Dardo Studios 的绘画风动画短片 Ghost Stride 如何完全在 Unreal Engine 中绑定与动画

Interviewed by

13 October 2026

Dardo Studios explains how Control Rig, Sequencer, Lumen, projected paintovers, and hand-drawn effects helped a small team create a high-octane cinematic without a traditional Maya-to-engine animation pipeline.

Dardo Studios originally planned Ghost Stride as a 15-to-20-second parkour sequence demonstrating its animation capabilities. The concept gradually expanded into a complete stylized short that brought together the studio’s character, environment, animation, lighting, VFX, technical art, and illustration teams. Its high-speed action takes inspiration from Spider-Verse, Arcane, Team Fortress 2, and Genndy Tartakovsky’s work, but pushes that painterly visual language into darker horror territory.

The team also used the project to test whether an entire cinematic animation pipeline could live inside Unreal Engine. Rather than rigging and animating in Maya or Blender before exporting the results, Dardo built its rigs with Control Rig and animated the characters directly through Sequencer.

Dardo supplemented Unreal’s real-time rendering with more than 60 painted keyframes and hand-drawn 2D effects. To discuss the whole process in more detail, we spoke with Kenny Carvalho, Co-Founder at Dardo Studios; Andrew Miller, Lead Technical Artist; Jesse Baumgartner, Principal Animator; Sacha Veyrier, Lead Environment Artist; Daniel Orive, Co-Founder and Head of Art; and both Dean Boadana-Krone and Thomas De Lacroix, from the Technical Art and VFX team.

For some of the newcomers, give us a short intro to Dardo Studios. What do you guys do? Who are the founders? What are their previous projects and current cool projects shipped?

Kenny Carvalho, Co-Founder of Dardo Studios: Dardo Studios is an art co-development studio with over 60 artists and experienced game developers. Founded in 2023, we work across the full art pipeline, either acting as a full art department for a game studio or as an extension of an existing team to help scale production. Our work spans art direction and concept development, character and environment art, illustration, animation, VFX, lighting, and technical art. When needed, we also consult on art department matters, including fractional art management and recruitment. That’s why we like to describe what we offer as an “Art Department in a Box™.”

Dardo was founded by Kenny Carvalho and Daniel Orive. Kenny previously led creative talent acquisition at Blizzard Entertainment, Riot Games, Wildlife and Phoenix Labs, while Daniel brings experience in art leadership and creating art himself at Riot Games, Blizzard Entertainment, Ninja Theory, Mercury Steam and Phoenix Labs. Together with a leadership team of experienced developers from studios like NetEase, Sloclap, Treehouse, Singularity 6, Airship Syndicate and Epic Games, we set out to build a studio that combines high-quality art with the collaborative mindset of people who understand game development from the inside.

Recently, we’ve been especially proud of our contributions to Invincible VS, across 3D character art, lighting, real-time VFX, concept art and key art, as well as our ongoing work on projects like Marvel Snap and Cyberpunk. We’ve also been working on several exciting projects with top industry partners that we look forward to sharing publicly soon.

Our goal is simple. We want to be a genuine extension of our partners’ teams, helping them tackle creative challenges with top-quality art while staying mindful of performance, budget and scalability. Ghost Stride is a chance to show another side of that, and to share what we can do when we take a project all the way from concept to final.

What would you say are the biggest challenges that companies are facing with premium-quality animation? What are the biggest challenges technologically, business-wise, and so on?

Kenny Carvalho: We think one of the biggest challenges is finding the right balance between quality, production time, and budget. Everyone wants animation that feels polished, expressive and full of personality, but getting there takes time, experienced people and a strong understanding of how the animation fits into the game or product as a whole.

Technologically, the tools are getting better all the time, but integrating animation into a game is still a complex process. It’s not just about creating a beautiful animation in isolation. It needs to work with the character, the rig, the gameplay, the camera, and the technical limitations of the game, and getting all of those pieces to work together while preserving the original artistic intent can be challenging. Cinematics give you a bit more freedom, since you can animate for a specific shot, but they need their own kind of problem-solving to achieve unique looks that capture the audience’s attention. That’s something we explored a lot on Ghost Stride.

