SHADER vs PHOTOREAL

Left is a live shader — running right now in your browser, never repeating, infinitely long. Right is a still from FLUX on the studio 4090. Same subject, same aspect. The point is not which wins; it is what specifically the photo has that the shader does not.

TIDE FLATS

Shallows vs a generated tide flat. Watch the ripples in the photo compress toward the horizon — that is the depth cue the shader was missing entirely, and why it read as a flat wall.
SHADER · LIVE
FLUX · STILLgenerated tide flat

FOG OVER WATER

Fogbank vs a generated fog field. The photo has aerial perspective doing all the work: the treeline loses contrast with distance and the reflection softens as it recedes.
SHADER · LIVE
FLUX · STILLgenerated fog field

SEA CAVE

Azzurra vs a generated sea cave. The photo wins on surface — wet rock, real caustics, the exact way water scatters. The shader wins on never being the same twice.
SHADER · LIVE
FLUX · STILLgenerated sea cave

WHAT THIS ACTUALLY SHOWS

The photo's advantage is depth cues, not detail. FLUX learned from photographs, so it gets aerial perspective, texture compression and correct reflection falloff for free. A shader only has what is explicitly written into it — which is why adding ground-plane perspective to Shallows changed it more than any amount of retiming the waves.

The shader's advantage is time. A plate is one frame. Loop it and a 12-hour spa day will make the loop legible — someone in the water for 40 minutes will notice. Shaders never repeat, which is the whole reason they suit this room.

So the hybrid is not "pick one". It is: use a plate where the subject is nearly still — fog, a water surface, wet rock, embers — and let the shader supply every bit of the motion and change on top. The plate gives photographic surface; the shader guarantees you never see the same minute twice.