AnimBreak challenge “Respawn” process (June 2017)

 My goal is to share my general workflow and thought process of taking a scene from an idea to a finished piece*. Brief postmortem at the end.

*Each Animbreak challenge is only ~3 weeks long and I ran out of time to completely polish this shot, everything is still relevant.

TLDR version

-Get an idea. Inspiration is everywhere! (Animator's Resource Kit)

-Gather resources. Rigs, props, scenery, etc.

-Layout the scene

-Blocking pass with cubes in place of characters

-Rough blocking with rigs

-Refine poses (personal workflow step)

-Blocking Plus

-Feedback

-Spline

-Feedback

-Polish

-Feedback

-Breathe, relax, celebrate

-Reflect

Coming up with an idea

This is my favorite part of animating. It's when the possibilities are endless and I feel like I can take on the world! (And usually the part that gets me the most in trouble)

First brain storm ideas-

I'll write down any ideas I have, even if I don't think I'll follow through with them. It's good to get the thought process going (sometimes starting with nothing can be intimidating) plus you can combine elements of sub-par ideas to make a great one.

If I'm having trouble coming up with ideas-

I like to look at rigs that are available. Finding a unique character can get you in a different frame of mind. A crocodile will act different than a robot... Or perhaps you can make a crocodile act like a robot. (Animator's Resource Kit is a good place to start and great resource in general)

Look at reference-

This can be anything! Movies, books, toys, objects, life experiences, music. Ideas are everywhere.

   

Context-

What emotion are you aiming for? Serious, funny, somber, ecstatic. Think of a verb or adverb and try to fit your idea into that.

Are there skills you'd like to practice?

Single character, multiple characters interacting, types of props, different environments, quadrupeds.

For this challenge nothing came immediately to mind like the previous “Happy” and “Death” challenges. It's not something that is universal to all games so I did some research. I've played and enjoyed Sunset Overdrive and it utilizes respawns to add character to the over-the-top world. Although the animations were great, they didn't inspire any unique ideas.

Outside of these challenges I keep a short list of animations I really want to do eventually. Whether I think it's a unique idea or something I'd have fun animating, I keep track of it. After a bit of brainstorming I fell back to this list. I've wanted to do a walk cycle of someone inside a hamster ball and I decided to turn that into a respawn!

Once I have an idea

I'll make a list of things I need in the scene such as props, rigs, and set pieces.

Think of how you're going to use a prop so you can plan if you need to worry about constraints, rigging if it's a more complex prop, or using deformers to bypass the need for a rig. Test a rig's motion and features to make sure it fits your needs. Scaling, FK/IK switching, foot roll, etc. Not all rigs are made the same and it can be a bummer to realize halfway through animating that a rig can't accommodate an awesome action sequence you had in mind.

If you're not into modeling look for models online, there are a ton of free resources. Be sure to credit artists!

Can't find a model you want? Make your own! I am by no means a professional 3d modeler, but I greatly enjoy modeling props for my scenes. It doesn't need to be a masterpiece, you should focus on the animation. The point of a prop is to add to the scene and not be the main focus.

Can't find the perfect rig you want? After some practice you can make some changes to a rig yourself. It can be as simple as overriding a texture or adding blendshapes. If you're not comfortable with rigging, don't sweat it, your focus is on animation. I always advocate learning, at least the basics, of the entire character pipeline!

Next step: Layout the scene, import props/scenery, setup camera

The first thing I like to do is setup the scene by bringing in all the props, scenery, and rigs I need. Making sure everything is all to scale and in place.

Next I'll do an animation pass with cubes in place of the characters. This is to get the general timing and layout of the characters' positions without spending time on posing the characters. This will help to give you an idea of camera positioning as well.

In this phase I have an 80-90% idea of what I'd like my animation to be. My animation process is more organic, I don't have every single pose or action in my head before I start.

Rough pass with rigs

At this point I like to roughly pose the characters by matching the layout blocks. I'll focus on general poses/silhouettes and only using a few main controls. Hips, legs, arms, and spine.

This could be considered an extra step, but it helps me in a few ways.

1. I'm motivated by results so when I get an idea I like to get it into Maya as quickly as possible.

2. If I missed any parts of my planning this step helps me realize that before I get too far.

Workflow is never set in stone and should evolve from shot to shot. For this particular scene I knew the ball would be a big focus so I animated it first and kept the keys splined from the get-go. The animation on the ball was not final, but it really helped getting a feel for it's general movement and rotation as I went to block the characters.

