Taken from another app so the semantics are not identical to ogre - I still found this helpful so for anyone interested here it is.

Cg Parameter

Name

Description

world

Contains the world transform matrix, i.e. the matrix that transforms a vertex from the model (local) coordinate to world coordinate system.

view

Contains the view transform matrix, i.e. the matrix that transforms a vertex from the world coordinate to view (eye) coordinate system.

worldView

Contains the concatenated matrix of world and view transform matrices, i.e. the matrix that transforms a vertex from the model (local) coordinate to view (eye) coordinate system.

worldViewProjection

Contains the world-view-projection matrix, i.e. the matrix that transforms a vertex from the model (local) coordinate to clip space (projection) coordinate system.

worldIT

Contains the inverse transponate of the world transform matrix, i.e. the matrix that can be used to transform a vertex from the world coordinate to model (local) coordinate system, under the assumption that the scaling is uniform.

viewIT

Contains the inverse transponate of the view transform matrix, i.e. the matrix that can be used to transform a vertex from the view (eye) coordinate to world coordinate system, under the assumption that the scaling is uniform.

worldViewIT

Contains the inverse transponate of the world-view transform matrix, i.e. the matrix that can be used to transform a vertex from the view (eye) coordinate to model (local) coordinate system, under the assumption that the scaling is uniform.

worldViewProjectionIT

Contains the inverse transponate of the world-view-projection transform matrix, i.e. the matrix that can be used to transform a vertex from the clip space (projection) coordinate to model (local) coordinate system, under the assumption that the scaling is uniform.

projection

Contains the projection matrix, i.e. the matrix that transforms a vertex from the view (eye) coordinate to clip space (projection) coordinate system.

projectionIT

Contains the inverse transponate of the projection transform matrix, i.e. the matrix that can be used to transform a vertex from the clip space (projection) coordinate to view (eye) coordinate system, under the assumption that the scaling is uniform.

Note: when using the inverse transponate transforms (xxxIT), you have to reverse the matrix multiplication, i.e. change the order of the arguments to mul() function. For instance, if you want to transform a vertex P from the view coordinate to model (local) coordinate system, you have to use this line:

float4 oPos = mul( P, worldViewIT );

instead of

float4 oPos = mul( worldViewIT, P );