Make It Jellybit

I'm Dave from Baltimore, and this is where I'm dumping the stuff I make.

[makeitjellybit]

I want to talk about something that changed a lot for me when it comes to drawing. For many years, drawing has been this monster of sorts for me. I defined myself by my art, and struggled with perfectionism. I used to be afraid of making a bad drawing, and being shown in very clear terms that I’m a shitty artist, thus making me shitty due to being defined as an artist. The anticipated drawing was bad, so I avoided it, and didn’t draw it. This went on for about a decade. I stagnated, while also hating myself for not drawing. I thought about it all the time, yet never did it.

A few years ago, I went to a life drawing session with someone to do gesture drawings. I remembered all the admirable voices in the past who told me that gesture drawing was super important for capturing life. I’ve studied animation all of my life, and pretty much every animation educational resource says to do gesture drawings. I hadn’t drawn in a while, so I agreed to go to this session to give it a go. 

This is what came about:

I felt awful that I couldn’t find whatever it was that I was supposed to find in these poses. My lines were super scratchy and light, two big signs of lacking confidence and conviction. There was nothing clear that I was saying with any of it. I left feeling bad, and being more afraid to draw.  I can copy what I see fairly well, and thought I knew what I was getting into, but that is a deeply different process from gesture drawing, which I didn’t recognize at first.

Recently, after attempting to animate for the first time in 3D, I decided that I needed to practice gesture drawing. I needed to understand how to capture the life of someone/something. So I watched a great video series on gesture drawings from Proko, loaded up my phone with pictures of gesture models that I bought from his site, and went outside every day at lunch to do gesture drawings for about 20-30 minutes a day.

I wanted to use pen to help build my line confidence, and not give me an “out” with light lines. I wanted to showcase my scratchiness as an unnecessary thing. I wanted to be able to break the feeling that these drawings needed to be accurate, and just capture the feeling of the pose. I knew that if I was going to get better at all, I needed to destroy the feeling that every drawing was/should be precious/permanent. So I decided to draw these gestures in a shitty ruled notebook that I grabbed from work. The goal would be specifically not to make beautiful drawings, but to fill it up with garbage experiments. The only way I could break free from perfectionism, is to draw garbage as a gift for a garbage heap, and not to please myself aesthetically. Whatever I was building was not going to be on the page, but inside of me. So that’s what I did.

I tried not to care about accuracy, and just exaggerate. Draw the feeling of it, even if it looks weird/bad. Have really long limbs, or bigger tilts. Do whatever it takes to break my focus away from accuracy. Between drawing for the garbage heap, and changing the focus toward exaggeration, I started to fill this notebook up fast. The image above is where I first started to see progress in this mindset.

Sometimes I’d lose my way, and forget to focus on the pose above all else, but those were still helpful days because I was drawing. I felt frustrated sometimes at the fact that Proko seemed like he was drawing slowly, making perfect lines to get a solid gesture in 2 minutes, but I was struggling and racing just to get something recognizable down. Then I’d regroup and remind myself that I’m SUPPOSED to be creating garbage. Piles and piles of garbage. The intimidation and frustration would then subside, and I’d get back to enjoying myself more.

I think the pen helped me gain confidence in my lines. I tried to lead the pose with the spine, but I couldn’t quite get it all the time. Sometimes it worked okay.

I tried to use the lines to aim the body into the action, straightening even when I saw it wasn’t straight in the model, so I could get the essence of what was happening.

I also tried to capture the main part of the pose first, then do the rest of the body, so that if I was short on time in my 60-120 second time limit, I’d still get the feeling down. Sometimes, I’d miss half the body, but I’d still get the gist of the feeling.

I became more comfortable with missing parts. Sometimes I felt it added to a pose by cutting out parts that didn’t contribute to the pose as much. Because I was prioritizing from the beginning, this culling happened automatically, which was a cool feeling.

These drawings went on for a couple of months. I liked when I could feel the weight of something. I think the hard flat lines on the arms on the left pose contributed a lot to that. Eventually, my drawings were something I actually felt some pride about. Ignore the Dora face above. I was trying to explain something to a friend. Or don’t.

This next one was from my last life drawing session, 1-5 minute poses (see the full resolution image here). I had some better captures of the gestures in some of my past drawings, but these are really pleasant to me. Plus when practicing art (or any skill maybe), it’s not about hitting the pinnacle of your abilities every time you sit down to do it, but about raising the average. Understanding that helped kill my perfectionism further, freeing me more and more.

Drawing most days, I didn’t really feel the improvement in the moment, but having those notebooks filled with garbage helped me to lay out my journey, and I see that I’ve actually traveled somewhere. It’s something I have to keep in mind with other long-term skill building exercises. I am going somewhere with these things. In fact, this blog was made to help me see that with a variety of skills that don’t fit into sketchbooks.

