3D Printing for Sculptors: Turning a Horse and Sidesaddle Rider into a Large Working Pattern

The finished sculpture

This is what the completed digital sculpture looks like: a horse mid-stride on a sculpted rocky base, carrying a sidesaddle rider in period dress with a bedroll and saddlebag behind her.

The horse

We modeled the horse and its equipment here, working from the artist’s direction, and finished it on its own before any rider was integrated. The saddle, the saddlebags, the bedroll, and the rocky base were all part of that work. Establishing the animal first meant the rider could later be fitted to a settled silhouette instead of both halves moving at once.

The sidesaddle rider was a separate problem

Once the horse was finished, the rider went out to a specialist modeler. That figure is not our work, and it is worth saying plainly, because it was a discipline of its own.

A sidesaddle figure cannot be treated as a symmetrical seated pose. Both legs and both boots stay together on the horse’s left side, one raised and one hanging lower, and that asymmetry has to survive every view and every revision. Approved front, back, leg-side, and non-leg-side views mattered more than usual here, because image-to-3D tools tend to invent hidden geometry and quietly restore symmetry. The face, bonnet, collar, clothing folds, hands, reins, and the contact between rider and saddle all needed direct review.

Our part in the rider was on the production side: confirming it would survive at print scale, closing small open boundary loops around the reins and hands, and planning cuts that kept every seam away from the face.

The details that decided the sculpture

The bedroll, the saddlebags, the straps, and the tack were modeled here as sculpture in their own right, not as decoration added at the end. Each one changed the silhouette, the rider’s contact with the horse, the printability of thin elements, and how hard the supports would be to remove later. A bedroll behind the rider stays visible from almost every angle, so its position, rolled ends, compression, and straps all had to be resolved properly. The period footwear shown here was part of the artist’s direction for the rider.

Who did what

The artist supplied the art direction and reviewed every detail along the way. Working to that direction, Cru Bear modeled the horse itself, together with the saddle and tack, the saddlebags, the bedroll, and the sculpted rocky base.

The sidesaddle rider was a separate commission and is not our work. Once the horse was settled, the figure went to a specialist modeler who handled the pose, likeness, clothing, hands, and reins. Cru Bear then took the combined model through production: file review, mesh optimization, scale verification, section and seam planning, connector and support engineering, multi-plate production, and the practical handoff for the next artistic stage.

That division matters, because the technical bridge between a dense digital sculpture and a usable physical form is its own discipline. It is the part most artists do not have the tooling or the time for, and it is the part that decides whether a large print is usable or wasted filament.

The delivered file was too dense and too large to print

The combined model arrived as a 1.10 GB binary STL containing about 22 million triangles across 14 overlapping shells. Nothing about that file was ready for a print bed.

The original was preserved untouched and all production work happened on a separate copy. That copy was scaled uniformly to exactly 16 inches tall, conservatively reduced to roughly 5 million triangles, and repaired where small open boundary loops appeared around the reins and hands. The reduction was deliberately conservative rather than aggressive, because the whole value of the print was the accumulated design detail. A re-import verified zero boundary, wire, non-manifold, or degenerate geometry, and the slicer confirmed the production model as manifold at 203.6 by 561.2 by 406.4 mm.

Cutting the model was design work of its own

At final scale the piece could not fit a single build plate, so it had to be divided. Cut placement balanced four competing pressures at once: where a seam would be visible, how each section would be oriented for its best surface, whether connectors could be reached during assembly, and the usable nozzle range for the section.

One real constraint shaped the plan. The cutting workflow is axis-aligned after the model is rotated, so a seam cannot follow an arbitrary curved feature. A cut cannot be tucked directly beneath a bridle strap, for example. It has to sit above or below it. Knowing that limit up front is what keeps a seam from landing across a face or a detail that cannot be reworked.

Six build plates

The final production project was organized across six plates. Splitting the work this way is not a compromise. Smaller sections, each oriented for its own best face, consistently produce better visible surfaces than one heroic single print.

Printing and handoff

Production ran at 0.24 mm layer height using a support-interface material so the undersides of the horse, the rider, the reins, and the overhanging details had something clean to build against. The complete separated effort took approximately six days, and the largest single continuous print ran right at three days. Support removal alone took well over an hour, which at this scale is real labor and the point where an impatient hand can tear a printed surface.

The rocky base printed as its own set of sections, with the leg positions built into it. The sections were handed over separately so the sculptor could start immediately, which is why no photograph exists of the white horse-and-rider section mounted on the black base.

Why this print counted as a success

The sculptor’s own assessment was that the print came out better than expected and was fully usable. Not because every surface was flawless, but because a working pattern is judged by a different standard than a display piece. Its job was to preserve the expensive design work already invested in the posture, the hand on the reins, the footwear, the tack, the bedroll, the saddlebags, and the horse type, and to hand all of that over as something physical.

From there the print becomes raw material. It gets cut apart, molded, resculpted in plaster or clay, reviewed by a client, scanned, expanded in foam, refined, and eventually prepared for a foundry. Marks in open areas can be reworked. Only the areas a sculptor cannot easily rebuild, like eyes, lips, and mouth, carry real risk.

What this project taught us

• Decide at the start whether the print is a finished part or a sculptor’s working pattern. It changes everything downstream.
• Orientation controls where layer lines and tool marks land, so it is an artistic decision as much as a technical one.
• If a large figure can be split into smaller purpose-oriented sections, that almost always beats one heroic print.
• Support-interface material protects detail underneath, at the cost of setup and print time.
• Break support trees into small pieces before pulling them, or they will tear the surface.
• Run a small orientation and support test before committing to a multi-day job when the surface target is strict.
• Preserve the original source file and do every reduction, repair, and cut on separate production copies.