Painted portraits of famous figures reveal a lot about their likeness, with the mastery of artists being in their ability to capture someone's essence. Yet, they lack the true three-dimensionality and plasticity of a face. And reconstructing these elements is exactly what keeps Brazilian 3D designer and researcher Cícero Moraes busy. You might remember him from Bored Panda’s previous feature on his reconstruction of Ivan the Terrible, where he used forensic-style methods and historical data to build a lifelike face from old remains and records.
Moraes has become a familiar name in the world where history, anatomy, and digital art overlap. His work focuses on facial approximation: creating a scientifically grounded “best possible” face from skeletal evidence, whether that’s a famous historical figure or human remains recovered in archaeological digs. It’s not the same thing as identification (DNA still wins that battle), but it can be surprisingly powerful for helping modern viewers connect with people who otherwise feel trapped behind oil paint and mythology.
More info: Instagram | ciceromoraes.com.br
Beethoven’s skull, photographed straight-on, gives the hard limits for face width, jaw shape, and eye placement
Image credits: Cícero Moraes
In this post, we’re staying in the world of classical music, bringing you facial reconstructions connected to Mozart, Bach, and Beethoven. And while each case can involve different source material (a skull, casts, old photos, or measurements), the underlying workflow follows a similar logic: start with the skull’s hard limits, then build outward using anatomical rules and population-based averages.
A side view makes the “profile math” obvious
Image credits: Cícero Moraes
First glimpses of the face appear when the skull scan is uploaded into a 3D modelling program
Image credits: Cícero Moraes
So what does that look like in practice? In Moraes’ digital workflow, the skull is first positioned and scaled in a 3D scene, sometimes reconstructed from published photos and measurements when a full scan isn’t available.
Before the face exists, Moraes builds the framework by overlaying reference guides to the skull
Image credits: Cícero Moraes
The skull gets “dressed” in a digital soft-tissue shell, showing how muscle and fat would sit over bone
Image credits: Cícero Moraes
From there, he adds soft-tissue thickness markers. These are depth guidelines based on real datasets across key points of the face to estimate how much “living” tissue would sit over bone. Next comes feature projection, especially the nose, which is notoriously tricky because cartilage doesn’t survive; to compensate, multiple established methods can be used together to estimate its likely shape.
Outlines and landmarks help compare multiple historical references
Image credits: Cícero Moraes
The skull is aligned with 2D reference geometry, letting Moraes test proportions and correct angle distortions
Image credits: Cícero Moraes
A particularly interesting step is using a “virtual donor” head scan (a CT-based reference) that’s digitally adjusted to match the target skull’s proportions, helping generate realistic facial volume before the final sculpting pass. Once the face is structurally complete, the more interpretive elements, hair, styling, and presentation, get layered in, and final detail can be enhanced with careful, human-supervised AI touch-ups rather than letting an algorithm freeload its way into rewriting anatomy.
Multiple profile outlines stack together, showing where sources agree and where artists may have flattered
Image credits: Cícero Moraes
The “raw” 3D head emerges, then gets refined by adding hair, skin texture, and lighting
Image credits: Cícero Moraes
A more human, less mythic Beethoven emerges
Image credits: Cícero Moraes
Bach begins with a different challenge: matching a 3D facial mesh to a historical portrait
Image credits: Cícero Moraes
The portrait becomes a measurement tool as Moraes maps facial planes and landmarks
Image credits: Cícero Moraes
A calibrated skull model, aligned profile, and head volume work together to shape the final model
Image credits: Cícero Moraes
Mozart’s process starts with reference imagery, turning a painted profile into a usable 3D guide
Image credits: Cícero Moraes
A second reference angle gets matched in 3D, helping confirm the proportions of the brow, midface, and chin
Image credits: Cícero Moraes
A second reference angle gets matched in 3D, and compared to previous attempts at reconstruction
Image credits: Cícero Moraes
Mesh-on-portrait overlay: the face grid is adjusted until key points sit where anatomy says they should
Image credits: Cícero Moraes
The skull model and the facial “envelope” get combined, showing how the reconstruction is constrained by bone
Image credits: Cícero Moraes
A clean 3D sculpt becomes a lifelike head as pores, shading, and subtle asymmetries are added
Image credits: Cícero Moraes
Mozart’s reconstructed face looks softer and more aged than in historical portraits
Image credits: Cícero Moraes