How do you find the camera direction vector in Python?

Coding Python Support
1

I can’t find a way to get the direction vector for a camera in the API. Do I have to figure it out from the object matrix? Same goes for the up vector.

2

Window.GetViewMatrix() returns the view matrix. There are some other view goodies in that module.

3

Thanks, I’ll give it a go.

4

For camera objects, the direction vector is the Z axis. so, you’d have to get the Z-axis from the object, which I think is the third row of a 3x3 matrix (I’m not sure though).

joeedh

5

For camera objects, the direction vector is the Z axis. so, you’d have to get the Z-axis from the object, which I think is the third row of a 3x3 matrix (I’m not sure though).

joeedh[/quote]

Think this will work


mat= camob.matrixWorld
dir= Mathutils.Vector(0,0,1,1)*mat
dir.resize3D()
dir.normalize() # optonal
# dir is the camera direction
6

Is there an easy way to get the perspective “camera” / View direction ?

I’m trying to project rays from the view TM, but can’t figure out how. I thought I had it. But is not right. Using the view spaces I think and getting 3x3 matrix and multiply by vector of 0,0,1. So Z is pointing / aligned from view into scene.

Thanks !

7

I’ve been trying to figure this out yet, and think I’m close maybe, but still can’t get it working right…

Seems like if I use this code, I can create a Camera/Set an object to be aligned to the Perspective TM/View, but the Project is not going in the right direction/using the proper vector.

How can I get the Local Z Dir from the Matrix?

I don’t even really want/need to create a Camera, but was using it as a test.
Mainly just trying to Raycast from the ViewTM, onto another Mesh. Raycast from VIew.

cam = bpy.data.cameras.new("Camera_ShrinkWrapper")
cam_ob = bpy.data.objects.new("Camera_ShrinkWrapper", cam)
bpy.context.scene.objects.link(cam_ob)
									
cam_ob.matrix_world = area.spaces.active.region_3d.view_matrix.inverted()
   
mat_camera = cam_ob.matrix_world * mathutils.Vector((0.0, 0.0, 1.0)) #area.spaces.active.region_3d.view_matrix.to_3x3()
trans_world =  mat_camera
trans_world.normalize()

Edit: Lol… and just like that, I got it working…! I looked at some of my other code I had/forgot about.
Basically just had to convert to 3x3 seems like, and still multiply in the Vector.

vec = mathutils.Vector((0.0, 0.0, 1.0))
trans_world = (area.spaces.active.region_3d.view_matrix.inverted()).to_3x3() * vec
trans_world.normalize()
8

vec = mathutils.Vector((0.0, 0.0, 1.0))
trans_world = (area.spaces.active.region_3d.view_matrix.inverted()).to_3x3() * vec
trans_world.normalize()

I tried this code in blender 2.8 but I don’t fully understand the second line from the quote.
If I try bpy.context.area.spaces.active.(There is nothing here in blender 2.8)

My main goal is to get the perspective view location.

9

It has probably changed in 2.8. The spaces / viewport stuff. I haven’t tried to convert anything to 2.8 yet. So I may hit that as well. Maybe I’ll try this weekend if the weather ends up bad.