Almost every 3D production that needs more than three characters ends up reusing one body. A single mesh is modelled carefully, rigged once, and then reshaped into everyone else with blend shapes, sliders, and proportion scaling. It is an enormous saving and it has one predictable side effect: everybody in the cast starts to look like a cousin of everybody else.

What a morph target actually is

A morph target is the same mesh with the same vertices moved somewhere else. The software stores the difference and can apply any fraction of it. Slide to zero and you have the original; slide to one and you have the variant; slide to 0.4 and you get something between. Several morphs can stack, which is where the combinatorial power comes from.

Because the vertex order never changes, everything built on the mesh keeps working. Weights transfer. Clothing transfers. Facial rigs transfer. That is the real prize, not the shapes themselves. Rebuilding a face rig for every character is what makes bespoke modelling expensive, and morphs remove that cost almost entirely.

What you gain

Speed, first. A new supporting character can be dialled in over an afternoon instead of built over a fortnight. Consistency, second: everyone deforms the same way, so an elbow fix applies to the whole cast at once. And flexibility late in a production, because a note like make her taller is now a slider rather than a rebuild.

There is also a quality argument. One base mesh that receives months of attention will have better topology than five meshes that each received a week. For deformation-heavy work, especially close scenes where bodies press against each other, that shared quality is worth more than superficial variety.

What you lose

Silhouette. Morphs move vertices, and vertices are limited to the surface they started on. You can widen a jaw but you cannot grow a new shape of ear. Over a cast this produces the flat feeling that many viewers describe as everyone looking the same, without being able to say why, because the faces genuinely are different while the outlines are not.

Proportion range is the other limit. Push a morph too far and topology stretches unevenly, texture density goes wrong, and the rig starts pinching in new places. Most base meshes have a comfortable band of maybe fifteen per cent in any direction, and beyond that you are better off building a second base.

Adding real variety back

The fix is to vary the things morphs cannot reach. Hair as separate geometry, since hair carries most of a silhouette at a distance. Clothing with its own volume rather than skin-tight shading. Posture and default stance, which change the outline more than facial structure does. And accessories with hard shapes, because a hard shape is visible when everything soft has blurred.

Test the cast as black shapes on white. If you cannot tell them apart, no amount of face sculpting will help, because at normal viewing distance the audience is reading outlines. That is the same test described in the piece on giving two women separate physical logic.

A cheap trick worth knowing: vary the eye spacing and the head-to-shoulder ratio rather than the features themselves. Those two numbers do more to change how a face reads at a distance than any amount of nose sculpting, and both are easy morphs. Readers identify characters by proportion long before they identify them by detail, which is why caricaturists work the same way.

Where one base mesh stops being enough

Sooner or later a design arrives that the base cannot reach. Accept it early. Building a second base for a genuinely different body type is cheaper than five weeks of fighting a morph that was never going to get there, and the second base can borrow the first one's rig approach even if it cannot borrow its topology.

Keep the whole cast fictional and unmistakably adult, and treat adult proportion as one of the base mesh's non-negotiable properties rather than a slider anybody may drag. Proportion trade-offs generally are covered in the stylised proportion notes, and the studio route for making original characters is on the studio page.