Kenny Carvalho: On the business side, premium animation is expensive and hard to scale. Studios often need a lot of animation in a short window, usually around a major milestone, a launch window, or a new season. After that, the demand becomes a little less afterward, which makes it hard to justify building a large-scale in-house team for those peaks. At the same time, animation is one of the most visible parts of a game, so cutting corners shows immediately. Finding a way to scale up quality without scaling up risk is a big part of why studios look for external partners.

From a production standpoint, communication and iteration are just as important as the technology and the animation itself. Animation is very subjective, and getting everyone aligned on what feels right can take a few rounds of feedback. When teams are working across different disciplines or with external partners, clear direction and a good feedback process make a huge difference.

Finally, there’s the challenge of finding people who combine strong artistic skills with a good understanding of game development. That’s something we take a lot of pride in at Dardo Studios. Our artists come from inside the industry, so they understand that premium animation isn’t just about making something look great. It needs to feel right, serve the character and the game, and work within the realities of production.

Ghost Stride is your experiment with taking the full animation pipeline into Unreal. It's absolutely lovely. How does your team look at Unreal as an animation tool? What are the advantages, and what are the things that it still lags in?

Andrew Miller, Lead Technical Artist: Unreal has become an amazing tool, not just for building games but, like you said, for building entire pipelines. For us, Control Rig and Sequencer made the whole animation process incredibly simple and tightly integrated with the work of every other discipline on the project.

Being able to build and iterate on the environments alongside rigging and animation was a game changer. Toward the end of the film, there’s a shot where Shadow is bouncing on power lines, taunting Ghost. We initially had the shot timed out and both characters animated, but realized Shadow lacked that sense of bounce and weight on the power lines that makes the final version so cool. Unreal let us build a Control Rig for the power lines after the fact, pop it into the scene, constrain Shadow’s rig to it, and animate. Iterating inside the actual scene was huge for keeping every discipline on the same page while we were actively changing shots throughout the film.

Andrew Miller: Sequencer is really the glue that holds the pipeline together. Being able to control not only character animation but also shader parameters, lighting, effects, and Blueprint events makes pulling a full shot together so much easier. Everything is in one place and can be iterated on at the same time. Shots that need precise timing, like when Ghost turns around and sees Shadow for the first time, and we catch her reflection in his glasses, could all be timed out together in Sequencer. Ghost’s keyframes and the shader parameter keyframes live in the same timeline, which is just so cool.

That said, Unreal still has some catching up to do as a pure animation tool. Compared to a dedicated package like Maya, some of the tools animators rely on every day aren’t as mature yet, and working inside heavy scenes full of lighting and effects means performance is always something you need to keep an eye on. None of these were deal breakers for us, and Unreal keeps improving with every release, but they’re worth knowing about going in.

Tell us a bit about your approach to these high-octane action animations. This is where the real meat of the craft is. If someone from the team could contribute and share some techniques, challenges, and approaches here, that would be great.

Jesse Baumgartner, Principal Animator: When this piece first started, the brief was simply to create a short animation that would highlight our capabilities as a studio. Jesse, our animator, planned out a 15 to 20 second parkour action piece that would have been a lot of fun to make, but probably would not have shown the full range of what we can do. Daniel Orive, our creative and art director, had the idea to make it bigger and more meaningful, and to bring in more disciplines so we could truly showcase our abilities as a team and create something that would really challenge us.

Jesse Baumgartner: For the animation style, we drew most of our inspiration from Spider-Verse (of course), but we also looked at Team Fortress 2, Arcane and Genndy Tartakovsky's work for personality, fluidity and intensity. Even with all that stylization, everything is still grounded in reality. We always start by gathering a lot of references to build a solid base and then push things a little further from there.

One of the things we are most proud of is that we rigged and animated everything directly in Unreal. The usual approach in the industry is to rig and animate in a DCC like Maya or Blender and then import the animation into Unreal for lighting and rendering. We wanted to see how far we could push a fully in-engine workflow, using Control Rig for the rigs and Sequencer for the animation, so the whole process lived in one place from the first blocking pass to the final render.