I’m so relieved to kill the majority of my fear of drawing. It was such a terrible burden. I hope these gesture drawings do in fact help me with my animation, but that’s not nearly as important to me as finding the freedom to draw again. I used to think my perfectionism was a great asset (which I would use to judge others... another false asset), but now I see that it had more negatives than positives. I will continue devoting effort to killing it in me. It will never die, but I need to show it that I’m the boss, that it can’t bully me anymore, and that I will call upon it only when it’s most useful to me.

[makeitjellybit]

Taking the time to do a high pressure self-test now and then is pretty important for getting me on track. Two game jams ago, I was able to do 3D art for the first time, but that self-test showed me that I didn’t know how to do an animated character, which is pretty important for most games. The game jam after that was my next big self-test, which showed me that I could make a character, but didn’t know how to texture my 3D art, which is lower on the game-necessity list, but still a pretty big thing. Just as I knew I had to learn rigging/animation after that first game jam, I knew what I had to do after the next one.

I figured that before I got into something as complex as characters, I should do some flat, tiling textures. I looked up tutorials on how to paint grass textures. It went over creating a special brush, which taught me a lot about Photoshop brushes. This was the result:

I’m very new to using color at all. Pretty much all my art in my life has been pencil/charcoal drawing, and color has been very intimidating to me. In the last game jam games, I used vertex colors, which was a simpler way to ease myself into making color choices. It turns out I still was pretty crappy at it, not only due to inexperience, but also because of the horrific realization that my laptop monitor was complete crap with color. There was angle from which I could view the monitor that would allow me to see consistent color/contrast. There was this maybe inch-tall row of decent pixels, but even those were way off from most monitors. So of course, this grass texture now looks nothing like what I thought I was seeing. It’s super neon and weird.

Another thing is that I never felt like I was actually learning how to paint textures. I can’t just make a special brush for everything that creates what I want by dabbing around randomly. I need to learn how to actually paint the shapes and gradients that I want. Then maybe I can make a brush to make that process easier in places, but I really need to know how to paint on a basic level to do painted textures for cartoony games. I looked around at what other people were doing and really liked this one that I found on Deviant Art:

Geez. So, this artist built pretty much all of the texture out of making those attractive fluffy tufts. How did this happen? How did they get this look? What did they start with? This is a commercial texture, so there was no chance of a tutorial. I had to reverse engineer it.

Here are some initial tests. As I noted before, my color perception was way off on my laptop monitor, so I just stole colors from the original texture to make sure I was on track. On the right, I did my first attempts. I went back to making a custom brush, because I did learn those skills more deeply with the last test. Then I realized I hated the result, and that I again need to learn how to paint it myself in some way. So I made some blobs on the left that went from darker on the outside to lighter in the center, and pulled out blades from the different parts of the blobs. That wasn’t what I was looking for either, so I made another attempt.

Here’s my attempt to paint the grass manually. The result didn’t feel nearly as lush and appealing as the texture that inspired me. It also didn’t feel that interesting. I felt that maybe if I can’t copy someone else’s approach, then maybe I should learn by creating my own. So I went with a little more vibrant palette, and made something more abstract.

This is what you can see in the main image of this post. If you want to check it out in 3D, you can go here. I kinda liked it, though it felt more like leaves than grass. I just kept it as is, but it made me wonder what it would be like to actually attempt to make tree leaves. I was inspired by the following fluffy tree.

As you can see, it’s an egg-shape with a “bunch of leaves” texture on it, and they added leaf planes all around the egg. There’s a top-view texture of the 3D leaf, and side views. Put them all together, and it looks like a little 3D branch. So I tried to paint my own leaves in the same way.

I didn’t want it to look realistic. I still wanted a cartoony look to fit a cartoony world. I don’t know if it’s that good or interesting, but this is the result I got. So with the branch textures done, I had to make the egg texture. I took the “top view” textures, and copied a dark version all around, then added lighter and lighter versions on top to get the egg texture. This is what I got:

It ended up looking more realistic than I was aiming for. This tiles perfectly in 2D, but not so well around an egg shape. Still, the result looked okay. The whole process made me feel super weak in painting, and this brought me to a conclusion to the year of cramming 3D art (which followed a year of cramming programming), and moved me into my next phase of being able to do better 2D art. The result of my last game jam showed me that I need to be able to do better concepts so that I can have more appealing 3D models, and this experiment showed me that I’m weak at texture painting. Hopefully my next phase will improve both.

All the textures that I’ve created here are under the CC0 license. Have at it.

[makeitjellybit]

You saw in a previous post that I’d tried to use my new knowledge of 3D art in a game jam. In another post, I talk about how immediately following that, due to my feeling the limitation of not being able to animate, I crammed for a week and did my first 3D animation ever. Well, in the last Global Game Jam (48 hour game development event), I decided to try to put myself to the test. The last time I did any rigging/animation, it took me over a week. This time I would have to do all of that in a day, in addition to designing and modeling the character. I also had to do background art and props. This was really intimidating, but I felt I should see what I had in me.

