Blab Entertainment
Argentum Studios Three-Year Short Film The Cradle Pushes Unreal Engine 5 to Path-Tracing Limits
Argentum Studio, a German‑based studio that specialises in real‑time cinematic graphics, completed a three‑year short film called The Cradle that was built entirely in Unreal Engine 5. The project was an experimental effort to bring offline‑style path‑tracing quality to a real‑time pipeline, and it required the studio to develop a suite of custom tools for hair, volumetrics, and animation.
The film was conceived as a research experiment during the early release of Unreal Engine 5.1. According to the studio’s CEO and Technical Director, Germans Alekseikovs, the goal was to test how far real‑time rendering could approach the visual fidelity of path‑traced production. Alekseikovs noted that the speed advantage of deferred real‑time rendering—often ten to one hundred times faster than offline path tracing—was a decisive factor. Even when viewport performance dropped to 1–10 FPS, the ability to iterate on lighting, camera work, and asset detail in minutes was far more efficient than waiting days for offline renders.
The Cradle was visually inspired by the surreal paintings of Zdzisław Beksiński. The studio did not attempt to copy the artist’s work directly; instead, it aimed to capture the same emotional tone—surreal, uncomfortable, monumental, and dreamlike. The result is a world that feels physically believable yet symbolically strange.
Because the default Unreal Engine workflows could not deliver the required quality, Argentum Studio built several internal tools. The first was a custom groom‑rendering system called the Hair Cinematic Tool. At the time, artists relied on undocumented console variables to tweak hair quality, a method that produced inconsistent results. The Hair Cinematic Tool centralised control of the engine’s Hair Strands system and added a post‑process layer that ensured viewport and final render consistency.
The studio also created a native VDB‑style rendering workflow. Existing plugin solutions struggled with translucency sorting and integration into Unreal’s rendering pipeline. Argentum’s team rewrote the volumetric cloud shader and incorporated it into the engine’s internal volumetric framework, allowing the film to use VDB data for smoke, fog, and other volumetric effects.
A third major contribution was an Alembic Cache Pipeline. Dense cinematic heads could not be imported reliably through standard Alembic workflows. The pipeline converted animation caches from Blender, Maya, or 3ds Max into skeletal meshes and morph targets that could be used natively in Unreal.
All of the film’s characters involved hair or fur. The moth creature, Diana, and floating figures required a unified approach to hair rendering. The studio used Ornatrix and Houdini for grooming and simulation, then adapted the results for Unreal’s Hair Strands pipeline. Human hair focused on shading softness and fine detail, while the moth’s fuzz emphasized volume and stylised readability.
Rigging varied by character. Diana used a pre‑existing human rig based on Chris Jones’ work and the Universal Human Blender Rig, allowing the team to focus on animation and facial performance. Other characters, including the moth and the cradle itself, were rigged from scratch in Blender, with additional physics for the moth’s wings.
Animation was a significant challenge. The film did not use motion capture; all facial animation was hand‑keyed. The moth’s movement also required careful animation to convey natural flight while remaining cinematic. The cradle’s hand movements underwent multiple iterations before settling on a shredding motion that matched the film’s surreal tone.
The project was completed without a dedicated budget and was produced in the studio’s free time. The three‑year timeline reflects the technical ambition and the fact that many of the engine features used were still unstable during the early UE 5.1 period. The studio’s own growth outpaced the project’s early decisions, creating legacy issues that slowed progress.
In total, 42 people contributed to The Cradle, with a core team of about 15. Alekseikovs said that the experience taught the studio the importance of starting with small, manageable projects. The film’s production demonstrates how aggressive experimentation can lead to new pipelines that benefit future work.
The short film is available on the studio’s ArtStation portfolio and has been showcased at several industry events. While it did not receive a wide theatrical release, it has drawn attention from other studios interested in real‑time cinematic pipelines.
As Unreal Engine continues to evolve, Argentum Studio’s work on The Cradle provides a case study in how real‑time engines can approach offline visual quality when combined with custom tooling and disciplined production practices.
The film was conceived as a research experiment during the early release of Unreal Engine 5.1. According to the studio’s CEO and Technical Director, Germans Alekseikovs, the goal was to test how far real‑time rendering could approach the visual fidelity of path‑traced production. Alekseikovs noted that the speed advantage of deferred real‑time rendering—often ten to one hundred times faster than offline path tracing—was a decisive factor. Even when viewport performance dropped to 1–10 FPS, the ability to iterate on lighting, camera work, and asset detail in minutes was far more efficient than waiting days for offline renders.
The Cradle was visually inspired by the surreal paintings of Zdzisław Beksiński. The studio did not attempt to copy the artist’s work directly; instead, it aimed to capture the same emotional tone—surreal, uncomfortable, monumental, and dreamlike. The result is a world that feels physically believable yet symbolically strange.
Because the default Unreal Engine workflows could not deliver the required quality, Argentum Studio built several internal tools. The first was a custom groom‑rendering system called the Hair Cinematic Tool. At the time, artists relied on undocumented console variables to tweak hair quality, a method that produced inconsistent results. The Hair Cinematic Tool centralised control of the engine’s Hair Strands system and added a post‑process layer that ensured viewport and final render consistency.
The studio also created a native VDB‑style rendering workflow. Existing plugin solutions struggled with translucency sorting and integration into Unreal’s rendering pipeline. Argentum’s team rewrote the volumetric cloud shader and incorporated it into the engine’s internal volumetric framework, allowing the film to use VDB data for smoke, fog, and other volumetric effects.
A third major contribution was an Alembic Cache Pipeline. Dense cinematic heads could not be imported reliably through standard Alembic workflows. The pipeline converted animation caches from Blender, Maya, or 3ds Max into skeletal meshes and morph targets that could be used natively in Unreal.
All of the film’s characters involved hair or fur. The moth creature, Diana, and floating figures required a unified approach to hair rendering. The studio used Ornatrix and Houdini for grooming and simulation, then adapted the results for Unreal’s Hair Strands pipeline. Human hair focused on shading softness and fine detail, while the moth’s fuzz emphasized volume and stylised readability.
Rigging varied by character. Diana used a pre‑existing human rig based on Chris Jones’ work and the Universal Human Blender Rig, allowing the team to focus on animation and facial performance. Other characters, including the moth and the cradle itself, were rigged from scratch in Blender, with additional physics for the moth’s wings.
Animation was a significant challenge. The film did not use motion capture; all facial animation was hand‑keyed. The moth’s movement also required careful animation to convey natural flight while remaining cinematic. The cradle’s hand movements underwent multiple iterations before settling on a shredding motion that matched the film’s surreal tone.
The project was completed without a dedicated budget and was produced in the studio’s free time. The three‑year timeline reflects the technical ambition and the fact that many of the engine features used were still unstable during the early UE 5.1 period. The studio’s own growth outpaced the project’s early decisions, creating legacy issues that slowed progress.
In total, 42 people contributed to The Cradle, with a core team of about 15. Alekseikovs said that the experience taught the studio the importance of starting with small, manageable projects. The film’s production demonstrates how aggressive experimentation can lead to new pipelines that benefit future work.
The short film is available on the studio’s ArtStation portfolio and has been showcased at several industry events. While it did not receive a wide theatrical release, it has drawn attention from other studios interested in real‑time cinematic pipelines.
As Unreal Engine continues to evolve, Argentum Studio’s work on The Cradle provides a case study in how real‑time engines can approach offline visual quality when combined with custom tooling and disciplined production practices.
