Bold plans require dedication that often falls into the league of the fantastical: Few would dare to tackle those, even less succeed in it… While the end is still yet to be seen, the idea of a walking creature of a beast, that only existed on paper for so long, has slowly taken its shape; over the long months that led here.

Link in case video is broken: https://www.youtube.com/watch?v=9Cz1Y-kXMRE

On with the work

When the model was completed (have a look at this previous long blog post about the model and future plans), it was time to begin the arduous process of diving deep into the world of rigging. Truth to be said that it was frightening in a way, as the scale of the production required a more professional, and consequently a more efficient approach; which was lacking from my roster. A solid month or two was just spent on making sure every inch of the model is rigged and was working in order. While it is certainly not perfect, a lot has been learnt over the almost 6 months that took to complete a part of the animation set, seen in the video above.

The sheer volume of pieces (several hundreds) made it extremely difficult to choreograph as well as to maintain equal spirit throughout the production; as nothing else, but animating in Blender was spent days upon days. However, knowing better, and having experience in making the impossible before (see the history of producing Endomorium as an example) have applied tactics that kept my sanity in check and kept the project alive at lightning speeds.

One part of it was learning more on how to optimize repetitive tasks (e.g., copying animation keys), or how to better rig certain parts of the character (e.g., what parts should move together, should be linked as such, and be in their own frame of reference). Also, diving deeper in how to better utilize constraints and be more efficient with laying out immediate tasks and planning ahead. All the while using visualization techniques, to have a better understanding of what is moving where. All in all, by the time of writing this article, it can be said that a lot more has been learnt in making a production ready character. In an earlier video, shown below, have included a small breakdown to show what the rig actually consists of.

The flow most continue

Certainly the biggest challenge this time around was to ensure that the animations are cyclic and realistic enough. It has to be said that it was a realization on its own to face the many questions of “how to do things”. Because, essentially every moving part of the creature had to be “calculated” in the head (for a lack of better term) on how would those react by the many forces of movement and interactions; something that is still being considered to be scrapped at all (to be done manually), if an automatic way of producing secondary motion will be somehow discovered (as so far, using physics objects in Blender did not yield expected cyclic behaviors; which is a must if you wish to make cyclic animations). And on top of that, synchronization played a huge role, as every motion had to be adjusted and delayed or shifted on the timeline; of which had to be anticipated beforehand. Not entirely sure why, but as an added bonus, the “cyclic modifiers” (in Blender 2.92) not always work the way one would hope (as in, those could and would break horribly). Thus it made it the more problematic to animate something in a cycle, when the middle of a period fell right on the end/beginning of the animation; and thus requiring frantic skipping from one end to another in hopes that it would all work. Times like these would wish for a mouse that had more buttons, because one must not underestimate the power of shortcuts. This might sound ridiculous, but when you need to use a function every other second, your muscles can cramp up at an alarming pace, when you are using a software solution that is packed to the brim, and requires a lot of pressed keys to work (e.g., pressing the control, shift and some other key on the board, to call a function). Similar experience was when painting the character (with a mouse), where pen pressure was simulated by applying the same brush stroke on the same place multiple times; over and over.

Link in case video is broken: https://www.youtube.com/watch?v=FfZNqbCSc8g

Balancing the act

Although the work spanned over a great time, a lot of split decisions had to be made (reminiscent of my last large project), which affected a long chain of operations, and if not experienced enough (certainly was not), one could easily fall in the many traps of not “knowing” any better (if a branch was moved, or was late in the animation cycle, it could clip, twist or “hitch” at any moment). Only positioning, stepping back and carefully inspecting the whole picture (letting the animation play itself) can (and later did) solve this conundrum. Also, have grown more respect for professional animators, who spend months on animating a scene at a time (which are only usually a couple of seconds long).

Times like these, it really shows that game development is not always all fun and games, and can get really boring and repetitive (for outsiders especially). That being said, there can be moments, where it all comes together to one cohesive thing and begins to shine (felt elevated when the character started to move on its own in the engine). Also, some of you might have noticed that the first video no longer uses outlines; for good reason. While it may look like a good idea to use such post processing features, the reality is that if you wish to make it look good (as in 99% of the time), and avoid certain edge cases, a lot of shader wizardry must go into it. So much so that it took several tries to land on one approach (thanks to the information on the web) that would look OK (seen in the second video). However, that still did not solve the issue with foliage completely (having those in such a dense formation creates black splotches), and finding time to correct it just did not seem to justify its existence: Outlines look nice, but not on everything. And as WOW was one of the main inspiration for this art style, thought to drop the idea of lines all-together; for once and for all…

Again, for the curious, here is a trello board, where all the tiniest details can be found, and perhaps agonized over (e.g., how the jittering of joints were solved, which come from moving disconnected objects at a fast pace). And the build itself (seen in the first video) is available on Itch.

Thank you for your time. Take care.

Useful links to check out

Official page (releases, blogs, etc.): Itch.io

Official blog: Gamedev.net

Live dev board (documented): Trello

Twitter: @theaaronstory

Headstart with Ko-Fi: @theaaronstory