I started trying to wing it with modeling a character, and hated the result. It had no appeal and felt super stiff. It seemed that a 2D character design and model sheet were necessary. I based the character on a friend of mine, but haven’t told that friend yet because I don’t know if that’s weird! Here’s the model sheet:

It’s pretty slapdash to get it done in time. As you can see in the video above, I didn’t end up doing the stripes on the shirt, but I got pretty close to the t-pose version. I needed around 15 different 3D props, the background, and a few character animations. I was able to rig her up, animate her, and get the rest done in a single day! I have ADD pretty bad, but was able kept on task the entire day thanks to the pomodoro technique.

I’m very happy about the success of this skill test. The work isn’t polished at all, but the fact that it all exists in one day’s work makes me pretty happy. If you want to take a look at the animations in real-time, here’s the sketchfab link. The main things I learned from this were that texturing is a weakness of mine that I need to build up, and that concept art is a HUGE deal, which happens to be another weakness of mine. Great! Now I know what to tackle next. Looks like I’m focusing on 2D art for a while.

[makeitjellybit]

Geez, I haven’t updated in a while. I’ve done so much since I last posted here, and I really want to keep this up to date. So yeah, please bear with me on the next few posts. I was digging through my old art and found these drawings. They’re from September 2014, when my game development group decided to have art challenges based on randomly generated concepts. The theme for this one was “Interstellar Rain Man”.  I’d been in programming world for a while by that point, and wanted to test my drawing/painting ability. The plan was to have a man in space made out of water with cloud hair. I did the first one with my drawing tablet and didn’t like what was coming out, so I jumped to good ol’ reliable pencil and paper for the second attempt. Obviously I didn’t finish the concept on either, partially because I was trying to submit at the last minute. I didn’t like either of these, but that’s okay, because my discontent is what led me to seriously try to build those skills later.

[makeitjellybit]

Learning ZBrush - More buildup (low on time) by Jellybit - 3D model

Not sure what’s going on with her right hand. I have to get back to a previous project, so I’m going to have to wrap this up real fast, but I’m having a lot of fun learning this.

Took a small break to learn ZBrush in October. I’ve been poly modeling for much of the past year, with some Blender sculpting earlier last year. ZBrush is the standard when it comes to sculpting, and I wanted to get a feel for how it works. I decided to go for this tutorial series on youtube.

I throw all my tutorial progress into this folder, so you can glance at a few stages of the T-Rex there (while ignoring my Blender sculpting from a year ago):

https://sketchfab.com/Jellybit/collections/tutorials

ZBrush’s interface is pretty maddening at times, but one thing that really helped was to pretend that it was designed by aliens and I sneaked onto a UFO to use their future technology. Once I let go of anything being intuitive, I started to really enjoy myself. After getting into the groove, I learned a lot about using reference, partially because I didn’t like the art direction of the tutorial. The tutorial made a helpful point to not think of ZBrush as sculpting clay, but as drawing in 3D. For some reason, that helped things click. Perhaps it breaks down the expectation that you should be feeling the material.

Unfortunately, I was not able to finish the tutorial. I got as far as the initial Sketchfab link above due to my laptop reaching its limits. By the end, I had cut the model up in 10 different pieces, but it was still chugging. Someone since told me to use ZSpheres instead, and that this would help performance. I look forward to trying it again with that knowledge, but I’ve moved on to other things since.

Before I moved on, I participated in the Halloween drawing challenge where you draw a different halloween themed object/monster every day in October. On one of the days, I wanted to see if I could use ZBrush to do a quick 3D sketch. It took me 2 days but I made this ghost that I dig:

The interactive 3D view of the ghost can be found here: https://sketchfab.com/models/6033da13545d427c85cdfbb08e66e4f4

Anyway, that’s it for my ZBrush flirtation for now. Last week, I ebayed for an old cheap Surface Pro 2 to use as a portable ZBrush machine. We’ll see how that goes now that I’ve been given some performance tips. Several pro ZBrush user friends say that they use(d) their Surface Pro 1 for ZBrush just fine and love it. Gives me hope!

[makeitjellybit]

In my last post, we left off at the point where we finished the game jam game. I decided to see what it would be like to make a character under the same time constraints that I was under with the environment art. Time constraints can be really valuable with self-education, because there are infinite rabbit trails you could follow when learning anything. A time constraint is good to remind you to keep on track. The character turned out a little crappy, but it was fun and fast to do. As I finished it, something hit me. 

I’ve had a lifelong dream to animate in 3D. I have always obsessed over animation, studying how things move, but there was this incredibly daunting amount of software knowledge between me and doing that in 3D. Just layers upon layers of vertices, UVs, axes, and bones. I had an epiphany that  I had a golden opportunity here. I had this throwaway model that I didn’t care about, meaning my perfectionism could be kept at bay, and I had built all this 3D knowledge so that I didn’t even have to think about that stuff anymore. I was just a hair away from learning how to animate! A step away from a life milestone! So I made a plan. I decided that if I learned rigging for a week until I had a working rig, I could be animating by Monday. With my eye on that prize, I began cramming.

