Walk the Walk

Finals came and went for Character Animation 1 and I can easily say I've learned a lot from the projects this month. The last animation project I have to talk about is the Bugsy walk. You may remember Bugsy from the Strong Poses project, but this is Bugsy 2.0 with a more robust rig, palette swaps, and a huge desire for riding elevators.

ooohh, the colors

Before even beginning to animate, developing some good pre pro was necessary. I went through a couple of ideas of how I wanted my animation to go with some exploratory drawings. This was to explore some of the possibilities and potential of the actions I had to create. Bugsy had to start and end somewhere and he had to hit a button. It was up to me how he went about that though. Plus exploritory drawings are a good place to see if the silhoette of your character will look nice in the pose you have chosen.

After choosing the poses I preferred, or a combination of a few, I moved on to storyboards. These are made to develop a good feel for the range of action and see what kind of timing and poses you are working with in a scene. In a production it's also a good gauge of how much a scene will cost (but we'll just leave that one out for now.)

Finally we have the action breakdowns. These are made to take a better look at your motion and see what all the keys and breakdowns would look and how you would pose them. You don't want a crazy amount of breakdowns, because at that point you're just 2d animating, but you want enough to define your motion, see your principals of animation, and convey your ideas quickly to someone else.

Now it's time to get into Maya and animate. The first step is blocking out the key poses to get contacts and the basic poses of the button press and start and end. For this I made two MEL buttons to speed up the workflow, one to key most all controls and one to select the controls that are needed to move the whole body. In this stage it's really necessary to make sure the feet are not sliding so it doesn't become a problem in the future.Animating the eyes this early helps to not make your character look like a zombie and also gives you more control over how your character is displaying his emotion (I hadn't done it yet.)

After the contacts come the passing poses, followed by the ups and downs (this is all in stepped mode still.) Working on the turn is very similar to the walk, the same rules apply in terms of key poses, but the curve progression looks a bit different.

This is a good time to check your curve progression in the graph editor and maybe even switch into auto tangent to test your walk. I had a few issues where my extreme poses were in the wrong keys so it took a bit of extra time to go back and fix it. Also while in auto check for any popping in the legs
on the in-betweens and fix those appropriately (This can be a result of having an up pose too high.)

The moving holds can be implemented in the stepped phase or auto. I did a little of both. This is also a good place to key the blinks and the mouth to further push the emotion. Secondary action like antenna bobs and hand sways will also lend to the animation, but above all get the main motion down.

I want to get back to the eyes for a moment because they are really important to the character. I put in a few extra keys for the blinks to have more control of their speed. Sometimes it's also necessary to pose the eyes in what seems to be an unnatural position in perspective, but they still look good in the camera. This is so the audience can properly read the eyes and there is not too much white space, giving a zombie like effect.

not very readable

better

 After all the basic requirements were met, I added a bit more movement to the elevator to put some personality into the scene.

Soft Bodies featuring Chowder

For my final visual effects project I had to create a simulation with soft bodies. The project requirements were very open ended, and many great ideas came out of that. I decided to go with a Chowder theme because our professor warned us to pay attention to swimming textures in maya (3d textures that do not animate with your geometry) and fix them as they happen. This made me want to implement them into my project intentionally, and Chowder was the only show I knew of that used them extensively.

I never understood swimming textures as a child

Back to the soft bodies though. Soft bodies can be used for many different effects, but they are great for secondary action, being that my interests are rooted in animation, that seemed like a good fit. I started with a basic chowder model I found online, which was ripped from the game Cartoon Network Punch Time Explosion XL (great game by the way, very under rated.) And it goes without saying, credit to the original creator. The model I used was actually a combination of two, there was the shirtless Chowder and the hat came from a regular Chowder.

Link to the Chowder model

I rigged the model myself but it wasn't really anything special (don't tell anyone, but the rig didn't even have control curves.) But it got the job done for what I needed and baring time constraints. It's all about putting out the best quality by the deadline, and the rig is not something the audience sees. I may go back to it in the future and retouch it for some practice. Other than the joints there was also some skin binding and weight painting (but I'll cover those in a rigging specific update.)