Between this step and the next step is when I more than likely nail down my idea completely. This is the easiest phase for thinking of poses, re-timing, and scrapping parts that aren't working. Don't get too attached to an idea or action, sometimes it doesn't play out like it did in your head and that's okay!

Refine poses

After I have rough blocking, getting the characters to match the layout cubes with weak posing, I'll go in and refine the poses.

I stay in stepped keys but focus the hero poses and cheat them towards the camera. I used to do the basics (body, legs, arms) and leave smaller things (fingers, face) for much later. Depending on the rig, fingers often have as many controls as the entire body. You don't need fingers to get the idea of a pose across and I didn't like spending so much time on fingers when I could achieve more readable results by posing the larger controls.

I do, however, now pose fingers before I move out of blocking. Animation doesn't need to be rushed and taking the time to come up with basic finger poses early will make the process go faster later.

I may pose a general facial expression if it's important to the scene, but I'll leave the face and lipsync (if I have any) for it's own pass after the body is nailed down. A lot of communication is achieved through body language and you don't want to rely on the face to get an emotion across.

Feedback!

It can be hard to show other people a work in progress, but it's important to get feedback early and often. It's also important to get non-animators to look at a scene (usually after you're out of stepped keys) as they see it from a different perspective.

Blocking plus

I use this phase to further define actions with in-betweens and refine timing. Depending on the action taking place I may change the tangents here. If it's a slow scene I'll stay in stepped keys, but if there's a fast action or multiple characters interacting, such as this scene, I'll jump into linear keys.

It's easier for me to keep track of how things move and are timed when the motion is fluid. I use linear keys because spline can make things too soupy and if your animation looks great in linear keys then moving to spline will make it that much better.

For this shot I had the bird and imp characters in blocking plus before I started the alien. I knew the alien wouldn't be controlling much of the motion overall so I left him until later. When I did get to him I animated him straight ahead instead of pose to pose. There was a lot of fast movement and he was being affected by outside forces so I found this the best solution.

Spline

I used to struggle with this phase because when you're in stepped mode your brain fills in a lot of timing and motion that isn't there. If you do enough quality work in the Blocking Plus phase converting the curves to spline shouldn't drastically change your animation (as stated above).

After I spline I'll take the time to make sure the curves aren't doing anything crazy in the graph editor such as overshoots or other jankiness.

This is where I start to look at arcs in more detail. I still try to stay on the keys I made in the previous phase here in case I want to make final adjustments to the timing. The fewer the keys the easier it is to adjust.

Polish

This part is fun for me because I don't need to worry about key cleanliness. Everything should be pretty solid at this point so animating on 1's is what I look forward to. I use this phase to really pay attention to arcs, fingers, and secondary controls such as hair, cloth, belly fat, etc.

That's just about it! At this point more feedback doesn't hurt, but keep in mind that you'll be looking for different notes than in earlier phases. Also know that you don't need to take or implement everyone's feedback. People have different tastes in style and acting, do what feels right to you.

I always animate for myself. I do ideas I think are fun, exciting, or challenging to me. When you try to please other people you may not be as motivated when it matters the most and do a half-hearted job.

Reflect

I enjoy doing a postmortem of each scene I complete. It helps cement new skills learned, new processes practiced, and discovered. You don't need to write it down, but after spending so much time on something be sure to gain something from it!

Lessons Learned

1. Planning for multiple characters

I've worked with multiple characters before, but not in a way in which they all affected each other (and props) at the same time. Usually you want to block out the character that is driving the scene first. In this case I chose the bird character as he was driving the ball and I wanted imp character to be more ancillary at the beginning.

The “use a cube to block” technique is good to get general timing, but I underestimated all the interaction and time I wanted to hold things to read. I ended up re-timing some parts multiple times, it wasn't arduous, just annoying.

Things that went right

1. Deformers

This is the first time I've used multiple deformers in a shot. It saved time and was a lot of fun! If you need a simple deformation you can skip the hassle of rigging an object and use some flexible tools in Maya. I used a flare deformer for the blue spawn tube at the beginning, a wire deformer to create the conveyor belt, and a squash deformer for the spring.

Credits:

Tribe Rigs – Animation Mentor

Too Rig- Boutique 23

Low poly assets (outside nature) – Twitter: @Quaternius