Every few months the argument resurfaces: practical effects have aged gracefully while computer imagery from the same period looks dated. There is real evidence for this, but the usual explanation — that physical craft is inherently superior — does not survive contact with the counter-examples.
The observation is broadly correct
Certain films from the 1980s still hold up under scrutiny. Creature suits, animatronics, miniature work and matte paintings from that era frequently read as convincing today, while digital work from only a decade later can look distractingly soft.
At the same time, some early digital effects remain almost invisible. The most cited examples are sequences where the computer-generated element was on screen briefly, moving fast, and photographed in conditions that hid its weaknesses.
Light is the variable that matters
A physical object placed on a set is lit by the same lamps as everything around it. Its shadows fall correctly, its surface picks up bounce from nearby walls, and the camera records it through the same lens with the same imperfections. All of that consistency is free.
A digital object has to reproduce every one of those relationships through simulation. Early renderers approximated lighting crudely: shadows were too clean, surfaces lacked the subtle grime that real materials accumulate, and the object often failed to interact with anything around it. The eye detects the mismatch immediately, even when the viewer cannot say why.
Why practical work also fails
The nostalgic version of this argument tends to forget how much physical effects work was unconvincing at the time. Visible wires, rubber that moved like rubber, miniatures betrayed by the wrong scale of water or smoke. Survivorship bias does a lot of work here: we remember the practical effects that succeeded because the failures are not the ones anyone rewatches.
Practical effects also impose real limits. A suit constrains movement. A miniature constrains camera moves. Many of the most admired physical sequences are admired partly because the filmmakers worked around those limits with staging and editing rather than fighting them.
What modern productions actually do
The framing of practical versus digital barely describes contemporary filmmaking. Almost every large sequence is a composite: a physical set extended digitally, a creature performer replaced above the neck, a miniature enhanced with simulated debris.
- Physical anchors. Even when an element will be replaced, having something real on set gives actors an eyeline and the lighting crew a reference.
- Digital extension. Building the first two floors and extending upward is cheaper and looks better than building nothing.
- Simulation for the impossible. Fluid, cloth, crowds and destruction are areas where digital tools are simply better, and no amount of craftsmanship changes that.
The scheduling problem nobody mentions
A significant share of weak digital work is not a technical failure but a production one. Visual effects sit at the end of the pipeline, which means they absorb every delay that happened upstream. When a release date is fixed and the edit locks late, the effects team loses the exact time it needed for iteration.
Physical effects cannot be compressed this way. If a creature suit is not ready, shooting stops. That hard constraint accidentally protects quality, and it is a real if unromantic part of why older work holds up.
How to judge an effect fairly
The useful question is not which technique was used but whether the shot obeys its own physics. Does the object have weight? Do the shadows agree with the light sources? Does the camera treat it like something actually present, with the same focus and motion blur as everything else?
When those conditions are met, audiences do not notice the technique at all — which is the entire point, and the reason the best examples of both approaches rarely come up in this argument.