3 ms·
I wonder how his ideas could apply to scenes in games, where after all we have far fewer constraints (we can make any landscape or plant or tree), while we have
by mmooss 2y ago
I wonder how his ideas could apply to scenes in games, where after all we have far fewer constraints (we can make any landscape or plant or tree), while we have other constraints that Olmsted didn't have (rendering).
For example,
Through all the designs that Olmsted created there runs one dominating and consistent conception. He always insisted on the subordination of details to an overall composition whose strongest and fullest effect was to act unconsciously on those who viewed it. He vigilantly guarded against distracting elements that would intrude on the consciousness of the observer. In the process, he simplified the scene, clearing and planting to clarify the "leading motive" of the natural site and heighten the effect of a particular quality of nature.
And at a greater level of detail ... do game studios work at this level of detail for the artwork (edit: I mean, do they think through the artwork conceptually at the level Olmsted did)?
When employed the style of the Picturesque, Olmsted introduced "complexity of light and shadow near the eye" to heighten another aspect of nature—its mystery and bounteousness. To achieve a sense of mystery, he used a variety of tints and textures of foliage that made forms indistinct and created a constantly changing play of light and shadow. At the same time, he planted profusely to secure greater richness and lushness of growth than nature would produce unaided. He planted one layer upon another, beginning with ground cover, then shrubs, then trees above them. To complete the effect he often added creepers that covered the trunks and branches of deciduous trees, keeping them green with foliage even in winter.
- reneherse 2y agoI too am interested to know if anyone has experimented with these ideas in a virtual environment. It seems there would also be an opportunity for enriching procedurally generated worlds with these concepts of landscape (and shelter) patterning. And not only Olmsted's ideas. Though I'm not up on the state of the art, a couple other useful landscape and architectural theories come to mind: -Jay Appleton's "Prospect and Refuge" theory of landscape painting analysis (and its extension into architecture by Grant Hildebrand) [1] -Chris Alexander's "A Pattern Language" and other books. I wonder how one would go about training a model using data from existing environments, paintings and photography that exhibit these patterns? [1] https://www.researchgate.net/publication/286123657_Prospect_and_refuge_theory_Constructing_a_critical_definition_for_architecture_and_design https://www.researchgate.net/publication/286123657_Prospect_...