I followed this wonderful tutorial on how to rig characters on Blender Guru. It had those snazzy control gizmos I see everywhere as you can see above. They always felt as classy as they were mystifying to me. Not only are they visually appealing, and seem to make movement easier, but they clean up the view of unnecessary information to help you think more clearly.

This is how that rig worked out for me. I simplified a couple of things, but the functionality is all there. As you can see, the arms/head don’t inherit rotation, because the zombie would probably be facing/reaching the same way regardless of body behavior. I really enjoyed the IK stuff, and the foot roll control was just magical for me. The logic of the bone relationships and constraints really appealed to me, though I’m sure I could get sick of it on a higher quality production.

I finished it late on Sunday as I hoped/planned. The next day, I would have normally gone to work, but I had off for Labor day, and it felt like a divine gift. How major was this for me? Well, decades ago, as a child, I saw the Amiga Juggler in a store. 

This may seem silly for such a crude animation, but it was like nothing I’d ever seen, and it turned computers from complex calculators into machines for art. Surreal art, that wasn’t like anything I’d experienced before. I just felt tingly inside, and it seemed like I was looking at magic. A love for animation and digital art was born in me then, and has been in the back of my mind since. The moment I finished that rig, that world of 3D animation was opened up for me for the first time. All that pent-up desire/energy that festered for decades poured out of me. I dove in hard and just started animating a walkcycle like the dickens. It didn’t all go smoothly.

One thing I added that the tutorial didn’t cover was a root node. I know others have used this, but I didn’t really know how it worked or why it was there. However, I hit a point in my animation where I had to start over completely. In videogames, the character animation is generally stationary on the character, so that the game’s programming and player controls can determine the character’s movement instead of the animation itself. 

A common problem in animation is that the character’s feet slip on the ground. I had to animate with the character moving forward in 3D space, so that I could make sure that when moving, the feet got planted firmly on the ground. This was working fine, but the problem set in when I then attempted to make the character stationary again. I kept the character in place, but the feet kept moving forward in 3D space away from the body. The IK keyframes were all wrong, and I couldn’t think of how to fix that. I didn’t get the hip node to stay reliably in the right place. I needed a way for the hips to work right in local space, and for it to seem like no one’s moving at all from the feet’s perspective. I remembered seeing this circle object below videos of other game character rigs, and thought about why it was there. I started to figure out how it would work, and implemented it. Basically, you create this node whose job is only to move the character, and you make it the parent of everything, including the IK handles, which normally aren’t parented to the root hip bone. Because it does absolutely nothing but move the character, there’s no weirdness with rotation or height changes. It just moves the character forward so you can test your feet placement and get a feel for the speed of walk.

I had to throw the attempt at an animation out and start over with this root node, but luckily I figured this out early in the day. I charged into the animation using the root node, and finished it by my local coffee shop’s closing time. I did it! I set a goal to finish rigging in a week, and animate on Monday. Here it was, Monday, and I finished my first 3D animation! It took 6 months of cramming 3D, and pushing on that door. That door has now swung open, and I’ve fallen into a running stumble into this new world that I’ve dreamed about since I was a child. Sketchfab for the walkcycle I made that day can be found here.

Here I am, a week after the animation, and I realized something. I thought back to my very first 2D animation in 2006, and it too was a zombie walk cycle. Pretty cool how those two firsts lined up!

Thank you for allowing me to share this journey with you. I’m happy I’m in a position to document this point in my life, as this feels like a pretty major milestone for me. I will do a bit more animation with this rig, then I will take on the more difficult quadruped goats. Wish me luck!

[makeitjellybit]

It’s been six months since I started my auto-didactic journey into 3D art, and so far, it’s just been a slow exploration of the process at my own pace. The weekend before last was a test to see where exactly I stood with this 3D art knowledge, and how it would manifest if I were on a project with a tight deadline (which is anything that might pay the bills). Some friends from Firaxis asked me to join them to make a game in a weekend for Ludum Dare. I’ve been in several game jams before, but among other things, when I did art, it was 2D. This would be the first time I had the skills to take on 3D without worrying about learning interfaces and concepts at the same time. I’d spent all this time building skill, and now I got to open that nozzle to see what happens when I focus on speed. 

I had to focus on speed a little more than normal on this game jam, since my father ran into health problems earlier in the week with multiple visits to the emergency room. I wasn’t even sure if I could participate, and told the team that I’d be away most of the time, but between visits with my father, it was kinda nice to put my brain on a task. My father is still in an iffy place, but that’s a different subject for a different venue. The point is, 48 hours is already a tight enough schedule, but I had to deal with far less, which I might even prefer, due to my experience in a previous post.

Time was tight. We had three character types and two types of buildings: houses, and a church. The characters I made in Unity using primitives, so they were capsules with capsule arms, some holding a torch with fire particles. They actually worked better than the primitive-based houses, so I decided to start by fleshing those out. As you can see in the main post image, I made some feudal houses with stone chimneys. For loose reference, I grabbed some pretty concept art in google images.

The exaggeration really appealed to me. I made a basic model and started painting a texture outline.

