There are a handful of conferences that talk about the research and innovation behind the game industry. Unlike other professional subjects, there are virtually no published works like other subjects in Mathematics or Computer Science, but there are still talks and presentations to attend.
One of the biggest conferences is GDC (Game Developer’s Conference) in the USA. And I just discovered via friendly reddit commenters that they have a YouTube channel with many of their talks uploaded for free, including some very specific animation talks that I was excited to watch.
This talk specifically is about the animation style in Guilty Gear Xrd. Guilty Gear was a niche fighting game series that had always focused on 2D sprites, but this recent entry was famously the first game in the franchise to use 3D models (in Unreal Engine to boot). And yet, during normal gameplay it is almost impossible to tell that the characters are 3D, even for a stickler like me. It’s only during the more cinematic 3D camera movements that the effect is ruined, but the artistry is still impressive.
And how did they do it? The video provides interesting details, such as only animating keyframes that were hand-set by the animator. I had long assumed this was a viable effect for mimicking 2D animation, but not using interpolation is not a set feature in any 3D animation software I’ve seen. The speaker didn’t say how they did it, but at least I know now that it is possible in some form. Other techniques like having a lighting source specifically for each character, hand-editing each normal vector to improve lighting effects, clever implementation of HD black outlines, and purposefully misshaping the animation each frame to create imperfections were all things I never thought about. It’s interesting that they put that much work into the look of the game (and people say me spending months on my 2D characters was insane), and mostly informative. While not perfect, this stands today as one of the best examples of 3D models to mimic 2D art in 3D games.
In addition to this, there was a video about the animation behind Skullgirls, which isn’t too innovative but is especially informative about the implementation of important animation techniques.
Also about the art behind Ori and the Blind Forest, which provided a lot of in-depth information about how their animation, modeling and lighting was achieved (in Unity3D).
And many other videos, including one that cites that over 55,000 indie publishers are operating today. And I assume even more developers, even if most of them haven’t finished a game yet. Still a scary number… I wonder how many filmmakers or musicians there are by comparison?
I also found ACM’s Siggraph, a bigger and more established computer graphics research group and conference, also uploads many videos on YouTube. It’s not as well shot, but I could easily spend weeks studying the videos from these two sources. If you haven’t looked into these YouTube channels yet, I highly recommend you check them out.