Watching a video about Disney’s proprietary “Deep Canvas” technology, especially part where the brush strokes are basically 3d objects, made me realize: hey, it was basically Grease Pencil
So the question is: how practical it would be to color the entire 3d scene with Grease Pencil, with level of detail comparable to that Tarzan scene? Will the average PC lock up from that? I’ll try it myself eventually, but wanted to ask someone with GP experience anyway
EIther you want to draw on 3d objects or draw 3d strokes from scratch it cannot be done for various reasons and PC performance is not one of them.
Disney’s Tarzan is 25 years old.
I have no idea to what amount of data those scenes were corresponding, at that time.
But a nowadays average PC should not be far from top-notch workstations of that time.
It is clearly explained that the shots were composited ; and not everything on screen was painted that way.
The same way, you will encounter performance limitations ; if you are using a lot of dense meshes in your scene.
GP objects corresponding to a lot of GP points will be costly.
It is quite easy to end up with a lot of GP points ; if you use brushes producing smoothed curvy strokes, strokes stylized by jittering points.
On the contrary, if you use textured brushes using a dot material ; you can have a lot of details corresponding to few GP points.
GP objects can be instanced, or simplified for several LODs. And that will be eased , when version 3 of Grease Pencil, handling geometry nodes, will be released.
It looks like, for Hero movie, artists preferred to paint vertex colors on meshes, for organic shapes.
GP drawing on surface was rather used for hard surface shapes.
If the drawing is not supposed to go over silhouette of mesh, vertex colors painting or texture painting are more practical than GP drawing.
GP strokes have the advantages to easily alter too straight perfect contours.