Kamujin's picture

Column Major?

Is Matrix4 (the float version) column major?


Comments

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
Inertia's picture
Kamujin's picture

I think less efficiently when I am tired. I think I see my mistake now.

Can you check my order here?

In row major, modelViewProjection = projection * modelView? (not modelView * projection)

Is this right?

Inertia's picture

ModelView * Projection if memory serves right. There's a whole chapter dedicated to matrix operation at http://www.glprogramming.com/red/chapter03.html

in OpenTK.Math matrices are concatenated like this

C = A * B; // results in C containing matrix A transformed by B
 
// should be the same operation as:
 
GL.Matrixmode( ModelViewMatrix );
GL.LoadTransposedMatrix( A );
GL.MultTransposedMatrix( B );
GL.GetFloat( TransposedModelViewMatrix, out C.row0.X );
Kamujin's picture

Yeah, I'm not using immediate mode though. I'm sending it to the shader as a uniform.

I wasn't confused until I started using the OpenTK math libraries with OpenGL via shaders. I just have to get it all sorted out in my head.

Thanks for the link. Its been a while since I've played with matrix ordering, so I am trying to get it straight in my head again.

objarni's picture

I use this "mind model" when thinking of matrix ordering:

newVec = matrix1*matrix2*...*matrixN*oldVec

.. so matrix1 is applied 'last' and matrixN 'first', on the oldVec-vector.

Kamujin's picture

Objarni, your "mind model" is row major, correct?

objarni's picture

My mind model comes from linear algebra.

To me multiplication between matrices is mathematically defined something like:

R(r,c) = (A*B)(r,c) = A(r,:)*B(:,c)    // for relevant r, c

.. where

  • R is the result matrix
  • A(r,c) means element on row r, column c of matrix A
  • A(r,:) means row r of A (a.k.a. row-vector)
  • B(:,c) means column c of B (a.k.a. column-vector)

The definition could be read aloud something like "the element on row r column c of the result matrix equals the scalar product of row vector r of A and column vector c of B".

Multiplying a row and a column vector means:

              |b1|
|a1 a2 a3| * |b2|  =  a1b1 + a2b2 + a3b3   // similar to scalar-product of two vectors
             |b3|

I'm not entirely sure what you mean by row-major; does it have anything to do with the ordering of the elements in a computer representation of a matrix using an array?

Kamujin's picture

I think thats row-major notation.

There is information on row-major vs column-major matrices at the following link.

http://lists.apple.com/archives/Mac-opengl/2000/Dec/msg00045.html