painting.

Whenever I find an image of a painting I will reverse image search it to see other images online of the same painting in case there is a better version available. Here are some the criteria I personally keep in mind when selecting the best images:

-Is the image of a reasonable resolution? If it is too small then we cannot get a good look at it and fully appreciate all the nuances of the image. A very small image may still provide some crucial information such as the composition or colour relationships, but it will be lacking many other qualities to make it a complete statement and shouldn’t be judged fully. I like to search for the largest available size of an image because I find they tend to be of better quality and more often come from an original source; the largest image also has the most information regarding detail, texture, and brushwork/technique. Google Art Project will often have massive images which are great for analysis (though I do find their colour reproduction is sometimes a bit off). Here's an Armin Hansen in which all we can see is composition really, the image is simply too small for a full analysis.

-Is the image in focus? Many images are blurry—either from when the photo was taken or from file compression artifacts. These are often not useful images in analyzing. This Stanhope Forbes painting is unfortunately useless for edgework and detail analyzation.

-Is the image in the right key? The image may be washed out or too dark overall or have the lights blown out. Compare it to other versions of the painting online to see if it is shifted in one direction or another, and use common sense. Even a high key painting typically has a few dark accents in it to keep it grounded (there are of course many exceptions, but it is a good rule of thumb). Many paintings especially in books will be reproduced a bit too dark.

-Is there glare? Especially in dark passages and on areas with thick, textural paint there may be glare which washes out important information and throws off the balance of the image. Note that some photos of paintings are taken purposely with a raking light to emphasize the texture. Here is a Krachkovsky painting in which the accidental glare interferes with the dark tree masses. And a Sorolla in which the lighting is intended to emphasize the physical paint texture.

-Are the colours accurate? This one can be harder to tell without having seen the original image, but even an untrained eye should be able to tell a fair bit with just a bit of consideration. Check to see if the overall image is too cold, too warm, too grey, too saturated, or shifted towards one particular colour. The lighting a painting was photographed in will dramatically change its appearance (incandescent lighting, fluorescent lighting, natural indirect light, natural direct light etc). The variety of hues in an image is important—in most cases if there is more hue variation then it will be closer to the original colours. This is a more complex issue as well especially with older paintings as other factors come into play such as the permanency of colours, the yellowing of varnishes, and the dirt accumulated on the painting over time. It is common for the original painting to no longer represent how it was when it was first painted. Typically paintings will be slightly warmer and darker if they are old. Here is a Jeff Jones painting that has been shifted heavily towards pink hues. Notice how monochromatic it appears.

-Are there watermarks? Obviously any watermarking will detract from the image. Try to avoid anything with watermarks directly on the image. A rare exception to this is with images from the Museo Del Prado in Spain, in which very high resolution images of good quality are on their website. The watermarking on the images are small and innocuous enough that they are not particularly detracting and are more than made up for by the rest of the quality of the image.

-Is the image the full painting? It is common for an image to be a cropped portion of a painting, so check to see that it is the full thing. If the painting was not photographed straight on, then it is a quick fix in Photoshop to warp the image and correct the perspective on it. 

This image is found when looking for Howard Pyle's paintings, but it is just a microcomposition in a larger image, seen below.

-Has the painting been restored? This is less common, but worth noting. A restored or cleaned painting will often look vastly different than before (and hopefully closer to how it was when originally painted). Some paintings will have both a pre- and post-restoration online, and while we hope that the restored version is better this is not always the case.

-Does it feel “off”? In art there is always that gut reaction we have, and we should trust it. If something feels off even though it seems to satisfy the logical criteria, then the image probably is off in some way. Trust your gut.

-Have you seen the painting in real life, or other paintings by the same artist? This is a best case scenario since you can better judge how it looked in real life. Many of these considerations are dependent on experience gained through seeing paintings in real life. I highly recommend going often to galleries in person to build this experience and also because even the best of photos and scans come nowhere close to the reality of a painting in front of you. Photos miss that certain “something”, that glow that paintings can emanate, they miss the physical textures of the paint and the effect they have as the light bounces off it, they miss the true size of the painting and the impact it has, they miss many subtleties regarding colour and edges. So even when seeing a good reproduction we should keep this in the back of our minds.

The points above are some of the things that go through my mind when I see a painting online. They are some personal criteria that I have come up with and should cover most bases when considering an image, though I may have missed some. As I mentioned before, it is useful to reverse image search any painting you find to try to find a better reproduction or a larger file.

But what if there are only poor reproductions online? If it is a small file or has many jpg artifacts then there is not much we can do. If the problem lies somewhere else, then we can correct some of it in Photoshop. I don’t really recommend playing with an image too much unless you are an artist yourself with some Photoshop experience, but sometimes a bit of adjusting an image can really bring out quite a lot. Surprisingly Auto Tone, Auto Color, and Auto Contrast work well as a starting point in many cases. I usually duplicate the painting on top of itself and play with it so that I can easily compare it to the original as well as lower the opacity of it to get a more subtle correction.  Here is an example with a bad reproduction of a George DeForest Brush painting on the left and a colour corrected version on the right. You can see how it originally was very yellow/orange and saturated from warm artificial lighting. I used Auto Color on it and lowered the saturation, and the resulting image is much more naturalistic in terms of colour and value. Recognize though that the image is still slightly out of focus and the image is a crop of a larger painting and, even though the colour has been corrected to a degree, I would not trust it to be a fully accurate representation of the original.

Lastly I would like to mention a couple ways to find good images. Just by going into specifics of Google Image Searches you can improve your chances of finding good images--I like to search within large image sizes. Some museums will have very high quality images on online collections so it is a good idea to check there. The Metropolitan Museum of Fine Arts, Museo Del Prado, and the Rijksmuseum all are examples of this. Another good place is to check auction houses. Since people are spending large sums of money on artwork, the auctioneers will often post very good images of them online. Sometimes you will need to sign up as a member to view these images, but it often is free and worth the effort.

This post ended up a bit longer than I anticipated, but I wanted to be thorough with the covering this before I begin talking about paintings and their design.