It hit me pretty quickly that I wasn’t using my existing skills, and on such a tight time-frame as this, I shouldn’t be struggling to learn how to do a full UV mapped texture. My goats (from previous posts) were UV unwrapped and textured, but they mostly lined up with solid colors that acted mainly as a palette. I only used deliberate UV mapping for the alpha on the fur fringe, which didn’t require a full sensical unwrap of the body. My work time would balloon dramatically with this approach, and I realized that textures are not necessary. One nice thing about a tight deadline is that you can cut to the heart of what’s necessary, which is a good skill to build. I could do what I needed with vertex colors and shading very quickly, so I switched to that.

I built the house, and initially started putting in all kinds of details like in the concept art, with lots of polygonal planks and stones, but then pulled back, again, to keep cutting fat and stay focused on just the necessary heart of the task/goal. Where I DID allow myself to put some detail is in the chimney. I ended up really liking the result. Several people who saw it assumed it was time consuming, and that I made individual stones, but really I just made a 3-face-wide grid, extruded/shrunk polygons and marked edge splits between them.

You can see the basic outcome of the whole house in the initial image on this post, but what that does not show is my full approach. One of the reasons the primitives stuck out to me in our initial version of the game, is that they looked way too identical. So I copied elements around the house (for example, giving the house four chimneys), and wrote some code to randomize which parts show, and the height of the house itself, like so.

If I had more time, I would have liked to randomize other elements, like the door location, stones in the wall, and slight variances in color. Here’s the final result on Sketchfab, so you can rotate around it. I used to embed these sketchfab models, but it messes the formatting up on mobile, so I’ll just leave it as a link.

Next, we needed a church. I took the house, enlarged it, and added a steeple section like so (Sketchfab here).

I also added a stained glass window at the top, because that’s a pretty clear indicator of it being a church. I implemented, but removed the clear religious symbol on the steeple, because since you’re a monster attacking the church in the game, I didn’t want it to be about attacking a specific religion. The flower is actually a religious symbol, but it’s a secondary one that most people wouldn’t recognize as representing the religion as a whole.

I wasn’t able to be there with everyone as they worked on the game, so I feel a little separated from it, but I was happy that I could contribute what I did in the limited time I had. 

After the art was in, the only thing left that stuck out was the characters. The game jam was over, but I decided to start modeling a character anyway, putting myself under similar a similar mentality of efficiency under time constraints.

I made this pretty quickly, and something hit me. Since bipeds are simpler to rig/animate than quadrupeds (as my goats are), I could use this throwaway model to learn rigging quickly without any risk or pressure to do it well. I set a goal of learning rigging in one week, so that I would be animating by Monday. In the next post, I’ll get into how that went.

[makeitjellybit]

We left off in the last post with a buncha 3D goats that had shapes that appealed to me, but still felt stiff and impersonal. I wanted to add some looseness to them while introducing more character exaggerations. I realized I never concepted my characters, so I dug out the drawing tablet and went for it. Using a weird vector-painting program called Mischief, I attempted to draw something closer to the feeling of the goats, without vertices and weird shortcuts getting in the way. The result is shown above. You can watch a small timelapse of the blue goat here, and the yellow goat here. I decided to go with unshaded art to see if I could get clear silhouettes with basic color differentiation on the main parts I wanted to make clear.

The palette could use some work, but it worked well enough I feel. The blue goat got a muumuu of sorts, I added fat pockets to green, red got a more exaggerated stance/proportions, and yellow I based more on my knock kneed cat who always has this weird curved posture. In the last design, her horns made everyone think she was a deer. I wanted to communicate the spryness of deer, but it turns out I went too far and made her a literal deer. Instead, I went for something streamlined like a concept race car/plane. I think it works better.

I brought the painting into my 3D modeling program, Blender. Positioning the concept art behind my model to match the angle in which I painted it, I got to see just how far off my model was, and where it’s lacking personality. At first I was afraid of messing with proportions because I wanted them to use the same skeleton, but I just tossed that concern aside in order to learn what I wanted to learn about character design.

You can see before and afters of my first attempts to match the concept art here for Yellow, Blue, Red, Green, and Purple. Though it’s missing comparisons, you can see the final result of shape matching below.

Link here for those who can't see the embed.

Now, obviously I’m missing details like facial features and fur, so it seemed obvious that next, I should attempt to match the looks of those. I’m still new to texturing, but decided that I’ll need to use alpha textures to get the fur looking the same way. I might take a different approach in the final version, but I’ll share my hacky clumsy first attempt. I created a palette texture for all the goats to share, like so:

On top of this, I added an alpha texture. Since I had to use the same UV set, I made copies of the alpha in each color. It ended up looking crazy, see?

This is likely crazy inefficient, but I didn’t really want to figure out how make a Unity shader that used two UV sets. At least not in this stage of my self-education. Later, I’ll probably try to use vertex colors with a far simpler alpha texture, but I’m not sure how complex that process will be just yet. Here’s how the UV layout looks on the red goat:

Here’s the result of that fur test in image form.

and here’s the Sketchfab so you can inspect different angles and such.