Jesse Baumgartner: That decision changed the way we worked. In a traditional pipeline, every change means going back to another software, re-exporting and re-importing, and checking that nothing broke along the way. Working in Unreal, there was no round trip. Jesse could block out shots very fast and rough out the animation at the same time, so there was hardly any downtime in production. Even more important, he was animating with the final cameras, lighting, and shaders right in front of him. For high-octane action, that is a huge advantage, because you can judge silhouettes, readability and impact exactly as the audience will see them, instead of guessing in a grey viewport and discovering problems later. It also meant layout, animation and lighting could evolve together, and feedback from Daniel could be applied and reviewed almost instantly.

Rigging in Unreal turned out to be one of the biggest wins. Control Rig was simple to work with and gave us almost all the flexibility we needed. Whenever Jesse needed an extra control for a specific action or pose, he could edit the rig, test it and be back to animating in no time, without waiting on a separate rigging pass or another export. For a small team working in limited windows, that speed made a real difference.

It wasn't all smooth, of course. Jesse comes mostly from Maya, so he was used to working a certain way with a set of tools he knew really well, and switching to Unreal for a cinematic like this came with a learning curve. Some of the tools animators rely on every day aren't as mature in-engine yet, and animating inside heavy scenes full of lighting, decals and effects meant we always had to keep performance in mind. But once we adapted, the benefits clearly outweighed the trade-offs.

For us, Ghost Stride proves that a full cinematic, from rigging to final animation, can be built entirely inside Unreal, and that's something we're excited to bring to future game and cinematic projects with our partners.

Could you share how you've been working with the camera, and especially share some of your thoughts on how to make the camera work well in Unreal? What are the opportunities you open up while doing these kinds of projects in a tool like that?

Sacha Veyrier, Lead Environment Artist: Camera work was essential on Ghost Stride, and it was one of the very first things we blocked out, side by side with Jesse's character animation. From the start, we treated the camera as a storytelling tool rather than something to add at the end, because in an action piece like this, where the camera sits and how it moves decides how every hit, jump, and fall feels.

Being able to animate and tweak both the cameras and the characters directly in Unreal, inside the final environment and lighting, was a game changer. Everything was instant, and the feedback loop was massively faster. There was no switching programs, no reimporting, and no discovering issues and discrepancies late in the process. What we saw on screen was what we were going to get. Unreal's cinematic cameras behave like real cameras, with real focal lengths, sensor sizes, and focus, so we could think like a cinematographer, choosing lenses for each moment and seeing the depth of field and composition live while the action played out.

Once the framing was blocked, it defined the bounds of the action and the environments. More importantly, it gave us a fixed canvas, which allowed us to work on each shot with the same level of care you'd give an illustration. We treated shots like paintings. Every asset was worked on based on what it could bring to the overall image, and it only needed to work for that camera angle and distance, nothing more. Because we knew exactly what the camera was showing and what it wasn't, we could hand-place every detail efficiently. In a gameplay setting, you can guide what the player sees to a certain extent, but a fixed camera gives you much more control and more time to push the art even further. This was also what made our paintover projections possible, since those brushstrokes were painted for one specific view and only hold up through that camera.

Sacha Veyrier: It also opens the door to some pretty nifty ways to cheat, tricks borrowed straight from the animated film industry. Some camera placements were impossible at first because our 3D buildings and walls were in the way, so we carved out or extended buildings and built different swappable versions of the same scene to fit each shot. Sequencer made this easy to manage, since each shot could have its own setup without affecting the others. What happens outside the frame is irrelevant, and that can be exploited to achieve a lot more with fewer resources. As long as it looks good on camera, it's fair game.

For us, that's the biggest opportunity Unreal opens up. It lets you work like a film crew and a game team at the same time, with the speed of real-time and the per-shot control of a film pipeline, and that combination is what allowed a small team to put together a piece like Ghost Stride.

Another huge element of the project that really does make it stand out is the phenomenal stylization and art direction. What were your main inspirations and choices that helped guide the general look of the project?

