12/22/10

Kevin O'Neill, Part 1: Painting and light.

Forums are a fantastic place to find communities of artists supporting each other. There are many spread across the internet, I happen to post on a relatively little one, The PA:AC. The forum is heavily based on critique and has a small core group of active posters, making for a personal atmosphere that discourages artists to just showcase art and run.  

Kevin O'Neill keeps the mood alive. Artist and moderator, he has provided  a massive amount of helpful and pointed critiques. The following quotes are from forum posts which I will do my best to contextualize and edit down, Click the link under each to go to the original forum posts.

I highly suggest reading the accompanying quotes, but my goal is that by scrolling through, you will see a progression in the lessons Kevin has to offer. This post will be long, but you can always bookmark it and come back!  

What Does Light have to do with Painting? 

Practically everything. Light can make the mood, the forms, and the composition. It can tell you how to color, what colors to pick, and why. If you are jumping into painting, digital or other wise, having some base knowledge of light will go a long way to making believable and interesting images.

One of the better, and simple explanations of light is the section in the Itchstudios tutorial, a staple of internet tutorials. 

We can start with bacons critiques of one of my own images, just to illustrate the difference between line and painting with light:
My original image is the 3rd image, when I was still playing wow freshman year in school. The other three are quick paint overs by Kevin. He pointed out some questions about the style:

I guess what I'd say is that I kind of (and this is just a personal opinion, right?) would like to see a stronger commitment either to the line-based style, or to a more painterly, light-based style. Right now it kind of hovers in between, without gaining the full benefits from either.  
What bugs me about the current state of your work is that the combination isn't quite bold enough that it read immediately as a coherent style, but rather something vascillating between two different styles.  
Original Post 
When I got that crit, I actually went the flatter, line based direction for a while. I think the advice is still sinking in. Now that I'm getting back into painting, I'm starting to pay more attention to light in general. Here Kevin breaks down how to simplify light.

Understanding how the tone changes as the form turns away from the light is essential to creating convincing form. This requires more than simply assuming a tonal value based on the distance from the light, or distance from the nearest edge, but a thorough understanding of how the object is constructed as a three-dimensional object.

Of course, looking at light in a planar manner is only the first step. Moving on from that are hard and soft edges:

Expanding on that a bit, you may want to try doing a 2-tone value study when starting out, just to force yourself to pay attention to your edgework, without getting caught up in the full value range from the get go. Just stick to a white and a flat medium grey, and use the fluctuations of edge quality to get the form to read. Think Mike Mignola inks if he airbrushed a few edges to soften and roll the form. I did some photoshop levels magic to strip down a lot of your values and tried it on your head here.
While your edges are generally fairly soft, you can see from my breakdown where I'm making them hard on drastic changes in plane, like the top of the tip, and on cast shadows. More rounded areas, like the cheek, are made softer. Watch how some edges start hard and then soften as they go along, not staying with one edge type along its length.  
A closer example of this:
Original Posters on the left, Kevin's example on the right. 
In the original, the edges are getting confused with each other- the edge of the cast shadow of the glasses is the same firmness as the form shadow of the cheek, the bottom edge of the cast shadow of the cookie flows into the the form shadow of the chin. Also, edges of the form shadow of the face are not consistent with each other, making them two distinct 2-d areas, rather than one area broken by a cast shadow.

When applying edges try to think about what the purpose of that edge is. If you look at mine, the form shadow edge is defined by one simple soft edge. The cast shadows are a crisp edge. As a result, the form appears continuous and solid, rather than two-dimensional. You can tell where the cast shadow of the glasses (crisp) meets the form edge of the face (soft), purely by the change in edge quality. 

It's easy when working from photoref to get caught up in rendering every piece separately, rather than putting your effort into figuring out how the form functions. The ambiguity of form and often unclear edge definition in anything but very good photoref makes putting conscious effort into how you design your edges essential, if you want to get a clear read of form. You need to make the form more clear and more understandable than the photograph.

