LogInThe legendary artist and designer Chris Jones has joined 80 Level to talk about how the CG industry has – or rather, hasn't – changed over the past decade, discuss his hyperrealistic human model, explain why he heavily dislikes generative AI, and share why he doesn't see himself as "secretive."
Please introduce yourself to those who somehow still don't know you. What have you been working on recently?
Chris Jones: I'm Chris Jones, 3D Generalist, Illustrator, Designer, Composer, etc. I'm best known for my digital human creations, as well as my 2006 animated short The Passenger, and to a lesser extent, 2018's String Theory (which YouTube's algorithm immediately swept under the rug). I've also worked as an artist on video games, TV shows, commercials, and children's books.
I'm currently taking a break from working on the human project after having burned myself out a bit while adding a musculoskeletal system, and am using the opportunity to finish off some music that has been waiting in the wings for the past couple of decades. I also plan to dust off some film ideas for which the human was originally intended, which will hopefully also serve as a more rigorous testing ground for it than the piecemeal animations I've been making to date.
When did your artistic journey begin? What initially inspired you to pursue 3D animation, and where do you draw inspiration from these days?
Chris: I don't remember there ever being a beginning, so I must have started sometime in between being born and having developed a long-term memory in the early to mid-1970s. Around '79 or '80, I saw a TV show about computer graphics, and was gobsmacked by what was essentially a rotating cube (which I more recently discovered was from the short "Sunstone"). That's when the seed was planted, but it wasn't until 1994 that I was finally able to get my hands on the technology when I discovered my university had copies of 3D Studio 3 (the precursor to 3ds Max). The following year, I bought a PC and LightWave 3D 4.0. I made my first attempt at modelling a human head shortly thereafter, and have been trying to get it right ever since!
Realistic humans have always been considered the "holy grail" of CG, so naturally, that made it an attractive proposition for someone as ambitious as myself. The initial incentive for my current incarnation of it, however, was to have a reusable generic humanoid for use in my own projects, so I wouldn't have to go through the ordeal of making one from scratch every time. It was supposed to be a side project that I could tinker with in between projects, but then it became the only project. No one had managed to pull off a completely convincing close-up of an animated human before, and the further I progressed, the more I was drawn by the allure of exploring this uncharted territory.
In my travels, I would occasionally come across examples of photorealistic still portraits, but invariably they would be sculpted into a pose and textured to be viewed from a certain angle, and if they were animated, the realism would break. Similarly, digital doubles that were made with animation in mind were inherently compromised in some way, such that they would appear uncanny even in a still render. I was inspired to try and achieve the same level of realism that was present in those portraits, but in a poseable rig; and then if I could animate it with the same attention to detail that was put into the model, textures and everything else, it could potentially be indiscernible from the real thing.
Within the CG community, some consider you to be one of the most secretive creators – you stay in the shadows, unveil a new masterpiece, then disappear once again for many months. My question is, do you create the aura of secrecy deliberately, or is it just a way you prefer to do things, and there's nothing to it?
Chris: I think that tends to be the impression of those who just follow my YouTube channel, as I only publish my most polished and finished videos there. Many don't seem to realise how hard it is to make those things... most of them are only 30-50 seconds long, but they take literally months to create. The longer ones take years. That's full-time manual labour, mind you, and it can't be remedied by throwing more rendering hardware at it, as some have suggested. And when you factor in another chunk of time for income-earning activities and other distractions, there's inevitably going to be a lot of radio silence between uploads.
I don't completely disappear into the aether, though; I regularly post incremental updates on the Blender Artists forums and Twitter, although I'm not as chronically active as has become expected in an age where broadcasting your every passing thought has become the norm. I prefer to spend my time being creative rather than trying to increase my online presence, and when I do post anything, I try to maintain a high level of quality and discretion (since it will be there for all to see, forever), and that also tends to drain a lot of time and energy.