Daniel Orive, Co-Founder and Head of Art: Thank you so much! The art style is probably what we're proudest of. Ghost Stride wasn't just a short for us; it was a chance to show the world the house style that defines Dardo Studios, something that sits between illustration and animation, polished but with an edge.

To understand where that style comes from, you have to go back a few years to LAX, a project some of our team worked on at Phoenix Labs. Daniel Orive, creative and art director on Ghost Stride, was also the art director on LAX, and a lot of the DNA you see here was born there. Daniel's inspiration has always been deeply personal. It comes from his youth in Madrid, where he was really connected to the underground side of the city, from his passion for tying every piece of work he does to music, and from his many years riding and being part of the BMX freestyle community.

We paired Daniel with the amazing Ed Laag, who co-led the art direction. Both have very strong painterly personal styles, and putting them together pushed the look further than either would have gone alone.

Our reference board was full of the usual suspects for anyone who loves stylized animation, like the Spider-Verse films, Arcane, and pretty much anything Alberto Mielgo does. Those projects proved audiences are hungry for 3D that feels handmade. But we didn't want to just echo them. From day one, the goal was to take that painterly, stylized language and drag it somewhere darker, toward horror. Painterly styles are usually associated with warmth and energy, so using that same visual vocabulary to create dread and unease felt like fresh territory.

The hardest part was the translation from 2D to 3D. Concept paintings have a looseness and life that can easily die once they become models and textures, and we were determined not to lose that vibe. Environments were the success story here, and the assets in Unreal stay remarkably faithful to the original paintings.

Daniel Orive: Characters were a different fight. Our first approach was painting textures with that illustrative look, and on a still frame, it worked. But once the characters moved, something felt wrong. The brushstrokes were locked to the surface, always in the same place, and the result was static and predictable, more like a painted statue than a living illustration. We realized what we really wanted was the brushwork to be painted frame by frame, the way it would be in traditional 2D animation.

The problem is that frame-by-frame painting is incredibly labor-intensive, and we simply didn't have the resources. So we built our own pipeline where we painted a handful of keyframes per shot and used that brushstroke information to project it across the rest of the frames. That gave us the living, shifting quality of hand-painted animation while saving a huge amount of work. It worked so well that we ended up using it on a few environment shots too, to push things Unreal couldn't do on its own, although that was kept to a minimum.

The final layer was 2D FX. We knew from the start that hand-drawn effects would be the glue between the 3D and the painterly look, so we sprinkled them throughout. But the real centerpiece, and our biggest headache, was Shadow's hair.

Daniel Orive: We wanted it to behave like a living shadow, starting as fire at the root and dissolving into smoke as it trails off. We tried everything to keep it inside Unreal. We ran 3D VFX tests; we even tried animating it in 3D, and none of it captured the feeling we had in the concepts. It was only when we started testing it as 2D FX that it finally clicked. It's now one of the elements we're happiest with, and a good example of how we approach problems, always choosing the solution that serves the image even when it isn't the most obvious one.

In the end, all of this, combined with a passionate team of artists at Dardo Studios, is what made Ghost Stride possible. It's why we feel confident calling this our house style, and it's only the beginning of what we want to do with it.

While working on the style, art, design, shaders, and other elements, how different is the process from doing characters for games vs. doing characters for film? There's so much more movement; it's much more directed, but at the same time, there are literally hundreds of moving pieces like fluid motions, effects, and so on.

Daniel Orive: At the start, the process is pretty similar. In both cases, you begin with a concept and try to match the art style as closely as possible, then you sculpt the high-poly mesh, and it's when you get to the low-poly that things start to diverge. Games today can handle a lot of polygons, but film still gives you more flexibility and bandwidth, especially on pre-rendered work. A game character also has to read well from every angle and at any distance, while a film character only has to look great through the camera in each shot, so you can spend your detail exactly where the audience will see it. Since we built Ghost Stride in Unreal, we still had some real-time limitations, but we pushed Ghost as far as we could so all the detail and shapes were in the mesh itself rather than relying on displacement maps.

That idea of building for the camera changes everything around the character too. In a game, the player controls the camera and can walk anywhere, so environments have to hold up from every direction. In film, you know exactly what each shot will show, so the world works more like a film set. You build and dress what the camera sees, and you can move props, lights, or even parts of the environment from one shot to the next if it makes the composition stronger.

