3D Artist On Substance 3D Painter's Role In Modern Pipelines

Substance 3D Painter has become a core tool in many AAA and indie pipelines. How do you see its role evolving within modern real-time production workflows, especially alongside game engines like Unreal and Unity?

The hope and the intention are that Painter evolves alongside other tools. We as a team are all quite passionate about 3D, and we tend to follow other software releases, as well as speak to users about their needs and changing workflows. I think our focus is on trying to alleviate as many friction points as we can, so artists can concentrate on creative work.

Painter continues to expand support for formats like USD and broader ecosystem interoperability. How important is open pipeline compatibility to Adobe's long-term vision?

It is at the forefront. As I'd mentioned earlier, removing friction is the keyword here – theoretically, that's USD's main purpose. I won't say we are quite there yet; it is a complex subject to tackle, but it's something we are working on. USD is a format that was designed to store all sorts of information related to every aspect of the pipeline, so different people can work on the same file structure at the same time and have little difficulty sharing information.

On a similar note, this year, all Substance 3D apps are working on OpenPBR support. For those who are unfamiliar with the term, it is a new material and shading standard, building on PBR that's been around for over a decade now. OpenPBR would make it possible to define advanced material properties (like fuzz and iridescence). Moreover, if everyone implements OpenPBR, materials will look the same in every software of the 3D pipeline. And we plan to transfer OpenPBR information via the USD format, as it is most capable of doing so due to its flexibility. 

How does Substance 3D Painter fit into the broader Substance 3D ecosystem today (Designer, Sampler, Stager), and how should teams think about using these tools together?

Once you get the hang of nodes in Designer, the possibilities are limitless for what sort of assets you can create, and we are not just talking about materials, this can be noises, alphas, images, and you can even expose parameters that can be applied in Painter later on. I am a Painter PM, but if you ask me, Designer is probably the most important tool in the system. Painter may be the most known but imagine if we didn't have Designer, the Painter library would be empty, we would have no parametric tools like Smart materials and masks, because their key components are generators and noises, which are all made in Designer. I think we, as artists, are all so used to it that we forget that Designer is where it all comes from.

Sampler, I would say, has a similar material-making role to Designer, but it specializes in materials made via photogrammetry, or in quickly combining and modifying materials from Substance 3D Assets (which, by the way, is a big part of the ecosystem as well, albeit it often goes unmentioned). Whether you prefer using Sampler or Designer, whatever materials you make can be easily sent to Painter (yes, we do have a one-click interop button), and that's where you would apply them and work on your 3D objects. Then, if you need to render your work, we also have a convenient button to Send to Stager. The advantage of Stager is that it's included with the ecosystem and is fairly easy to understand. It won't replace something like Unreal by any means, but if you are after an easy beauty shot, it's worth giving it a go.

Needless to say, I was a Substance 3D fan (back in the days of Allegorithmic) before actually working for Substance 3D. What I still love the most is that you are given a choice – you don't have to know all of them, you can just stick to one and do everything there, but if you want to really have fun with texturing and push the limits, then the combo is definitely worth diving into.

From your perspective, what are the biggest challenges artists face today in texturing workflows, and how is Painter evolving to address them?

That's a hard question. I think it really depends on the level of expertise. For new artists who do not have a 3D background, the challenge is to get into Painter and understand how Painter and 3D work, because UVs, baking, and topology are hard for everyone, but if you are a 2D designer or self-taught and don't have a school behind you to help, everything can seem very daunting. We always try to make changes in the UI to facilitate the learning process. Something I'd been working on recently is a very basic MeetMat tutorial to show the main tools of Painter, and how to approach them if you are starting from no 3D background.

For more advanced users, the challenge is probably performance and workflow optimization. These days, projects are getting bigger, so the need for resolution and mesh density is higher. We'd recently migrated to Vulkan, a newer graphics API than OpenGL, which can boost performance for mid to large projects, and we are always asking users to report cases and share their work so we can reproduce and potentially fix other issues. As for workflow optimization, the Flatten feature, for instance, is a step in this direction.

Optimization is probably the trickiest thing to solve, because first of all, there is no one workflow that rules them all, as everyone uses Painter differently. Secondly, very often the topics are technically challenging. For example, with the rise of USD, there is a demand for material variant management, which would imply redesigning the logic of Texture Sets, but also the entire mesh import and management system, which is not something that can be solved overnight. But we do strive to evolve with these needs, although we do have a lot of design, iteration, and experimentation to do before we get somewhere. 

