블로그 이미지
자신의 단점을 메꾸는 것을 단(鍛)이라 하고 자신의 강점을 갈고 닦는 것을 련(鍊)이라 하여, 두가지를 합친 것을 단련이라고 부른다. 붕대마음

카테고리

전체목록 (667)
참고사이트 (8)
Goal (4)
Travel (10)
My Life (105)
Game (35)
Game Review (7)
Game Plan (0)
Books (5)
English (1)
Optimizing (12)
Study (218)
유용한 것들_etc (44)
유용한 것들_func (20)
Unity (48)
Unreal (87)
작업장 (54)
RenderMonkey (6)
정리요망 (1)
따라잡기 시리즈 (0)
링크용 (0)
Total345,595
Today85
Yesterday123

PartialPrecision

Study/Graphics / 2011. 4. 21. 22:33

// D3DXSHADER_PARTIALPRECISION
//   Force all computations in resulting shader to occur at partial precision.
//   This may result in faster evaluation of shaders on some hardware.

이 옵션을 사용하면 shader의 float의 정밀도를 반으로 줄인다.
그럼으로 인해서 shader들의 연산 속도를 높인다.



// 그 외 기타 shader 옵션
//----------------------------------------------------------------------------
// D3DXSHADER flags:
// -----------------
// D3DXSHADER_DEBUG
//   Insert debug file/line/type/symbol information.
//
// D3DXSHADER_SKIPVALIDATION
//   Do not validate the generated code against known capabilities and
//   constraints.  This option is only recommended when compiling shaders
//   you KNOW will work.  (ie. have compiled before without this option.)
//   Shaders are always validated by D3D before they are set to the device.
//
// D3DXSHADER_SKIPOPTIMIZATION
//   Instructs the compiler to skip optimization steps during code generation.
//   Unless you are trying to isolate a problem in your code using this option
//   is not recommended.
//
// D3DXSHADER_PACKMATRIX_ROWMAJOR
//   Unless explicitly specified, matrices will be packed in row-major order
//   on input and output from the shader.
//
// D3DXSHADER_PACKMATRIX_COLUMNMAJOR
//   Unless explicitly specified, matrices will be packed in column-major
//   order on input and output from the shader.  This is generally more
//   efficient, since it allows vector-matrix multiplication to be performed
//   using a series of dot-products.
//
// D3DXSHADER_PARTIALPRECISION
//   Force all computations in resulting shader to occur at partial precision.
//   This may result in faster evaluation of shaders on some hardware.
//
// D3DXSHADER_FORCE_VS_SOFTWARE_NOOPT
//   Force compiler to compile against the next highest available software
//   target for vertex shaders.  This flag also turns optimizations off,
//   and debugging on. 
//
// D3DXSHADER_FORCE_PS_SOFTWARE_NOOPT
//   Force compiler to compile against the next highest available software
//   target for pixel shaders.  This flag also turns optimizations off,
//   and debugging on.
//
// D3DXSHADER_NO_PRESHADER
//   Disables Preshaders. Using this flag will cause the compiler to not
//   pull out static expression for evaluation on the host cpu
//
// D3DXSHADER_AVOID_FLOW_CONTROL
//   Hint compiler to avoid flow-control constructs where possible.
//
// D3DXSHADER_PREFER_FLOW_CONTROL
//   Hint compiler to prefer flow-control constructs where possible.
//
//----------------------------------------------------------------------------

'Study > Graphics' 카테고리의 다른 글

빛의 감쇠 계산  (0) 2011.06.22
조명효과 - 램버트 확산 조명  (0) 2011.05.05
PartialPrecision  (0) 2011.04.21
Gemetric Modeling Systems  (0) 2010.11.27
3D 모델링의 개요 및 종류  (0) 2010.11.27
shader 2.0 & 3.0  (0) 2010.08.28
Posted by 붕대마음

댓글을 달아 주세요

최근에 달린 댓글

최근에 받은 트랙백

글 보관함