Rigging and animation are another area where the two worlds split. In games, a rig has to work for hundreds of animations that blend into each other, react to player input, and stay within a budget, so everything is built to be clean and reusable. In film, every shot is its own little world. Animation is directed for a specific camera, so you can cheat poses, fix silhouettes shot by shot, and add details that only need to work for a few seconds. Timing is different as well. Games usually aim for smooth, high frame rates, while film lets you play with timing and frame rate as a stylistic choice, and for an illustrative look that can make movement feel much more hand-crafted. That freedom is a big part of why cinematic animation can feel so much more expressive.

Daniel Orive: Lighting follows the same logic. Game lighting has to work in real time for any situation the player might create, so it's often more general and systemic. In film, you light shot by shot like a cinematographer would, placing rim lights, shaping shadows, and pushing contrast exactly where the story needs it, even if it isn't physically consistent from one shot to the next. Camera effects like depth of field and motion blur also become storytelling tools rather than settings the player can switch off.

On the shader side, there are still a lot of similarities, but because we were working in a real-time pipeline, we had to be very conscious about performance. Even though everything ends up exported as a movie, you still need machines that can handle all those renders and elements while the team works. The environments had a lot of decals, and the main character used two different meshes, one for the main shader and another for the outline shader, which already comes with a big performance cost. Once we started lighting each scene and adding VFX placeholders, everything got slower to handle. The upside is that in film, a shader only needs to look right for the shots it appears in, so you can tweak it per shot instead of finding one setup that works everywhere.

Daniel Orive: Where film really opens up is after the render. Because the final product is a movie rather than a game, we had the freedom to build extra pipelines on top of Unreal, especially for comp, paintovers and 2D FX. The dynamic, illustrative look you see in the short comes from painting over key shots and projecting that work back onto the 3D, and that simply wouldn't be possible in a real-time game, where the player can rotate the view at any moment. In theory you could take the more than 60 paintovers we did, bake them into the main textures and swap them throughout the piece, but we don't think that amount of work is practical, and we don't think you'd get the same result or pixel resolution.

The same goes for 2D FX. We tried flipbook-style effects in Unreal, but they didn't give us the freedom we got from animating them frame by frame later with traditional techniques. Shadow's hair is the best example. In a game, it would have to be a real-time effect that works from every angle and in every situation, but in film we could hand-animate it for each shot and make it feel exactly the way we imagined.

The funny part is that because we built everything in Unreal, a lot of what we made could almost be reused for a game. In the end, there's no right or wrong answer. You need to find the tools and solutions that fit your project, and every choice you make leads you down a different path to solve the problems in a way that works for your needs.
8.

For smaller teams and smaller projects, what's your overall take on Unreal? Is it more cost-efficient than other pipelines? What is it lacking? What could be added or changed? I'd be very interested to learn more.

Dean Boadana-Krone and Thomas De Lacroix, Technical Art and VFX: We think Unreal, and especially Lumen, can be a great choice for small teams making their first short film. You can get good results quite fast without having to build everything yourself. Lumen and Nanite make it easier to create detailed environments with realistic lighting, and Niagara gives you a powerful tool for particle effects in the same engine. You can get a lot done without needing a huge team or a render farm. The final result is not always as polished as with a proper offline renderer, but the main advantage is that you can work fast and see the results almost immediately.

For our project, we chose Lumen instead of path tracing, mainly because we already knew it well from working on games. It was good enough for the look we wanted, and we could light a shot, move around in Sequencer, make changes, and see the result straight away. Having everything in one engine was also useful because people from the game side of the studio could jump in and help with the film.

Dean Boadana-Krone and Thomas De Lacroix: However, Unreal is still a game engine, and you can run into limitations when using it for film production, especially if you don't know about them from the start. Lumen can struggle with some lighting setups, and lighting and reflections can sometimes be inconsistent between frames. You also need warm-up frames to let the lighting settle, and emissive materials and thin geometry can be tricky to handle. These are small things on their own, but they add up. Making something look good in the viewport is one thing, but getting a reliable final render is another.

