PackDev's Founder Miloš Baskić told us about the studio's history, its production pipeline, and the financial advantages of selling 3D environment assets.
Introduction
PackDev is the Unreal Engine Marketplace – and now Fab – native company. What do I mean by this? We started as a seller on the UE Marketplace in 2019. We received an outstanding amount of support from the Epic team, and this allowed us to focus all of our efforts on the Marketplace. This meant that for the first two years of our company, we could sustain our seven-person team, exclusively via sales there. Since then, we have been able to grow to a 15-person team and spread out as a company, selling on other platforms, developing our own games, and offering our services to other companies (most notably Cathedral Studios, co-developing The Bornless, an upcoming game which will also be on Epic's platform).
Our main office is in Kragujevac, Serbia, although we have members from Belgrade, Novi Sad, Glasgow, and Warsaw. I have attached a picture from one of our team meetings below:
Experience With Unreal Engine
Many members of our team have been in the industry for many years now, and I myself have just recently hit the 10-year mark. Through the years, we have used many of the industry standard tools. Even when the company was formed, we were working across a wide range of tools (ZBrush, Maya, Blender, Substance 3D, Houdini, Unity, Unreal Engine, etc.), however, we found ourselves quickly converging on a few tools and workflows.
Namely, I’m talking about Unreal, Substance 3D Painter, and Blender. We found that the quality and accessibility of these tools allowed us to quickly onboard new members, and keep costs reasonable. Furthermore, with the advancements of the Unreal Engine modeling tools, the need to go into Blender is diminishing every day. Specifically, there have been several examples in recent weeks where I have been able to: sculpt additional details, retool/decimate meshes, correct pivot locations, correct in-engine scale, procedurally texture assets using material graphs, procedurally scatter small meshes on larger meshes, create LODs, etc., and all that without even leaving Unreal.
Many of these features have been in the engine for a while now, but their amount, reliability, and overall quality/usability have increased enough in recent times, for them to cover all aspects of my workflow, making it more convenient to just work in the engine rather than going between two programs. With a backlog of approximately 400 packs, we have more than 4000 models/assets. So every little optimization and reduction of friction in the workflow is a lifesaver.
Included here are pictures of one of our recent packs (Hydra Lair). This pack is a textbook example of the versatility of Unreal Engine. Our Prop Artist Andrija Nikolić was able to create a very bare-bones collection of props, and from those simple props, Level Artist Filip Karzić could assemble a breathtaking scene with immense complexity.
Without the ability to mix, match, adjust, and optimize these assets directly in-engine, this pack would have required hundreds of custom props.
The Wild West Series
A great deal of our decisions are made based on the statistics we get from the Fab backend. We look at our production costs, customer satisfaction, price per unit, and overall market needs in order to decide which packs to prioritize. Our first successful implementation of the idea was the Rocket Launch collection. In that case, we saw a significant spike in popularity with our first pack with that theme. Using our existing art, we decided to double down and produce a few packs covering the entire niche of real-life-inspired rocket launch scenarios. We stopped this collection when it seemed that the market niche was saturated.
Likewise, with the Wild West collection, based on our sales data we concluded that it was a theme that had potential and that with our existing backlog, we could produce a wide range of asset packs.
As mentioned in the previous paragraph, with our theme identified, we set out to cover all niches, ensuring that the theme is well-saturated once we publish our packs. The bakery, for example, represents a good intersection of topics and highlights how we aim to cover a wide range of games. Of course, a Wild West game might need the main street for a shootout at noon, but we wanted to make sure that a game focusing on the daily life of a Wild West town could also fully rely on our pack, including places of work, entertainment, worship, and so forth.
Many of the same tools/methods also apply to the previous example. The bakery was created using many pre-existing elements (like modular walls, filler props, etc.) and some custom-made props too, which might not be found in any other context. Once again, Unreal was instrumental in the workflow. With our basic props created in Blender and textured in Substance 3D Painter, all other work was done in Unreal Engine.
As an example, the planks were assembled into walls, merged, LODs were generated, and additional UVs were not necessary because we use Lumen. To give an extra level of cohesion, a master material was used, which applied grime and dust across the scene.
Lastly, an interesting quirk of our pipeline is that we find Unreal so convenient, we make all our packs there, and then use an in-house utility to convert the packs to Unity. From there, our Tech Artist Luka Aleksijević touches up the materials, lighting, and so on. A good example of why we do this is LODs. Auto-generating LODs in Unity is a pain and an involved process. So with one click in Unreal Engine's Property Matrix, we generate LODs and then migrate these assets into Unity.
I am including a screenshot of our in-house tool for the engine conversion. The UI is somewhat basic, but I figured you might find it interesting:
Advice on Selling Assets
For anyone considering selling assets, I would strongly recommend Fab. In fact, if anyone reading this is on the fence, they should finish reading this article and sell their work immediately afterward. The process is so frictionless that it is by far one of the best hourly investments an artist/developer can make.
Our marketplace originally started with five packs I had published, consisting of some old personal projects I had lying around. These personal projects were uploaded over the course of a weekend, and over the years, they have earned thousands of dollars. In terms of time spent, compared to potential returns, this is by far the best use of your time.
I can’t give a perfect example of what publishers should focus on, so I will say this: starting is the best way to figure out what to focus on. If you publish three packs, and one performs better than the others, then that is where your focus should go.
Lastly, as it relates to all developers and artists, working in any engine, I would still highly recommend selling your assets when possible.
In the past, we used this to offset our own game development costs by selling elements of the game as packs, and even offsetting costs for our clients! This is possible in cases where the client does not mind us re-selling some of their more generic assets (furniture, vegetation, architecture, etc.) This ability to offset costs of production has made us very competitive and allowed us to implement projects that would otherwise not be feasible.