concept - SSAO, Opmizations and pipelines using OpenSceneGraph -
i have questions ssao technique implementation:
does need sec (or more) pipeline every geometry? mean, found tutorials , stuff give directions without entering in farther details.
is there optimization possible? i'm using osg , i've got impression if send textures cpu , throwing 1 time again gpu isn't best solution possible.
is possible create shaders generate texture samples depth in buffer , send sec pipe line using quad screen, colors, depth of scene , depth tests? i'm using osg , couldn't find how in documentations.
in general, ssao best suited beingness implemented part of deferred shading approach. strictly forwards shading approach possible, still require 2 rendering passes, , ssao can added sec rendering pass of deferred shading engine. in ssao, need finish depth buffer of scene able calculate occlusion, short reply section 1 of question yes, ssao requires 2 rendering passes.
note in deferred shading, although there 2 rendering passes, complex geometry (i.e. models) rendered during first pass, , sec pass made of simple polygon shapes rendered each type of light. you're suggesting in section 3 of question.
with regards section 2 of question, when set correctly, shouldn't need move intermediate textures cpu , gpu between 2 rendering passes; simply create first rendering pass's textures available resource sec rendering pass.
openscenegraph concept ssao
No comments:
Post a Comment