How would you say the digital art industry has changed over the past decade? In what ways has it improved, and on the contrary, in what ways has it gone worse?
Chris: From the perspective of someone who confines themselves to personal projects for years at a time, and is consequently somewhat disconnected from the "industry" as such, 10 years isn't all that long… and for the most part, the changes have been relatively minimal, I'd say. In fact, the changes over the past 20 years have been fairly trivial compared to the 20 that preceded them. As a point of illustration, up until a couple of years ago, I was still using a PC from 2008, and the machine that replaced it is orders of magnitude more powerful, and I was finally able to leverage features in software versions that I couldn't run before. This didn't translate into orders of magnitude faster output (except when it comes to rendering, but as mentioned, that isn't the main bottleneck most of the time). Rather, it allows more headroom to raise the quality bar a bit, while making things more comfortable to work with.
For example, I'm now able to load all the layers of all the face textures into Krita at once, whereas they previously had to be split across multiple files. The workflow remains much the same otherwise, and I was soon back to maxing out the RAM just like before, but the convenience means I can spend less time juggling files and more time painting textures. Files load and save faster, and I can work with models at higher subdivisions and render previews in near-real time, all of which appreciably increases my productivity. But instead of producing more "content", this manifests in the form of more refinements and higher quality renders.
In that regard, the technological advancements have obviously been an improvement. Generative AI, on the other hand, has clearly been a considerable change for the worse on multiple levels. And the continued focus on quantity over quality and the extreme aversion to risk taking (particularly in the entertainment industry) have certainly been negatives from both a consumer and prospective employee standpoint – but could also be seen as positives for independent creators, since they make it easier to stand out and compete.
And what about the tools, what can you tell us about the capabilities of 3D software then, compared to now? Are there things we can do today that weren't possible 10 years ago?
Chris: I can only speak in terms of my transition from LightWave to Blender, as those are the only 3D tools I've used in the past decade. And I can't comment on LightWave's evolution, as I was only using the one version of it throughout the four-year production of String Theory. I had already been testing Blender prior to that, and was waiting for its render engine to catch up to LightWave's before exploring further. This had transpired by the time I finished String Theory, and within a few months of learning Blender, I had sailed past virtually every obstacle I had been bumping into in LightWave. I was able to resurrect the otherwise stalled human project and reconstruct the basis of the rig in a fraction of the time, and push forward with ideas that had previously been out of reach.
My Hand Test II video exemplifies this. For the original Hand Test, I could only crudely emulate the tendons in LightWave by chaining together and elongating a series of spherical effector nulls, which deform the skin surface. For the sequel, however, I was able to actually model the underlying anatomy in Blender and then shrinkwrap the skin onto it.
Conversely, I've been unable to achieve Tension Maps quite as successfully as in LightWave, partially due to my inexperience with Geometry Nodes, but also because Blender's hair system doesn't react to node-based deformations in the Shader Editor (i.e. when the skin wrinkles, it passes through the hair). This issue is more pertinent to the face however, and dynamic wrinkles weren't implemented in the original Hand Test video.
What have been the standout moments of your artistic journey? Which year would you call the most challenging, and which one the most successful?
Chris: I completed String Theory in 2018, and in 2019, I fully moved over to Blender, so those were both notable years. Career-wise, 2020 stands out as my most successful year, as that's when I started selling my Universal Human assets, and I no longer needed to rely on freelancing as my primary source of income.
2022 was already a candidate for the most challenging year thanks to a variety of unwelcome circumstances, but those were eclipsed by the uncertainty posed by gen AI and its threat to undermine everything I had been working towards. It's been heartening to see the pushback from the art community in the intervening years, though, which feels as though it's reaching some sort of critical mass... and awareness seems to be trickling into the wider community as well.
Finally, what would you wish for the community in 2025 and beyond?
Chris: I hope for a fair and prosperous future for (actual) creators to look forward to. Failing that, I wish you all a merry apocalypse and a happy new world order!