How do you see teams integrating flattening, like with the new layer flattening system, into existing pipelines? For example, when passing assets between Painter and other DCC tools or game engines.

Flattening layers was one of the top requests from artists for years, so I think they won't have a hard time integrating the feature into their workflow. The need to flatten layers could be purely for organization; sometimes you might paint or texture on 10 different layers, but once you are done with them, you actually don't need all that data. Another big use case performance – users' layer stacks can get very heavy, and they just want to get rid of a chunk of texturing they won't be touching again. It could also be useful to flatten a part of your texturing before reimporting the mesh if you know that the new mesh has moved and textures won't reproject correctly.

We also added a little extra to the Flatten feature, allowing to directly export of the flattened result to disk. This was a request we had from some studios (they tend to only make masks in Painter and blend between tiling materials in the engine), to be able to click on several layers and export masks directly, instead of setting up the custom templates and channels via the classic export.

Painter 12.0 introduces a revamped set of post-processing effects in the viewport, including lens flare, film grain, and improved depth of field. How important is it for artists to evaluate near-final lookdev directly inside Painter rather than relying on external renderers?

It may not be the most important functionality for intermediate or advanced artists, but a good portion of beginners who have yet to master another rendering software actually render their work in Painter. I think they would probably appreciate this feature, as it really improves what we had before (the effects are of a better quality, and there are more of them available). Besides, we had the opportunity to use an internal library of post-processing effects, which will be deployed in other Substance 3D apps, which once again comes back to your questions about ensuring interoperability and a consistent pipeline.

Do you see Painter evolving further toward a lookdev or presentation tool, or will it remain focused primarily on texturing?

We are most interested in the texturing process and adding tools for artists to boost their creative flow (but also enjoy their creative process). As a by-product of having internal teams outside of Painter that work on lookdev/presentation, we may have other opportunities, like with post-effects, to integrate their libraries and make visualization better than it is currently, but I would not say that it's Painter's main ambition.

Are there specific optimizations or backend improvements in 12.0 that artists might not immediately notice but will feel in day-to-day use?

We worked on redesigning the New Project window, because we know that beginners tend to get lost in its settings, but what I think will actually be useful to experienced users is a few small improvements the Project Configuration, specifically that we now remember the path to the mesh file across sessions, and it is possible to change settings that don't need a mesh reload. 

Where is the Substance 3D Painter team focusing next, workflow efficiency, AI-assisted tools, rendering, or something else?

We don't generally give away our roadmap, so I can't say anything too specific, but our focus is on the friction that comes before an artist can start the texturing process. Texturing is not the annoying part (for most, I'd imagine), it's actually getting there that can be a lot of back and forth. Part of it is the baking process – at GDC, we've announced an open beta, so don't hesitate to try out our new skew map painting (alongside a few other improvements).

For artists or studios not yet using Substance 3D Painter, what do you think sets it apart most in today's landscape?

I think there are several apps out there with which you can get the same final result as you would with Painter, but the difference for me, again, is the journey.

Painter makes the whole process more fun and convenient – and a big part of it is default assets. It's so easy to get a good base with a Smart Material and then swap out grunges and tweak parameters to get something unique. The recent Stylization filter is just as powerful and versatile. And the coolest part is that you can just save your setup as your own Smart Material and use it on another project. Changing your mind about what color you want a certain effect to be is not a problem when working with fills and masks. With the release of the path tool, even brush strokes aren't so static anymore. There are so many ways to work in Painter, and I feel like it gives an artist so much freedom.

We often get feedback from artists that, as far as 3D software goes, Painter is quite accessible, fairly easy to pick up and use, so I’d say the team takes some pride in that.

Is there anything else about the 12.0 update that you think deserves more attention from the game development community?

The Warp to geometry is worth mentioning here; it's a toggleable button we added to the Warp mode, the projection you get when you place a decal or a graphic on the surface of the mesh. Before you'd have to adjust warp vertices manually to a non-uniform surface (for example, a face or a car). Now you can toggle Warp to geometry, and vertices conform to the surface automatically. Manual tweaks are still possible afterwards, but the automatic mode gets you at least 80% there. It was a recurring request from the product design community. Under the hood, this feature is more complicated than it seems, and we worked in collaboration with researchers to integrate it. Fun fact: before the release, we referred to Warp to geometry as the "Super sticky mode", and it had a cheesecake icon.

Valeria Gerontopoulos, 3D Artist & Senior Product Manager

Interview conducted by David Jagneaux