We've contacted several experienced artists to find out what they think about the future of 3D production, the importance of AI algorithms, and how toolkits like Unreal Engine V will change our workflows.
A couple of months ago Epic Games shared the first look at Unreal Engine 5. The team presented a demo running on the PlayStation 5 featuring the power of GI, super realistic details, millions of triangles, advanced physics, and more.
The fifth generation of the engine will introduce at least two new core features. The first is Nanite that will let you create as much geometric detail as the eye can see. What this means is that film-quality source art comprising hundreds of millions or billions of polygons can be imported directly into Unreal Engine (anything from ZBrush sculpts to photogrammetry scans to CAD data). Nanite geometry is said to be streamed and scaled in real-time and there are no more polygon count budgets, polygon memory budgets, or draw count budgets.
The second big feature, called Lumen, is a fully dynamic global Illumination solution that immediately reacts to scene and light changes. The team states that the new system renders "diffuse interreflection with infinite bounces and indirect specular reflections in huge, detailed environments, at scales ranging from kilometers to millimeters."
Why are we talking about the demo again? The 3D industry will obviously change and the introduction of features like Nanite means that some 3D tasks like retopology will become obsolete at some point. We've decided to talk to fellow artists about potential scenarios. How will the game and 3D production change in the future? What steps will be automated? Will we rely more on AI techniques and machine learning? What do artists think about open-source tools and their potential? Another big question is whether there will be an all-in-one free solution like Unreal Engine that will feature all the needed tools. Let's discuss these questions!
Art by Inka Sipola
Inka Sipola: With the extremely fast development of new procedural tools, I see workflows becoming a lot more tech-based. It is undeniable that using software like Houdini to generate entire environments is a huge time saver, and since the industry is all about money in the end, I believe that is the future we are looking at. You will always need an Artist for creative work and storytelling though. However tedious tasks like placing modular objects, UVs, reactivity, and VFX in environments could (and should) be handled by an AI.
I absolutely love the fact that there are solid open-source tools available. As a matter of fact, it’s unlikely I would’ve ever even gotten into 3D at all if Blender didn’t exist. The cost of other software is insane for someone just starting on their own. Blender’s potential is enormous, and I love the hotkey-based, fast workflow with so many different tools to get the job done quickly in one software. All you need is an up to date team passionate about their product and constantly listening to feedback and making improvements.
I believe it’s not possible to include all the best industry-standard tools in one software, however. If there is even one software that does the job of one part of the pipeline better and faster, it’s going to be used instead.
Art by Antoine Destailleurs
Antoine Destailleurs: I’m really curious to see the games and the evolution of 3D graphics in the future. With the recent PS5 demo of Unreal Engine V, I was totally blown away by the capabilities of the engine. The presentation surely had commercial purposes, and I don’t know how a 30+ hours game could have this quality in the coming years. But still, this demo was really exciting!
I think that with this new generation of consoles we might see a lot of the 3D production pipeline being automated: the retopology, the unwrapping, the optimization... All of those things are already kind of automated in Zbrush, not game-ready, but it’s already incredible to be able to sculpt, optimize, unwrap and texture a quick 3D concept in a day or two.
I think open-source tools have a lot of potential. I’ve started learning Blender this year because it seems really promising and game companies like Ubisoft seem to be interested in it. The community is really active, and the software is regularly updated. I wouldn’t be surprised to see the game industry switching to Blender. The software offers 3D modeling, retopology, unwrap, animation, drawing, and sculpting tools for free.
This could be an all-in-one free solution for graphics production, like Unreal Engine V might become one for game development. Maybe we’ll see a new ZBrush competitor in Unreal Engine V, who knows? Still, I think companies might continue to develop their own engines to focus on what makes their identity and push it further than ever.
Finn Meinert Matthiesen: With Unreal Engine 5 showing up at the horizon, a lot of artists seem to perceive it as a game-changer in production and while I'm also usually on the skeptical side when it comes to tech demos, I admit that features like Nanite and Lumen have the potential to be revolutionary. As usual, there are risks and chances connected with such a fundamental pipeline change. Game developers and game or 3d artists are always aware that they constantly have to adapt to new technologies and crucial pipeline improvements and, of course, a lot of the expertise artists need these days (performance optimization, LODs, proper normal map workflow, etc.) might be obsolete in a few years. Even today we can see that 3D artists from the areas of arch-viz and the movie industry start to use game engines for prototyping and visualization, so while game artists and offline rendering artists were sort of separated camps a decade ago, they moved much closer together in the past years.
I can also imagine that AI and automation tools will have a huge impact on the entire production of 3D as well, at least when it comes to scenes and environments that consist of generic elements, like nature scenes for example. Procedural landscape and vegetation scattering tools are already in use these days, especially when creating the base for huge open-world environments. Combined with a strong base library (like Quixel Megascans or Speedtree) plus an optimized lighting setup from startup, I'm pretty sure that it won't be much of a challenge to come up with movie or photoreal landscape scenes in the near future. Programs like VUE have been providing offline rendering solutions for landscapes and nature scenes for years already and there are quite a few groups and companies working on projects for Unity and Unreal that focus on exactly these aspects. Epic/Quixel, Speedtree, or Substance are probably the most prominent ones to mention at this point.
Nevertheless, a lot of creative tasks will rely on 3D artists. And while more generic environments might be easier to achieve, artists will probably have to craft unique sets for architecture or meet special design requirements with the use of their own creative skills. AI recombination and scattering tools, procedural landscape generation or advanced photogrammetry workflows might greatly improve the quality in certain artistic production areas, but stylized, specific, iconic or unique geometry, like fantasy or sci-fi architecture and designs, will need to be handcrafted for quite a while I think.
Well, so much for my rough assumptions about what's going to happen at some point. Pretty sure that the future will prove me wrong in a lot of aspects like it always does somehow. But I sincerely hope that this insight was at least a bit helpful and interesting for you, so thanks a lot for reading and all the best with your creative projects.
Tarek Abdellatif: I'm afraid that the asset creation will almost be automated and that might be bad news for artists like me because you'll see many scanned assets that might influence artists' jobs. Maybe the only things that will still be there are the stylized environments and characters because you can't use scanned data to create them. Technology keeps updating every single day and machines might replace most of what we do now.
I believe open-source tools are great for all of the artists, especially for indie game developers or artists and indie studios that can't afford pricey tools, but I don't think that giving away scanned assets is a very good idea. Nowadays artists don't try to create scenes by themselves and rely too much on free assets, and I think that's not good for art because the main thing that makes me love this industry is the art we make, not just dropping assets.
Art by Craig Richards
Craig Richards: In a decade or so we could potentially be in a position where AI learning could have advanced so far that it will be able functionally to generate, set-dress, and texture entire environments - with minimal direction from its human counterparts. This sounds incredible, but what are the downsides of AI learning? Well. the most obvious to me is in the same vane as the apprehensive attitude towards automation - will this lead to job losses in the industry?
In this hypothetical future where a game studios art department is now 60-70% automated, what need is there to maintain a team of hundreds of artists? That is a dour thought for the future. But the argument could be made with time being freed up by AI learning, developers could use that extra development time, resources, and manpower to push for even larger richer environments.
I also think VR will play a big role in game development in the future - not only with game testing using the headset but even working inside a game engine in virtual. We've already seen early tests of developers using the Unreal Engine's toolset in virtual reality, potentially this could shed new light on our current way of working in game engines.
Art by Craig Richards
I can imagine from an artist's perspective how interesting it would be to set-dress or texture a room in VR - for example, if an artist wanted a room to look derelict and abandoned, they could simply hurl furniture, broken shards of wood and glass into the corner of the room and use dynamic physics to calculate how these objects position themselves in a 3d space. How practical this all would be in an office is debatable, but it would certainly provide much-needed exercise to studio developers!