This is a big part of the reason that it is often very obvious when people are working from photoref and not from life- simply replicating 2d shapes, rather than presenting a coherent representation of the 3d form.

Simplifying your light and properly noting your edges can help you quickly lay out a painting, here Kevin goes into the details of over detailing.
original posters on the right

I think you're running into a lot of problems as a result of trying to work out the drawing's structure and the painting at the same time, which is leading to an overemphasis on details, and too little emphasis on the light. For example, in trying to work out the folds of cloth, you've made each fold extremely contrasty and very much its own sort of object, rather than being subservient to the broader, overall forms- the block of the torso, the cylinders that make up the arms, etc. 

It's also leading to you forgetting things like the broad brim of the hat casting a shadow on the coat, the coat casting a shadow on the feet. 

I also suspect that while you're trying to design this character, you're trying to figure out the local colors- 'oh this is white, this is blue', so you stick those colors in- but when you think about the light and where it's coming form, those colors would be stuck into shadow or effected far more by the light's color than the local color, so it ends up looking out of place, and unrealistic.

I did a quick paintover here, but look at what I did with it- it looks more realistic and solid, but how I got there was stripping away and simplifying detail, not adding it. To be sure, in a more finished illustration, there would be more detail, but being able to figure out this broad, overall read is what will make the subsequent additional details hold up.
  Some step by step build up. These are rather self explanatory, but the original posts are still linked, if you'd like to see the images in context.



I think the best examples of light and mood come in his environment critiques. 

Original posters image is the top left. 
Darkness combined with bright, local manmade light sources means high-contrast, which isn't coming across here. That means you have to pay a good deal of attention to where light would be coming from in a real environment- in a place like a hospital where someone may be pulled up outside the front doors bleeding to death, there's going to be a lot of light so the emergency crews aren't screwing around in the dark. Entrances in general are likely to be lit up so they are easy to find, and general safety. Streetlamps are going to be on the sidewalks. Moonlight isn't consistent or bright enough to provide a significantly usable amount of light in a modern city, so don't rely on it in a painting of a city either.

So I threw together a paintover to add those light sources, and bring the areas that wouldn't be lit down to a more appropriate level. Note how the colors of the light sources are handled: yellow/orange safety safety light, dingy orange for the off screen streetlamps, green/blue glow of the entrance florescents.

Compositionally, the added lighting of the entrance, the color contrast of the area with the rest of the picture, and the addition of some space to the right to make the entrance more centrally placed makes it clear what the most important part of the piece is. In the original, it's difficult to tell what is meant to be the focus of the piece, which ends up giving it a rather dull, monotonous read. 

It isn't enough to simply draw the objects you want in a piece, it is necessary to establish a hierarchy of what is important and what is not in a piece through value, color, and composition. 

Up next, we can see some of the benefits of how a quick change of light can effect the mood and focus of the piece. 
Original Posters image on the right.
I would suspect it's mostly the lighting that's the biggest problem here, since it seems to just feel like this drab green-blue wash over everything. All the textures and models are great, but the lighting isn't serving to separate what's important to the scene/player from what isn't.

Given, there's limits to what you can do with light, and Maya's viewport rendering is never going to give you the lighting results you really want simply because it's not a game/movie quality renderer, but the point stands. If you can, try lighting the scene without an ambient light (or at least, have it set to emit a very,very low level of light), using directional lights for primary lighting and broad fills, and spotlights for focused areas- you can get a lot more subtle, effectively contrasting lighting this way. (This may be problematic for same game engines like Doom3 where all lighting is done in real time, but not for a good portion of them that bake lighting anyway (Quake/Source/Unreal). I believe you can also bake the lighting into texture in Maya for nicer real-time rendering than you'd get in the viewport normally.

What you might want to do is do a screengrab and do a quick paintover of what you've got in order to figure out where your focus is; ideally, you'd have a good piece of concept art to go by from the get go, but doing this mid-stream can save you some time if you just find yourself noodling around for direction. Did an example to try to center the focus around the big central crystal- maybe not the mood you're going for, but with something like this in hand, I'd have a pretty clear idea of how I'd want to set up my lights.

