4x4 matrix triangles or individual verts?
Author
I am wondering if 4x4 matricies typically contain values for a single vertex, or does the matrix define a triangle (3 verts). I would assume that since its 4x4 it would contain vert info for a complete triangle.
Matrices generally store transformation information which are then applied to an arbitrary number of vertices. A matrix by itself can store scaling, sheer, rotation and translation, but inherently does not store point information.
I'm by no means an expert :), but I think that most people use a 4D vector or 3D to contain a vertex individually (most likely 1) and then use a 4x4 matrix vector to descibe operations done on that vector. Though, I suppose any vector , say (1, 1, 1, 1) can be broken down into
1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
Does that answer your question?
1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
Does that answer your question?
Author
So verts aren't stored in matrix form when manipulating them? Let's say I have a triangle with the following points:
PX0 = 4, PY0 = 9, PZ0 = 1
PX1 = 0, PY1 = 0, PZ1 = 0
PX2 = 7, PY2 = 5, PZ2 = 2
Are they, if this is how verts are organized when manupilating them, arranged like so:
P0 P1 P2 n
X 4 0 7 0
Y 9 0 5 0
Z 1 0 2 0
n 0 0 0 0
PX0 = 4, PY0 = 9, PZ0 = 1
PX1 = 0, PY1 = 0, PZ1 = 0
PX2 = 7, PY2 = 5, PZ2 = 2
Are they, if this is how verts are organized when manupilating them, arranged like so:
P0 P1 P2 n
X 4 0 7 0
Y 9 0 5 0
Z 1 0 2 0
n 0 0 0 0
A vertex is treated as a vector that is multiplied by the matrix. While you can batch vertex multiplication by storing multiple vertices in a matrix before being multiplied by the transformation matrix, the mathematics are no different than interpreting each as a single vector and multiplying them in turn.
Author
Thanks for the clarification. I'll go and experiment with it and see what I get but I'm sure I will be back with more silly questions about it.
Quote:Original post by Caitlin
Are they, if this is how verts are organized when manupilating them, arranged like so:
P0 P1 P2 n
X 4 0 7 0
Y 9 0 5 0
Z 1 0 2 0
n 0 0 0 0
Ok, the way this is put is a bit confusing to me, but I'll try and see if I can answer it with this general explanation.
if you had three points, say (4, 0, 7) (9, 0, 5) and (1, 0, 2) and you wanted to do something to them , you would manipulate them by applying a 4D matrix to them. This gets kinda tricky because those things have only three points. I don't want to confuse you more (because I was in this boat a little bit ago), but most people usually just add a "1" to those 3d points at the end (4, 0, 7, 1), etc. Now, say you wanted to do something to those vertices. You would just simply apply whatever transormation you like to each one.
here, in example, is something I use in my own code to modify a vertex:
[source lang="cpp"SHYR_INLINE ShyrVec3 ShyrMat3::operator*( const ShyrVec3 &vec ) const { return ShyrVec3( mat[ 0 ].x * vec.x + mat[ 1 ].x * vec.y + mat[ 2 ].x * vec.z, mat[ 0 ].y * vec.x + mat[ 1 ].y * vec.y + mat[ 2 ].y * vec.z, mat[ 0 ].z * vec.x + mat[ 1 ].z * vec.y + mat[ 2 ].z * vec.z );}Granted, this example is just a 3D matrix and 3D vector, but you could apply the concept to anything. This Matrix transforms the vector by multiplying it by whatever I want.
If you would like some references, check these out:
http://www.gamedev.net/reference/articles/article695.asp
http://www.upl.cs.wisc.edu/~psilord/docs/local_axis.html
The second is a bit tricky to follow, but very informative. Let me know if I can help you anymore in any way.
Author
Oh, ok thanks. That's how I thought it worked, and had it working before, but then confused myself. :)
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement