An ongoing series of hyperrealistic CGI head studies that sit between emotional realism and surreal abstraction. Each piece combines pore-level surface detail, layered skin materials and effects such as glitch, particle erosion and compositing distortion. Selected works have been shown at exhibitions and festivals internationally, including a public screening at Times Square, New York.
This ongoing series of hyperrealistic CGI head studies explores the intersection of emotional realism and surreal abstraction. Each piece is a personal experiment in digital portraiture, blending sculptural precision with unexpected distortion to create characters that feel both familiar and otherworldly.
The heads combine intricate surface detail, complex materials, and layered effects, including glitch, particle erosion, and compositing techniques to push the boundaries of identity and perception in digital character design.
Selected pieces from the series have been showcased at exhibitions and festivals across the globe, including a public screening at Times Square, New York, drawing international recognition for their striking aesthetic and technical depth.
This body of work continues to serve as a testing ground for advanced sculpting, look development, and rendering workflows, while offering a deeper reflection on the human form in the digital age.
This was a houdini fx & procedural simulation project. See how I work on houdini fx & procedural simulation.
What makes a CGI portrait look hyperrealistic?
Not detail density, and not render resolution. Almost every CGI head that fails is failing on how light behaves, not on how much geometry it has. Skin is translucent, layered and uneven, and the eye is remarkably good at noticing when it is being treated as a smooth surface with a picture wrapped around it.
Three things carry most of the realism in this series:
- Subsurface scattering that varies across the face. Ears and nostrils transmit far more light than a forehead does. A single uniform scattering value is the fastest way to make a head look like wax.
- Two specular layers. One broad and soft for skin moisture, one tight for oil. A single specular layer is the most common cause of the plastic look.
- Breakup at every scale. Pores, fine wrinkles, uneven colour, small blemishes, and asymmetry between the two halves of the face.
These heads are deliberately surreal, distorted, eroded, glitched. That only works if the surface underneath is believable. Without a convincing skin, the distortion has nothing to push against and the piece reads as an abstract shape rather than a violated face.
How is a hyperreal 3D head made?
The ordered steps are listed above. Two things about the sequence are easy to get wrong and expensive to fix late.
Skin look development comes before grooming. Hair changes how you read the skin beneath it, so grooming onto an unfinished shader means adjusting both against each other indefinitely.
Lighting is set before the final texture pass. A texture that looks perfect under flat, even lighting can fall apart entirely under a single hard key light, and finding that out at the end costs a day.
The eyes deserve their own note. They are wet, refractive, never perfectly aligned and always in slight motion. A portrait with excellent skin and dull eyes reads as a mannequin; a portrait with average skin and correctly built eyes usually passes. It is the highest return per hour anywhere in the process.
Which software is used for hyperrealistic CGI portraits?
| Tool | Used for | Why it is the choice |
|---|---|---|
| ZBrush | Sculpting anatomy and pore-level detail | Handles very high polygon counts for displacement |
| Houdini | Grooming, simulation, FX, scene assembly | Procedural, so a change upstream propagates instead of forcing a rebuild |
| Redshift | Shading and rendering | GPU rendering with a mature subsurface and hair model, fast enough to iterate |
| Nuke | Compositing and grade | Pass-level control, lens character, final integration |
| Texture painting | Skin colour and imperfection maps | Layered maps that survive a close crop |
None of it is prescriptive. The same result is reachable through other packages with different friction; what matters is that each stage stays adjustable, because a portrait is finished by iteration and not by a single correct setting.
Why do the effects sit on top of realism rather than replacing it?
Glitch, particle erosion and surface deformation only carry meaning against something the viewer has already accepted as a person. That is the whole premise of the series: build a head that reads as human, then break it in a way that is legible as a violation rather than as noise.
It is also why the series doubles as a testing ground. Every FX layer applied here, dissolution, displacement, compositing distortion, is a technique that later has to survive a client deadline.
Where has this work been shown?
Selected pieces from the series have been shown at exhibitions and festivals across the world, including a public screening at Times Square in New York.
Can this be commissioned?
Yes. Character-led commercials, campaign key visuals, music video sequences and digital fashion all sit on the same foundation, and the same pipeline runs behind them.
Frequently asked questions
What makes a CGI portrait look hyperrealistic?
Imperfection and light behaviour. Pores, peach fuzz, asymmetry, uneven skin tone, subsurface scattering and slightly wet eyes do more for believability than polygon count or render resolution.
How is a hyperreal 3D head made?
In six stages: sculpting the head, developing the skin shader, grooming hair and fuzz, adding fabric and secondary simulation, lighting and rendering in passes, then compositing with lens character.
Which software is used for hyperrealistic CGI portraits?
ZBrush for sculpting, Houdini for grooming, simulation and scene assembly, Redshift for shading and rendering, and Nuke for compositing and grade.
Why does CGI skin often look like plastic or wax?
Usually because subsurface scattering is uniform and too strong, specular is a single clean layer instead of two, and the surface has no fine breakup. Real skin is oily in places, dry in others, and never one material.
Do you need a 3D scan to make a photoreal head?
No. A scan shortens the path because it carries real asymmetry and skin topology, but sculpting from photo reference works and gives more creative freedom at the cost of time.
How important is hair for realism?
Very. Eyebrows, eyelashes and facial peach fuzz are noticed subconsciously even in a close crop, and a head without them reads as synthetic regardless of skin quality.
What is the uncanny valley in practice?
It is what happens when a face is realistic enough that the viewer applies human expectations to it, and then something small, eye movement, translucency, symmetry, violates those expectations.
Where has this series been shown?
Selected pieces have been shown at exhibitions and festivals internationally, including a public screening at Times Square in New York.
Can this work be commissioned for a campaign?
Yes. The same techniques feed directly into character-led commercials, campaign key visuals and music video sequences.