*Disclaimer: This post is about my learning process, which means that the first few images aren't that great. If you just want to see my good stuff, just skip the text and scroll down towards the bottom.
For all my artistic talent, I am notoriously bad at judging value--how light or dark something is; some people call it what "shade" it is, although in the art world "shade" means mixing something with black, which also changes the color. It sounds like using the correct value would be so simple, but there are about a million mind tricks going on in your eyes before an image even reaches your brain that make it difficult. Just to scratch the surface, there's an example here in which square A and square B are actually the exact same value. It's so hard to believe that even after studying this exact image in school and linking it here as an example I had to go into Microsoft Paint and compare the colors myself juuuust to make sure. And yep, they are exactly the same. I think that should be enough to give you an example of how difficult judging values can be.
I'm great at line drawing, pretty good at using interesting color... but getting the values wrong has been the #1 reason why some of my paintings have turned out to be horrible failures. (Case in point: the first couple of these. And don't try telling me they're good to make me feel better because you know I can do better; you've seen me do it.) I have a couple old black-and-white still-lifes from my oil painting class forever ago that are pretty awful too... maybe I'll put up some of those to demonstrate just how bad at this I have historically been.
I'm also pretty terrible at landscape. Sure, I have done a few cool landscapes and cityscapes in watercolor, but most of those were so abstract I'm not really sure they count for what I'm talking about. What I'm mainly talking about is just knowing how to compose a landscape and how to paint things in perspective so the mountains look distant and the sky is bright enough, etc. Fortunately I have lots of book knowledge about this; I just haven't had much practical experience, and every time I try and it turns out horrible I've wasted expensive oil paints in the process, which was really discouraging. And that's how I justified buying a drawing tablet to hook up to Photoshop, so that I can experiment all I want with absolutely no cost consequences and far less wasted time. Which brings us to the last few weeks.
Pretty much all of these images are honestly in the top results if you Google things like "landscape photography," "Pacific Northwest landscape," or "Still Life Photography." I had a couple rules for myself: I used a relatively small "canvas" for each of these, to force me to resist the temptation to get lost in the details. I wanted to practice lights and darks, particularly the large patches that really define the image, not the details. I used a palette of grays that started out with 7 shades and then grew to 9 and then finally 11 (which I found to be the best--enough to cover a large range, but not enough to confuse me). I never quite got to 50... lawl. Each of these took me about two hours; some of the sloppier ones took a little less and some of the more difficult ones or the ones that I just got really into took a little longer, but I don't think any of them took more than two and a half hours.
This was my first one. I started this using a color image to evaluate how well I could translate colors to grays. This was my initial block-in.
Aaaaaand this is when I got roped into filling in a lot of detail, despite my rule. Whoops.
At this point I turned my color image into a grayscale image and compared it to my initial drawing. This drawing is after my adjustments; if you compare this with the last image you can see that I was way off. The previous one looks like some weird glowing alien wasteland, while this one just looks like a pretty normal scenic pasture.
This one actually looks very little like the image I was working from, but it looks cool so I just left it instead of "fixing" it. My end goal is to make cool landscapes, and if I accidentally make one by breaking my own rules, I'm not going to complain about that. (All of these were painted in grayscale; I added color filters to some of them just to make them a little more interesting.)
This was my next attempt. As you can tell, I made a lot of the same mistakes I made on the first one. It was definitely an improvement though.
I accidentally deleted all the comparison shots showing how off I was at the beginning and how much I had to change. Oh well.
This one was a challenge but I actually was pretty accurate on this one without too much "cheating." This is basically the point where I began to feel like I knew what I was doing.
I found a sharpening filter that made these look painted. Not until I'd done this did I realize just how blurry and hard to look at the unfinished images are.
I removed the filter and started working with my artistic license. I lightened some areas and darkened others to increase the drama. Then I reapplied the filter, added a color mask, and the final result is the image on top of this post.
For some reason I made the canvas really tiny for this one. Whoops.
This was one of my favorites to paint. I really love forests and I would love to get really good at painting them but they are huge challenge. You are working with so many different objects that each have their own color independent of how much light they're receiving, but then those same objects have totally different values depending on what part of the image they appear in. Fortunately, I'm finding that I do better with high-contrast images to begin with, so this one wasn't as difficult as I expected--just a lot of fun.
I made some lighting changes here because the lighting in the other one didn't quite make sense. I'm not sure how much of an improvement it actually is. It's certainly more dramatic, because of the glow effects I also added.
At this point I decided to switch to still-lifes, because I was getting a little bored of landscapes and also I'm just as bad at still-lifes. I know you're going to look at this one and think, "You're not bad at still-lifes, this looks great!" That's because, again, I accidentally deleted all the steps working up to this, which I was going to use to demonstrate how bad it was at first until I not only put the original image in grayscale but started using the eyedropper tool (which tells you what color or what value is in the reference image) to figure out what the values were supposed to be because I was very lost and getting very frustrated.
This one seemed like it was going to be really hard at first, because I was worried about all the different colored fruits. Turns out that even though they were different colors, they were all pretty much the same values, which made it a lot easier. The only thing I don't like about this one is the grapes. If you're thinking, "what grapes?" that's exactly what I mean. They're in the bottom right but I didn't really make them stand out in any way except for the one that's glowing because of how the light is passing through it.
Can you tell which part of this image made me so frustrated I finally just gave up? If you guessed the very unconvincing drapery, you are right! I actually really like everything else about this image, but the drapery was just obnovious. I haven't really painted drapery like this since, so we'll see if I handle it any better next time.
This is the first still-life I have ever done that I am completely satisfied with. I really like the style I used and how I rendered the textures; I'm especially proud of the paper on the bottom right with the fake sheet music marks. I think I did an okay job portraying sheet music considering how little detail I actually used. Which is the trick to convincing paintings: you give just enough detail to suggest what something is and then let the viewer's imagination fill in the rest.
I also really like this one, though I don't think it's as impressive. I just really like painting bottles. Especially now that, thanks to these exercises, I actually know how.
I wanted to try my hand at a cityscape. I guess this one wasn't exactly "typical" of cityscapes but it looked cool and it did help me work out some of the problems I've faced with doing cityscapes before.
This one was positively infuriating. I started out trying to do it on my own, but it was completely impossible for me to figure it all out. I'll confess that by this time "cheating" with the eyedropper tool had become a regular part of my technique, although I tried to use it sparingly. With this one I just couldn't do that at all. The light transitions were just way too subtle for my high-contrast sensibilities. I used the eyedropper for almost every single stroke. (To clarify, I didn't pull colors off the photo; I measured them there, then used my 11-color palette to draw with the closest approximation.) I think the end result is really cool, but as I said it was a long and frustrating process. It was kind of a good one to finish this round of exercises with though, because it sort of humbled me. It's obvious from looking at the progression that I have learned a lot and improved immensely, but this one humbled me and reminded me that I have a long way to go if I want to be a serious artist.
Somewhere in between these, I painted a few black-and-white portraits, but I think I will make the portraits their own blog entry. Faces are complicated y'all.