I believe this post is long enough, so I will leave it at that. I hope this information finds people outside of our little community and inspires them.


12/4/10

Its actually difficult to do a steady stream of one-off drawings and get things that really have a functional place in a portfolio. Because of this, I'm making up a game. Soon, the relationship will make sense, until then, you will just have to wonder what this book has to do with this girl:

11/23/10

Controlling Photoshop: Color Picker, Part Two!

So now that I've blogged about the technical reasons for switching your color pickers mode. Now I want to go into how to think about this while coloring, which will be easier with some visual aid. heres a doodle I'm going to color:

Dragon Suit McStabby

Alright, so lets put some base colors on this. 

So at this point its time to start thinking about light. As we established, the default color picker is in "hue" mode. I think thats the reason that I see a lot of pallets like this when people first start digitally painting in PS:
Now, for one, I don't like making the little blob of on canvas color swatches. I generally just use the color picker when ever I need a new color and use the eye drop tool to gain access to colors if used already. The only time Ive used these blobs is when I'm specifically using limited color.  

Anyway, Heres that pallet in action. 

Okay, so other than some sloppy quick application, you can see this is a functional selection of colors. Still, I like retina burning rainbows and toxic purple secondary light sources. The problem is finding those colors in this:  


to me is really tedious. Finding a blue that will go over my orange involves a lot of messing with the color slider and adjusting the brightness and contrast in the color field. Really what I wan to do is start with my orange and simply add blue, not start with a blue and adjust its brightness. So lets look at this chart from the last blog of the color picker in red mode.

In red mode you can adjust based on color rather than brightness and saturation of the hue you already have picked out. The amount of red will be locked by the color slider, and you can modify the amount of green and blue in the color field. 

Here is the base orange
Alot of red, a little green and a smidge of blue.

To get a shadow color, I'm going to add some blue, and take out some green. 

This gives me a salmon-y pink, but note that the color slider will display what will happen if I start to also take away red with the hue I currently have selected. I'm going to take away some red and green, and add some more blue. 



Now I have a purple, sweet. It might look a little odd as a color blob, which is why I hate color blobs. for proof of concept lets throw that purple over the base colors:
Hey, thats not too shabby! Way more vibrant than the hue pallet equivalent. Of course, when actually working this is all about flexibility and experimentation. I like to play with different color pickers, layer modes, and overlaying textures while I color. I hope this helps explain the color picker and gets you to try something new in your next PS session! 






11/22/10

Controlling Photoshop: Color Picker

  Using Photoshop can sometimes overload new digital painters with options, while stripping away some of their basic instincts. Still, my philosophy with software is to dig deep into it and play with all the options available. The more familiar you are with your tools, the better decisions you will make while creating a work flow.

 There is a lot out there about fighting the lack of texture and achieving better brush handling, I find that there is not a lot about color, so thats what I'm going to focus on.  The color picker seems fairly technical, but if you familiarize yourself with it, it can really help you find that exact color your looking for while painting, saving you time and allowing more flexibility when you are guessing and checking.



The color picker that is default for Photoshop is the "hue" mode, seen here



There is noting wrong with Hue mode. But when painting traditionally, I rarely lay out a color with a drop of pure black and pure white to mix with. This picker doesn't work with how I want to think about colors while working, but luckily there are other options. So first lets label the color picker so I can talk about specific parts:



Here in the color picker we have three major areas, the color field, the color slider, and the mode buttons. To change the color picker around, You'll be playing with the mode buttons. Generally when I suggest this to people, they click the mode buttons and find the new spectrum extremely confusing. The first thing to understand is that the mode buttons are applying themselves to the Color Slider, and modifying that subsequently effects the color field.

To express this visually, I want to explain what each one does. Lets break down the default mode, Hue:



In hue mode you are controlling the spectrum of color with the Color Slider. Once you pick a pure color from the spectrum, your Color Field controls the brightness from top to bottom and the hue saturation from left to right. The top right color is the pure color which is a value of 255. When we are looking at pure Red, R is at 255 while both the G (green) and the B(blue) will display values of 0. Pure white is always all three colors at 255, and black is all of the colors at 0. 

