JHDD 3D Modeling Report — 2026.08.21

Undertraining’s recent project, “Bus Ride Straight Out Of Dream,” offers a stark look at the evolving expectations for virtual environments.

The disparate announcements of new game releases and creative projects reveal a quiet acceleration in how artists are approaching the subtle manipulation of perceived reality within digital spaces, prioritizing evocative experience over raw technical fidelity. This shift is less about achieving perfect photorealism and more about crafting specific emotional or experiential outcomes through nuanced control of digital toolchains.

Undertraining’s project, “Bus Ride Straight Out Of Dream,” produced with Blender and 3DGS, demonstrates a subtle but significant shift in how virtual spaces are being conceived. The detail of “undertraining” in its creation suggests a deliberate embracing of outputs that deviate from conventional perfection. Mainstream industry discourse often fixates on pushing polygon counts and ray-tracing fidelity as the ultimate markers of progress in hyper-realism. This view, however, misses the point that perceived reality is often more compelling when it is subtly distorted or imbued with an unexpected logic. True immersion for the user is not always about matching photographic realism, but about crafting an environment that resonates emotionally or conceptually. This involves a nuanced command over lighting dynamics and procedural generation to evoke a specific mood or memory, rather than simply replicating objective light physics. By mid-2027, a significant portion of innovative virtual experience design will deliberately incorporate procedural “imperfections” and unconventional rendering artifacts to create unique aesthetic signatures, moving beyond the brute force pursuit of photorealism.

The ambition to create “bigger worlds” for titles like Mortal Shell II, and the foundational focus of Black Pony Immersive on expansive single-player, first-person action RPGs, inevitably relies on advanced procedural generation techniques. The conventional wisdom states that procedural tools exist primarily for efficiency and scale, allowing developers to fill vast virtual landscapes without manual placement of every asset. However, this perspective overlooks the profound artistic potential of generative systems. The perceived “randomness” of procedural outputs, especially when combined with sophisticated lighting algorithms, can create spatial dynamics that feel genuinely emergent and unpredictable, fostering a deeper sense of exploration and discovery. This is distinct from simply populating a scene; it is about algorithmically composing environments that surprise and engage. Within two years, leading studios will establish dedicated teams focused on art-directing procedural generation’s output, moving beyond merely setting parameters to actively curating the unexpected results, especially regarding atmospheric lighting and environmental narrative.

The primary resistance to this artistic evolution stems from entrenched production methodologies and a market often fixated on measurable, quantifiable metrics of fidelity. Large publishers and development teams often prioritize predictable outcomes, tight asset pipelines, and performance benchmarks that align with established expectations of “hyper-realism.” The perceived lack of direct artistic control over procedurally generated “imperfections” or unconventional lighting schemas creates a hesitancy to embrace methods that might appear less efficient or harder to quality-control against a rigid standard. This is compounded by an industry culture where “polish” is often synonymous with visual perfection, leaving little room for deliberate artistic grit or unsettling strangeness, favoring instead a more conservative approach to what constitutes visual quality.

A working 3D modeling professional should dedicate time this week to experimenting with procedural tools, such as Blender’s Geometry Nodes or Substance Designer, not merely for asset creation efficiency but for aesthetic exploration. Focus on generating subtle variations, controlled imperfections, or deliberate spatial anomalies that would typically be “cleaned up.” Then, rigorously test how these unique procedural outputs respond to a range of dynamic lighting conditions within a virtual scene. The goal is to understand how calculated “mistakes” or unexpected algorithmic results can enhance visual storytelling and evoke specific atmospheres, rather than striving solely for a perfectly rendered, unblemished surface.

---

Curated References