So one way you get soft bodies to act the way you want is by setting goals of them. By making a copy of Chowder's geometry and making that a soft body, the original can be a goal, and the soft body will attract to that goal. The amount of attraction is determined by the goal weight and can be refined by painting goal weights all over the model to get varying degrees of motion. You can probably tell where the goal weight was high and where it was less. The less the attraction the more the freedom to move.

That's good and all but how about when chowder smacks his belly? Is there some type of collision in the arms? After all soft bodies are still particles and can react in many different ways. What I actually used was a radial field that was animated to push the particles the moment Chowder hits his belly. There was also a turbulence field active throughout the whole scene to give the belly some randomness. I also had a subtle gravity field on the hat to effect it's motion.

Once I had the simulation done, I added the other two models of Panini and Kimchi for a bit more personality.

Panini Model

Kimchi Model

Then it all fell apart...

From what I understand my particles and my material somehow got disconnected, so every time I would close my scene and reopen, Chowder's geometry would disappear. Reapplying the material only resulted in a lower resolution version of the original model not suitable for rendering. The fix for this was to just work without closing Maya which worked until render time came along. When I went to batch render the model would still not show up, despite it showing up just fine in the single frame render. What I concluded from some online searching was that particle caching or geo caching was the answer, I tried both with no success. You may see where this is going. I ended up single frame rendering each frame to get the desired look, this was 240 frames and it took around two hours, plus one for troubleshooting the batch render.

Batch Render Result

Single Frame Render Result

Despite all the issues, I did learn something new, while meeting the project requirements, and hitting that deadline. But I did need to drop a few features like the swimming texture and making Kimchi a particle cloud. I await your feedback, until next time.

Steel Mill Particle Simulation

For this project I really wanted to give a different vibe than my first visual effects project, my first project was bright, colorful and happy. I wanted to make this one a bit darker and even unsettling. I think experimentation in different styles is very important even in CG.

The steel mill scene was provided to us along with all the geometry and lighting. We were responsible for the particle effects and the overall entertainment value. I learned a lot about particle simulation with this project.

Particle emitters produce the particles but only affect them until they leave the emitter, after that they react based on the world around them. This can mean a variety of different things: A gravity field will cause them to fall on the axis you set the gravity to, setting the particles to collide with geometry will do just that (based on the settings you can have them stick on a surface or bounce off of it, or react in other ways), or they can just go on forever if you have nothing affect them.

Point particles with a gravity field

Point particles set to collide with geometry

Point particles with no additional fields or settings

I tried many different effects for this project. The fire was multistreak, the water and lava was blobby surface, the smoke was smoke, and the sparks were tube. My favorite effect though was the blobby surface, there is just so much potential with those particles and I love how they clump together based on their proximity.

The thing about all of these different particles though is that some require maya hardware rendering and some require a software render, like maya software or mental ray. So rendering in passes is almost a requirement, and certainly recommended if you want to alter the effects individualy in a compositing package like Nuke or After Effects (I used After Effects for this project.)

Count on particle simulation renders to take a while, especially with mental ray, rendering this all in passes took around 28 hours.

Why so Spoooky?
Like I mentioned earlier, I really wanted to change styles from my first project but for the longest time I had no inspiration as what to do (I had a simple pan of the whole scene with no real character in the camera motion.) Late one night I decided to make the camera first person to give the effect of actually being there, and then I realized how scary the steel mill actually is and I thought how it would feel to be there alone after hours while something was watching you. That really got me thinking about the sound design and what would actually be the payoff at the end. I decided to go with the Bugsy rig I was using in my animation class because hyper-extending the jaw is really creepy. Hope you can sleep easy after this and I'll see you next time.

Strong Posing with Bugsy

So in my last animation update I mentioned the importance of keeping a character balanced. This holds very true when you are trying to convey a certain emotion. That's not to say you can't break the rules, but you have to know them first. But realistic weighting is only one aspect of a character and this project. Another is silhouette and negative space. You're well on your way to a successful pose once you've developed a nice readable silhouette.

So what this project was exactly and what the requirements were was to make seven clean readable poses using this character named Bugsy (we'll be seeing more of him soon!)

Try to guess the emotions I was going for, the last one's a freebie since it's not really an emotion, it's a balanced on one leg pose.