Put your color picker at pure red, and then move your cursor around in the window while looking at the values. You will notice that the G and B values are remaining the same (give or take 1) as you change the red. 

The two next modes are saturation and brightness, which apply themselves to the color slider.
Looking at the two of these should start to explain how the picker is truly functioning. I personally don't use either of these modes very often, but I find myself clicking over when color I already have picked isn't dark enough, and I need to better see both its brightness and saturation so I don't muddy my colors.

 



Under the hue, saturation and brightness options are RGB mode buttons, which I use a great deal. Much like H/S/B this will apply your selected color to the Color Slider. If you are used to mixing colors with paint, this wont exactly replicate that, but if you get used to this mode you might find it slightly easier to find colors you want.

Heres how the mode works:
Red will be 255 across the entire color field, and can only be adjusted by the slider in Red mode.

Here in red mode, you can now work with green and blue and modify your colors based on that. I find this the most convenient way to deal with "This color I just put down is not quite blue enough." Working in the color field gives you a larger plain to work with the spectrum than have every color laid out in the slider. If you are particular with your colors, its worth your time to give the RGB modes a shot.

 The modes to the left of H/S/B are "Lab Color". Lab color is supposed to approximate human vision. L is for luminance (light) while A controls how red or green a color is, and B controls how blue or yellow a color is. This mode, admittedly, is too odd for me to use while painting, but perhaps you will find them intriguing. Heres the picker in Luminance mode:

This is cooooonfusing

That.... is probably not a good way to chart it out. I will spend some more time with it and try and explain it again later. Soon, I will add some visual examples that practically apply the different color pickers in painting situations.

On to Part 2!

11/17/10

Continuing to work on landcapes and cities, they are staying relatively simple as I just figure out how to even process what a city looks like when I try to draw them. I've been trying to expand my knowledge of landscape artists and speed painters outside of Craig Mullins and dudes who worked on StarWars. Now with Animation Backgrounds on hiatus, I've been mostly using DA. Check out my collections over there if you are looking for some random eye candy.

11/8/10

MSPA


Did a little fanart for MSPA Better known as MS Paint Adventures, a comic by Andrew Hussie. Its latest iteration, "Home Stuck" has exploded the fandom that the site has. Andrew builds a world that really reminds me of being 13 and seems to have captured a captive audience of actual teenagers. 

Personally, I've been trying to figure out how and why. Creating something that is worthy of being obsessed over is not as easy as people tend to think. Plenty of comics start out with huge epic worlds in mind, but never get very far by ways of story telling, a good readership base, and a long lasting impact. Homestuck is doing all of these things relatively effectively, if its rapidly growing deviant art groups are any indication.

 For me, Homestuck has alot of the little parts of being a 13 year old gamer, in the beginning. Stuck at home, its snowing, talking my friends on IM to try and get them into some random game. Its set up perfectly, with the well crafted  "Pesterchum" logs really hitting the beats of instant messaging. With its adventure game elements, the story manages to pull those gaming tropes out into the characters reality. Its gone from being a skynet version of The Sims, to a Final Fantasy destiny killing epic. His characters all have obsessions, which we cling to desperately while trying to figure out our identity in those early teen years, and they are slowly evolving and growing with each kid.

So, part of Homestuck's effectiveness is coming from careful observation. For me, its nostalgia. For some kids, its probably even more true to their reality now than it was for me 9 years ago. I think the other factor is how Andrew creates rules for his worlds and not just complexity. You can see this in how Homestuck has really grown from what he learned in the previous story archs. The inventory systems that the kids use to carry around their stuff are convoluted math problems, to set the tone for how the worlds and timelines would eventually be jumbled together. His slow build helps ease the reader into what is now a crazy little universe.

I love the comic, though I miss the less log intensive beginning acts. If you havent read it, its worth taking some time to start reading, and see if you don't get sucked in.