Hi all, really quick question, don't want to be wasting your time: I'm rendering a .x model using C++ DirectX9, using a D3DXMESH object (well, a LPD3DXMESH object to be exact). I would like to move this mesh, but not the camera (as there are other objects in the scene). What is the correct way to do this, (a) lock the mesh's vertex buffer, increment (x,y,z) of each vertex by the new value, unlock then render again (b) somehow at the render stage offset the object by a certain amount (x,y,z) using a Direct3D function (I don't know how this would work) (c) call a function on the mesh so that DrawSubset() renders the object at a different relative 0,0 location, using a Direct3D function (Again, I don't know how this would work) I'm not sure I know what I'm doing here; all I'm certain is that locking the vertex buffer, incrementing all the values, unlocking then rendering seems to me like a very, very costly way of doing things. Any help appreciated :) Thanks in advance

Quote:Original post by dpadam450
3D transforms. use a translation.

Thanks, I'm actually looking for more of an answer to as why ID3DXMesh itself doesn't encapsulate a translation function, perhaps some typical methods to overcome this, or some kind of explanation as to which tree I should be barking up, regarding translating a ID3DXMESH.

That's perfect! Thankyou so much, you've been such a great help, Sc4Freak :D

My thinking made the whole thing seem really odd previously, but knowing that makes a lot of things fall into place in my head now.

Thankyou again :D

WOOHOO!! BTW just to let u know it works perfectly :)

Thanks again :)