The Bugsy rig wasn't to hard to work with, which was nice because it gave me time to focus on the requirements of negative space from the fingers to the antennas and even develop a strong line of action (which I'm sure will come up again in the future.)

The unsung hero of this rig was the jaw in my opinion, though I didn't use it in every pose, it really helped in some poses to nail in that last layer of character. The cylinder prop was also really helpful to back up some of the poses and assist in their line of action.

Overall this project really got me thinking of all sorts of animation and staging concepts that at first glance may not seem to have a huge impact. See you soon!

Intro to Rigid Bodies with the PowerPuff Girls

Visual effects in Maya is a vast domain of many possibilities, but rigid bodies seems like a great place to start. Active ridged bodies act on each other and react based on many different conditions (mass, gravity, inertia, etc). They are also being dynamically simulated every time you hit play. So even though it may not look like it, they act differently every time, at least until you bake the simulation. Passive rigid bodies do not react as much as active ridged bodies, but they can serve more as walls and containers while still affecting active rigid bodies. They can also accept any keys you have placed on them which active rigid bodies do not do.

Another big part of visual effects that not only relates to ridged bodies is fields. Fields come in many different forms, like gravity, turbulence, newton, and others. A big one that helped in this project was the gravity field because I was trying to simulate the effect of a real life rube goldburg machine with parts that react as they do in the real world.

In addition to rigid bodies and fields, constraints were another big part of bringing this project to life. Constraints hold back geometry to a place of your choosing. The constraint that helped me the most was the hinge constraint. How it works is just like a door does, wherever you put the constraint, the geometry will rotate around that axis, just like a door, all while dynamically reacting to your rigid bodies. You can see these in action on the hammer right at the beginning or any of the doors in the scene.

I also learned a few new things and refreshed some old skills not directly related to dynamics by doing this project.

1) Visibility swaps are your friend. This goes hand in hand with only worry about the camera. What both of these basically mean is that if the audience can't see it, it's fair game. Many things go wrong in simulation but as long as it can be hidden or covered, its A-ok.
2) UVing and Texturing. Though it didn't progress the simulation at all, giving some color to the scene can really help bring it to life, and I think everyone is tired of seeing the default lambert at some point.
3) Moving Textures. This made me so happy because now I just want to use them everywhere. Like having a scene where someone is watching tv but the screen material is a moving texture and no longer needs to be added in compositing. Makes me happy!

Why PowerPuff Girls?
To be honest it was an on the spot idea. I knew PPG had a lot of heros and villains to choose from and I wanted to make something unique for every piece of the rube goldburg machine. I also wanted to tell a story which the sound really helped with. It's like a journey through an episode. Plus I just really like the show. Really hyped for the revival in 2016! Who's with me? One thing you may have noticed was the emergency phone towards the end of the clip. That was just a quick model I made in about thirty minutes using the basic tools as well as a lattice on the chord to curve it a bit. The texture was one that I found online but touched it up for what I needed.

After rendering in Maya, I took the frames to after effects to add the sound and get the single movie file.

Overall a really fun first step into visual effects and something I would encourage everyone to try if it sounds interesting to you! Oh and don't forget to bake your simulation, you'll thank me later :)

Walking with Eggy

This project was to make Eggy (the egg with legs) walk from his starting position to the scale, turn to face the scale, and then jump onto it.
Eggy was a nice character to work with because it allowed me to focus on the legs and make sure the animation on those is done right. Learning the legs is important before tackling a whole body that applies for 2d as well as 3d.

Back to Eggy though, his rig was easy to work with and the feet in particular had some nice controls to streamline the process. Having controls like Lift Toe and Lift Heel made it so I didn't have to rotate the foot for every little movement, making the check for sliding much more efficient.

This project was also really helpful for learning the appropriate curve progression in the graph editor for a standard walk cycle, how certain keys need to favor the extremes, and how working in stepped is a good way to block out your poses before moving on to secondary actions and the foot roll.

The turn and jump were also a unique hurdle in it of themselves but they still required the same basics in terms of keeping a steady and natural curve progression and thinking in arcs. Having the character looking realistically balanced is another big part of bringing them to life. But I'll cover that in my next animation update, so stay tuned.