Link here for those who can't see the embed.

Overall, I think I got it close to the art. It has some problems, but I look forward to improving it in the future. I’ve brought you up to the current day with my goats, but not with other things. In the next post, I’ll go over my first real test of my newly built 3D knowledge.

[makeitjellybit]

Goat progression

Sorry for the delay. I’ve been getting into a lot of exciting things that have distracted me from documentation of my journey. My goal is to catch up this week to the present day. When we last left off, I had moved into the world of polygons from voxels to make goats for a possible game project. As you can see above, I went for a smoother style based on one of several basic shapes shown in a previous post. My goal was to find good character gestures/flows. I’m pretty inexperienced with character designs, so bear with me. While I haven’t designed many characters, I have had a good deal of obsession with the process of designing characters and animating them all of my life, from my earliest memories. In that time, I formed some ideas on what makes strong characters. This doesn’t mean much when you aren’t testing your understandings by actually making characters, so this process was pretty enlightening. For now, you have to deal with an awkward phase. I’m STILL in that awkward phase. Anyway, the point is, I have these untested ideas about the importance and approach of shape and flow for characters, and what you’re seeing now are my initial tests. In the previous post you saw that I’m dealing with several goats of different shapes and colors. These goats have to be immediately recognizable from each other, as you’re dealing with herds of ‘em, and the type of goat matters. Here’s my first attempt to follow my basic shape sketches as a test for feel and differentiation.

Link here for those who can't see the embed.

I went with fat, average, and skinny to get a feel for range/contrast. I also worked a bit with horn shape to match. They weren’t extreme enough, and felt too samey to me. Their masses and shapes felt too similar, even though I was going for extremes. I went back and exaggerated some things further, and revisiting the original pixel art in the case of the blue goat. Here’s the result:

Link here for those who can't see the embed.

Blue goat got an overhaul to look more like the source material. Added some beards, and decided to contrast the characters more in weight distribution, rather than weight amount. I think it works better. As you can see, I also was doing some material tests. I wanted to see if using flat shaded specularity on the hard parts like horns would make it feel hard visually. I decided it didn’t quite work. It looked a little too abstract and didn’t quite convey the feeling of the material or the characters. In the end, I felt that the personality wasn’t quite where I wanted it. So I did another test based on lessons I learned by studying a friend’s art.

Link here for those who can't see the embed.

It might be hard to tell, but I played around with placing hard edges in areas that I felt would accentuate the shape/feel of the character’s body. For instance, on the red goat, I tried to make it look more muscular with harder lines. I also reverted the horns back to a basic material, which works a bit better. I liked this more, but it still wasn’t there. I found the shapes appealing, but just didn’t connect with the characters themselves. That’s when it hit me that I never drew concept art before modeling. I was just feeling it out in 3D. Drawing is something I’m more comfortable with, so I thought the characters might come out better there. Then I can try to capture that feel in 3D. In the next post, I’ll go over concepting, and translating that concept painting into 3D.

[makeitjellybit]

First polygonal goat attempt

Alright. The voxels didn’t work out for me. I got bummed for a day or so because I thought I was being skilled out of our project due to not having any 3D skills, but then I decided to dive in anyway and see what happens. You may have seen my previous posts about going through the character modeling tutorial. On the last post I made about that, I mentioned being ready to try out my new skills on the goats. And here we are. Now I attempt to make something of my own design.

One thing I remember from drawing cartoon stuff is that simplification is hard, and it’s best not done out of ignorance, so study how the real thing looks/moves first. The best references I could find for modeling were some anatomical drawings. They weren't optimal for learning the real deal, but the perfect side view was important to me, and I felt they'd come in handy when rigging comes into play.

So I took that, and made a slightly exaggerated cartoony version, trying to take note of subtleties as I went along. The result was the following:

Link here for those who can't see the embed.

I personally think it as an okay start for a beginner. But still, though it’s stylized, it felt a little too steeped in realism for me. I wanted something more cartoony. Something where the body itself said something. After getting a basic idea of goatdom, I sketched out some more expressive shapes.

I thought for half a moment about modifying my current model, but I thought back to a speech by Louis CK that I won’t get into here (maybe another day), and thought about how I’d learn a lot more by making another from scratch. I opened up my sketchbook, and went for the upper left shape in that series of sketches. The following is what came out of that.

Link here for those who can't see the embed. Alright. I feel decent about how well I captured that flow, though I ended up fixing a lot of subtle stuff along the way later. So I have this. In my next post, I’ll get into where I went from there.

[makeitjellybit]

A friend I haven’t talked with since 2007 contacted me recently, soon before my post about learning 3D in fact. He’s done fairly well for himself since we last made a game together, starting his own company, and making successful games, one for Cartoon Network. Anyway, he brought up the last game we made together (shown above). He wanted to remake it. I got pretty excited about revisiting and animating those goats. I was even feeling nostalgic. I missed those little furry folks. We agreed that the new one should be 3D. Since I didn’t know 3D modeling, we tried to figure out how we could make the game with skills readily available. He suggested we use voxels. I did some tests. My first test attempted to be literal with the source material. Here it is below.

