With this month’s blog post I wanted to highlight an absolutely crucial part of the process; Sometimes animations just don’t work out. It happens all the time, and it’s totally fine! It’s actually a good thing! After all, where’s the learning without the failure?

There are a million valid reasons not to cross the finish line. It might be because of time constraints or maybe your shot exceeds your current skill level. Sometimes it’s just time to move on to something new, or completely restart with a new approach! I talked a little bit about this in my post on creating an idle animation. We can noodle our animations into infinity, but that has diminishing returns. Eventually we stop learning, and start obsessing. And that’s no help to anyone, especially not to you!

For every shot that I’ve posted here, there’s MANY more that haven’t seen the light of day. But today, let’s change that! Every so often I want to highlight an animation or two that I’ve never finished. Posting completed work is all well and good, but I hope there’s something helpful about posting uncompleted work too. So let’s give it a try!

Before I started my Juice Box Heist animation, I was really struggling with character and object interaction. The year was 2021 and I wanted to create a simple animation where a character grabs a small object. Simple enough, right!? Maybe not so much! At the time, I was still learning (and mostly misunderstanding) Maya’s constraint system. And so creating motion that had an object cycle between being constrained and unconstrained to a character’s hand was a slight challenge. With that in mind, here was my first attempt:

Okay, maybe not too bad! But I never ended up finishing it. There’s a couple reasons. The first being the constrain system I was using. Instead of creating a system of constraints that uses locators to take some of the translation data (allowing you to still animate the object while it’s constrained), I just constrained HealthBot’s main control to the main control of the hand. This is fine for this animation, since the object doesn’t really shift around in the character’s hand. But I already had plans to change the motion a bit. If I wanted to have the character to fiddle and rotate the object in their hand, I’d be out of luck. The parent constrain won’t allow you to translate anything under the direct influence of the constraint itself. It just follows the translation of the hand control.

I was also struggling a bit with the repetitive motion that the character does midway through the sequence. They start repeatedly pressing the button out of frustration before slamming the robot back onto the table. This reads well when it’s in this simple blocking stage. However, when I started splining this animation, the motion became really difficult to read. The computer does a little too good of a job interpolating between the frames. Without enough keyframes to give it more structure, things felt floaty and mushy. Nowadays, if things look a bit floaty in spline, I know I need to go back into blocking. I’ll add some key poses with more progressive spacing to remedy the problem. But I didn’t quite know that at the time, and I felt a bit overwhelmed.

So I decided to restart!

This time, I fixed up the constraint system. I made sure to use locators, not just parenting one control to another. Here was the second try:

I’d say that this second attempt has a bit more appeal than the first. The object not only gets picked up by the character, but it also moves around while it’s in their hand! AND there’s a bit more interaction at the start; The character brushes away an apple to get to the bottle. An overall improvement! But you might notice that it’s still in blocking! Not quite as smooth as a finished shot.

Well my ambition got the best of me here, and that additional movement of the bottle was my undoing. Having it spin around in their hand looked good in blocking. But it was a total mess when I brought this motion into spline. And try as I might at the time, I just couldn’t get it to look as consistent as it did in blocking. So I had to move on.

The blocking-to-spline pipeline was a re-occurring issue with many of my earlier animations, this one included. I just wasn’t quite sure what keyframes were needed to keep things looking decent after interpolation. Thankfully, I have a much better understanding of progressive spacing these days. This is especially true when creating key poses, but you can read more about that in my latest posts.

Well there it is! A small tasting of shots that I never finished. These were far from disasters, but at the time I couldn’t invest any more time into getting them across the finish line. Maybe sometime I’ll try splining the second shot again, but for now I’ll leave it be. I have many many many more where these came from, but I’ll save those for future posts. Thanks so much for reading, and take care

Link Rig by Christoph (The Stoff) Schoch
Malcolm Rig by AnimSchool
Healthbot Rig by Raphael Cancellier

My Links: ArtStation  / Cara / Vimeo / YouTube / Instagram