I can certainly see more engines being free at the point of use - but the royalties-based payment system seems like a model that is here to stay. There will always need to monetize a game engine, even to merely assist in the maintenance and updating of that piece of software. I'm sure there will be more open-source engines that hit the scene in the years to come, which is fantastic - particularly for smaller developers out there - but will those engines ever hit the bar in a quality set by the likes of Unreal or Cryengine?
Nicolas Morel: I think real-time tools will take a real part in the movie Industry. Tools like Substance Painter are becoming more and more powerful, allowing artists to visualize complex shaders almost instantly.
Open-source tools are becoming bigger and that is what makes this industry move forward real quick. Now I am a believer that there aren’t any “must-have” tools. They are just tools and their purpose is to help people create anything they want. And each software has its ups and downs. And the good thing is that they all are becoming more accessible. Now I don’t think one software can become an all-in-one free solution. Because They all have their specificity that will make them suitable for a situation. Video games and movies don’t have the same technical challenges. You will probably never see a character in games with 200k polygons. In the movie industry, it is really common.
Art by Angel Fernandes
Angel Fernandes: I think there will be more and more work for hire, especially in Latin American countries. It is something that is growing a lot lately. There are many professionals, really capable people here. We are countries with a weak currency where the dollar weighs heavily, and for outside companies we are cheap.
I also see many people specializing in different areas. 4 or 5 years ago it was difficult to find a Material Artist for example. The idea of the Generalists has almost been abandoned.
Regarding the visual or artistic area, photogrammetry is covering a lot of ground. Nowadays a studio that wants to make a game with realistic graphics almost only has to download the Megascans library. I am not saying that photogrammetry is the solution for everything as there are things that you cannot do with them. Not to mention that if you are aiming for a more stylized aesthetic, it is not as useful there. But we've already seen realistic games made almost 100% with photogrammetry, like Star Wars Battlefront. My recommendation is to study these new workflows.
As for automation, I am not sure really. Today I think that what steals us a lot of time is unwrapping and retopology. I hope that is automated soon and leaves us more time for the funniest parts of the process.
An “all-in-one free solution” sounds like a paradise to me. I think sometimes we look at open-source tools with mistrust. Since they are not made by a large corporation, we think that they will not work, or that they will be full of bugs, or that they will be insecure. It is good that nowadays there are tools like Blender that are proving people wrong. (mental note: I have to learn how to use Blender)
Art by Paxton Klotz
Paxton Klotz: I feel like from what I’ve seen texturing is something that has the potential to be heavily automated, with more sophisticated smart masks, and through using scans. I recently saw the video for the beta of Quixel Mixer, and I’m really excited to try that out. I think that being able to use scanned textures directly like that has the potential to cut down on texturing time and the results that I’ve seen look really strong (all of this at least for the PBR pipeline, I don’t see hand-painted workflows being automated soon).
I’m a fan of open-source tools, I honestly want to dive into them more in-depth. I’m always glad when tools are widely available both to use and to learn about. I am excited about Epic’s habit of making new tools that they purchase free to use, and giving more creators the ability to access these tools and create games will lead to an increase of marginalized stories being put into the spotlight. I think there’s a lot of potential when indie games have the capability to look as good as some AAA games, as many people who play games find it difficult to look past less advanced graphics.
Art by Anton Syrvachev
Anton Syrvachev: A lot of technical mundane tasks will get automated. Manual retopology, LODing, for the most part, will go away. Not completely though as there are always some edge cases that you have to fix manually.
Cleanup of scanned data, creation of tileable materials from scans, and photos, LODing, placing small props in 3D environments, cleanup of mocap data are areas where neural networks could excel.
Looking at Blender, there is definitely a lot of potential to open-source tools, but Blender has a 25-year history of development. There is also a promising open-source game engine Godot, but I am not sure it would be able to compete with Unreal’s flexibility.
Where will all the new technologies lead us in a year or two? What's your take? Don't forget to share your thoughts in the comments and discuss how Unreal Engine V and AI will change 3D production.
Author: Arti Sergeev