Yeah, no. This tests made me realize just how much is left to the imagination in pixel art, and that voxels don’t leave that same room to mentally fill in the gaps in 3 dimensions. The lighting especially makes each block painfully obvious, when in pixel art, they could blend together better. This voxel stuff would require a new approach. So I looked into how other games like Crossy Road used voxels. After some studying of how voxelers approach simplification, I worked out some ways to do it. I realized that what they and others do is design in 2D, then thicken the art. An animal, being longer than wide, would generally be designed from the side. So I gave it a go.

Here’s the main 2D design I went for using this method. I heavily borrowed from Crossy Road as a learning exercise, though there are differences, like the head being pushed forward and down, the horns are on top of the head instead of behind, and other details. Here’s how that design looks when pulled into three dimensions:

These look a bit appealing compared to the first test. On the end there, I tried to see how I could work with different fur markings. These designs were ok, until I oriented the goats in the same way they appeared in our original game.

Eh… the readability is pretty low. The horizontal goat pupil eyes help to see that they’re eyes at all from the front, but it’s still pretty lame design-wise. No wonder the recent voxel games on mobile use an isometric view. That really is best for that style of simplification. Let’s orient the face closer to the front to see if we can fix this.

Hmm, that’s ok I guess. I like it a bit more than the last ones, and they have more personality. It’s doable for a quick project maybe, but I wasn’t wild about it. We dug further into voxels, and found that it was pretty hard to import any decent geometry into Unity. It was extremely inefficient. I think my friend felt the same way I did about the designs, so the technical difficulties just pushed me over the edge to admit that it had to be proper polygonal 3D. I got really sad at first because it seemed I was being skilled out of our own game. I definitely did not have enough time to learn something as complicated as 3D art in time for the game pitch. But I decided that I would try to cram 3D art against all odds. The worst that could happen is that I could gain some new skills and catch up with an old friend. Determined, I dove in. You already got posts of my attempts to follow that character tutorial at the beginning of my 3D cram here and here. The next post will show how I attempted to apply that knowledge into goats.

[makeitjellybit]

Tutorial Character 08 by jellybit - 3D model

Use Sketchfab to publish, share and embed interactive 3D files. Discover thousands of 3D models from games, cultural heritage, architecture, design and more.

It seems I haven’t updated in a bit. I’ll post a thing or two to catch us up. Here is where I left off on my tutorial. The polygon modeling was a ton of fun, as was the sculpting, but something happened during the sculpting. As long as I was in the technical part of learning, I moved quite fast. Once I got to anything resembling artistic, I started to slow down tremendously, adding more and more detail unnecessarily, on something I didn’t even care about. My old anatomy lessons wanted to jump in and contribute. 

After realizing what I was doing, I talked to some friends. They suggested not going through tutorials thoroughly, but to instead learn exactly what I need to learn, and apply it to my own art, revisiting for clarification if necessary. I took a look at how many skills I built going through this Blender introduction that I hadn’t used yet, and saw the work I was putting into something I wasn't even supposed to care about. I decided to end the tutorial here, and pour what I’ve learned into my own work. It makes a lot of sense, because so much of learning is problem solving, and you might as well solve problems in something you care about. The main problem with this approach for me, is that perfectionism rises with my own work. We'll see how that goes.

[makeitjellybit]

Tutorial Character 05 by jellybit - 3D model

My current project requires some 3D work, so I decided to dig into learning 3D properly. I am going through a tutorial in Blender, and I’m having so much fun. Here’s the base mesh I’ve put together. The next step is sculpting, then I move on to rigging.

I’m a pretty big geek about the craft of animation, so I greatly anticipate the rigging portion. I also feel fear regarding the barrier between me and the 3D object for when I do get into 3D animation. 2D just allows for so much hand waving and mentally filling in of the blanks. I hope I don’t drive myself insane.

As a recovering perfectionist, one major skill I’ve been trying to cultivate in the past couple of years is having patience with myself, giving myself room to learn. That’s worked great for coding, but me and animation have a history. My animation critical eye is just too big and pulsating. My new skill is going to be pushed to the limit. Hopefully I don’t fold under that pressure.

My next post will get into the face sculpt.

[makeitjellybit]

I found this interesting bit on a fascinating wiki called The Cutting Room Floor. There's a Developer Mode menu in the arcade version of Jojo's Bizarre Adventure (1998) by Capcom. In this menu, you can find some unused animations/art. In the case of Anubis with Silver Chariot, we get a peek into Capcom's production process. In the first image above, we see the original concept art for this character done in pencil. When you go into the Developer Mode menu, you can find that this concept art was scanned in and saved in the sprite's resolution (though the palette is messed up for this debug menu). The third picture shows how they blocked out the sprite by drawing over the scan, and colored in with a solid color. This overlay wasn't shown in the debug menu, but I thought I'd demonstrate how well it fits. The fourth image shows the sprite pixels alone. Finally, I included a sprite used in the game to get an idea of the final color/shading. It's in the middle of a move, so it has some smear effects, but you can get the idea.