One of the hardest parts was compositing. Getting separate render passes out of Unreal for use in After Effects or Nuke is not as straightforward as with a traditional renderer. With path tracing, for example, it's easier to get passes like beauty, depth, or ambient occlusion. With Lumen, we had to research how to get the passes we needed and build our own solutions by enabling or disabling different features, and even then the results weren't always the same as proper passes straight from a renderer. Lumen also relies on temporal accumulation, which can cause noise or ghosting in some of those passes. And when something goes wrong in Movie Render Queue, tracking down the cause can take a lot of time.

Dean Boadana-Krone and Thomas De Lacroix: Looking back, since we were rendering everything through Movie Render Queue and relying heavily on comp anyway, a more traditional offline rendering workflow might have served the final image better. Still, we don't regret using Unreal. Working in real time is what made it possible to rig, animate, light, and shoot everything in one place, and it worked well for what we wanted to achieve.

If we could change one thing, it would be better support for rendering separate Lumen passes, similar to what's available with path tracing. A clearer and more reliable rendering workflow would be really helpful, especially for small teams without a dedicated rendering specialist. We understand Lumen was designed mainly for real-time rendering in games, but making it easier to use for film production would open up a lot of possibilities.

Overall, we think Unreal is a great tool for small teams and people starting out, because you can do a lot in one place and get strong results without a huge budget. But it doesn't automatically make every project cheaper. The time you save during content creation can end up being spent later on optimization or solving technical problems. It also depends on the type of project, and once you need more advanced compositing or a traditional offline rendering workflow, other tools might be a better fit.

For us, Unreal was a good choice for our first film project. We learned a lot from it, and it gave us great results, but it's important to understand its limitations and know what you're getting into.

Overall, what were your biggest learnings here? Something that you'd love to share about the Ghost Stride experience?

Daniel Orive: Every project teaches us something as artists, and once it's done, we always look back at the things we would do differently. With Ghost Stride, that's even more true, because it pushed us beyond our usual boundaries and helped us take a big step toward showing what we can do on the animation and cinematic production side.

The biggest lesson for us was understanding where a real-time pipeline shines and where it starts to push back. Unreal gave us really fast iteration and dramatically reduced render times, but for the level of comp and lighting fidelity we were going for, getting the render passes we needed and keeping performance manageable for the whole team was a real challenge. A pre-rendered pipeline would have given us more options there, with the trade-off of needing powerful machines or a render farm to put it all together at the end.

Another important lesson, or at least a good reminder, is how much a steady cadence of work matters when you're building something as a team. Ghost Stride was a side project at the studio, worked on in the gaps between client projects, so availability came in limited windows and it was a challenge to bring everything together at the pace we wanted. Keeping momentum and team synergy is really important for us as creatives, but there's never going to be a perfect time to do the things we want to do. You have to take what you have and do your best with it, and hopefully that leads to opportunities where future projects get full resources.

Ghost Stride also grew a lot along the way. It started as a quick animation scene to show our animation capabilities, but we wanted to aim for something bigger and start telling the kind of stories that inspire us. We couldn't just press a reset button and start from scratch, so we took the work we already had, reused pieces here and there, and shaped it into a short with a story behind it. That story ended up being a metaphor for any artist out there, and especially for us at Dardo. We all have to face our fears head-on, even when there's negativity or uncertainty around us, and what matters most is to keep moving, take on our challenges, and learn from them whether we succeed or fail.

Daniel Orive: We also want to give a big shout-out to our team. Our people are incredibly passionate and skilled, and they're the reason a piece like this exists. People already recognize the quality of Dardo's concept and illustration work, since that's the kind of work we can share most easily. We hope Ghost Stride gives a deeper look at what we can put together as a studio, and shows that we're a true co-development partner that can match any game or cinematic project out there.

From all of us at Dardo Studios, thank you to 80 Level for the opportunity to share our work and our story with the community, a community that means so much to us. We hope people enjoy watching Ghost Stride as much as we enjoyed making it, and that potential partners get to know more about our creative capabilities.