Website powered by

ZBrush 4R8 Beta : 3-D Scientific Illustration Process

My end goal was to come up with a set of nice renders that mimic classic scientific illustrations. I have the skull of a juvenile male black-tailed deer that I dissected and cleaned up myself a while ago, and I began my process of re-constructing it as accurately as possible within ZBrush by taking photogrammetry of it. I then made a rough 3-D model through AgiSoft PhotoScan, imported the raw scan data into Zbrush, and sculpted to refine the data into a more accurate representation of my specimen.

I then had a nice sculpt of the skull, but it was only a solid mesh with superficial details. A model with the internal sinuses, braincase, and cranial sutures carved out would be far more valuable from a scientific or medical perspective. I found a series of CT scan images of a similar species that I used to create a rough starting point for subtractive, internal Boolean shapes, then sculpted them to more accurately depict my skull.

Skull set up for 3-D scanning with photographs

Skull set up for 3-D scanning with photographs

Mandibles set up for 3-D scanning with photographs

Mandibles set up for 3-D scanning with photographs

Cameras set up inside Agisoft PhotoScan

Cameras set up inside Agisoft PhotoScan

The raw photogrammetric data

The raw photogrammetric data

Rough 3-D reconstruction based on the photogrammetry, imported into ZBrush

Rough 3-D reconstruction based on the photogrammetry, imported into ZBrush

3-D scan sculpted and refined in ZBrush

3-D scan sculpted and refined in ZBrush

3-D scan sculpted and refined in ZBrush

3-D scan sculpted and refined in ZBrush

3-D scan sculpted and refined in ZBrush

3-D scan sculpted and refined in ZBrush

Cross-sectional CT scan images imported into ZBrush as a texture on a subdivided plane, then I used the texture to Mask by Intensity, grouped the mask, and used PanelLoops to quickly give volume to each slice of the CT scan

Cross-sectional CT scan images imported into ZBrush as a texture on a subdivided plane, then I used the texture to Mask by Intensity, grouped the mask, and used PanelLoops to quickly give volume to each slice of the CT scan

After a quick Dynamesh, those 2-D CT scan images were reconstructed back into a 3-D model

After a quick Dynamesh, those 2-D CT scan images were reconstructed back into a 3-D model

By simply flipping the normals of the model from the CT scans images, I got this preliminary braincase shape to use as a subtractive Boolean

By simply flipping the normals of the model from the CT scans images, I got this preliminary braincase shape to use as a subtractive Boolean