[makeitjellybit]

I had a great time last night at our Bmore Indie Games meetup. There were several VR projects, and a lot of other things in the works. My piece of shit google cardboard thing fell apart. It's had a hot glue gun to it four times now, and I think I'm going to call it quits. Time to find a better one. My friend Greg let me use his, and it worked beautifully. The image was just so much better, and it didn't fall apart. It also didn't cut into my face like Freddy Krueger. I'm going to pick it up soon. So yeah, it was great to catch up with people. A lot of people liked to play Velocity Squared/Lava's a Battlefield, and a couple of great discussions really excited me. We're cooking up some very cool stuff. Hopefully I can talk about it soon. Had a lot of fun, but I'm glad to put that little Virtual Boy build behind me and dive back into my my other work.

[makeitjellybit]

Welp, between me, two roommates, and a girlfriend, we all got sick for over a month with everything this winter season had. It started the day before the VR meetup, and echoed through the holidays. This was pretty disappointing. I gave myself a week to finish my project for the meeting, and though it took me a week and a half, I did get something together. I couldn't hang out with everyone, but I was proud of what I squeezed out. Luckily, we set up another meeting for January, and I decided to bring it. I've embedded a video above about what I decided to do in that week and a half. As I mentioned in an earlier post, I had that google cardboard thing, I had an old prototype, and given the color scheme of the cardboard thing, I decided to convert that old prototype to a pretend Virtual Boy game. According to my plan, this involved a couple of things: 1) I had to have two cameras, one for each eye. 2) I had to redo the art to match Virtual Boy's color palette. It ended up being a little more complicated than that, with problems in both areas requiring extra work. Adding a camera wasn't enough. Since the cameras were orthographic, there was no depth to the view. Using perspective cameras went against the Virtual Boy's visual style, because distant pixels would be higher resolution. I decided, maybe foolishly, to fake an orthographic view through a script. I put the background layers under control objects, and had the camera dynamically resize them when it moved on the z axis. I was a little concerned that this would cause problems with the physics, but it turns out it only has to do adjustments on the first frame, since past that, I could simply zoom in and out with a change in the FOV. Foolish to take that on in a week-long project, but I was proud of the result. The next problems showed up with the art. I realized that I didn't have enough visual markers to perceive 3D. I had to create a background, and a few background set pieces like stalactites. For the foreground, I needed to give the lava some texture for the eyes to focus on. I ended up making the following animation.

Finally, with testing on my cardboard, I had to shrink the view to make it fit within my visual range. This made the characters very small, and they looked distant. The whole thing looked static too. So I put in a zooming effect to bring things close while keeping all of the action in view. I wanted this in the original prototype anyway, so it was nice to have an excuse to put it in. Several bugs are left, like the highlight color making the blocks invisible, and the strange disappearance of the death effect. Oof.

Well, that was that. There are a million more things I could do, and the lack of polish is embarrassing, but I gave myself a week limit and went over a few days, so I called it done. I am now back on some previous work, after the month and a half break from being sick that is. I have some thoughts on that, but I'll save them for another time.

[makeitjellybit]

I first tried to get into the games industry as an artist, but wrote up game designs on the side. Later I got hired as a designer, and started learning programming on the side. During that time, there was one thing I did that helped me grow tremendously. I foolishly, as an absolute beginner, attempted to participate in a 4 hour game jam.

We had to complete a full game in 4 hours. I'd learned a little Game Maker before that, but I never made what you'd consider a game. I failed to make a game in 4 hours, but did finish in 8 hours, and very much succeeded in many other ways. I was forced to plan as small as possible, and this allowed me to learn a whole lot in a short time. I had others participating, so I had a lot of motivation to finish. I learned a lot about cutting down to the simplest form of an idea, which has helped me in countless ways ever since. It also taught me significantly about how to approach/learn the valuable skill of prototyping. I called the resulting prototype "Velocity Squared".

In this prototype, as you can see above, you play against another player, picking up pieces of the floor and throwing it at each other to knock each other into the lava. The lava slowly rises as a loose time limit that you can partially avoid by stacking boxes. You both make the ground more precarious by trying to win, taking away your own stability and trading it for an offense.

Last April, I remade this prototype in Unity in a week out of nostalgia maybe, but also to learn Unity's relatively new 2D system. It was a fun diversion. I packed it away on Bitbucket, and moved on to other things.

As I mentioned in my previous post, I mentioned that I would be taking an older 2D project and making it 3D for a VR meetup. Well, this is that project. As I mentioned before, I want people to play works in progress, so here's the Unity version if you'd like to try it out. This is my first time sharing it with the public. The xbox controller setup may not work on Mac, as their axes are different, but whatever. You can throw in different directions by holding a direction when you throw. Throw downward to make yourself double jump. Give it a go and let me know what you think. It's far far far from polished, but it gives a general idea.

The VR meeting is Saturday. I'll post soon to show what changes/improvements/breakings I've made in the past week.