Sign in to follow this  
Geometrian

Emulating gluUnproject

Recommended Posts

Hi,

I set up a view.
The modelview matrix is the identity matrix:[source]1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1[/source]
The projection matrix is 45 degrees, near 0.1, far 100.0, resulting in a matrix:[source]1.8107 0.0000 0.0000 0.0000
0.0000 2.4142 0.0000 0.0000
0.0000 0.0000 -1.0020 -0.2002
0.0000 0.0000 -1.0000 0.0000[/source]
By the documentation on gluUnproject, I multiply (MV*P)[sup]-1[/sup] by a normalized device coordinate to get an object coordinate. Because in my case, MV is I, that means that MV*P = P, and (MV*P)[sup]-1[/sup] = P[sup]-1[/sup]. Matrix P[sup]-1[/sup] is:[source]0.5523 0.0000 0.0000 0.0000
0.0000 0.4142 0.0000 0.0000
0.0000 0.0000 0.0000 -1.0000
0.0000 0.0000 -4.9950 5.0050[/source]
Now, this is a problem. In the next step, the normalized device coordinate is multiplied by this matrix to get the object coordinate. However, notice that any Z value in the NDC will [i]always[/i] be zero, because index 2,2 of this matrix, (MV*P)[sup]-1[/sup], is zero! Clearly, something is going wrong. Help?

Thanks,
-G Edited by Geometrian

Share this post


Link to post
Share on other sites
[quote name='Geometrian' timestamp='1312690020' post='4845654']
Now, this is a problem. In the next step, the normalized device coordinate is multiplied by this matrix to get the object coordinate. However, notice that any Z value in the NDC will [i]always[/i] be zero, because index 2,2 of this matrix, (MV*P)[sup]-1[/sup], is zero! Clearly, something is going wrong. Help?
[/quote]

Nothing is going wrong, you've just miscalculated something. With this matrix, z value in the result will only be zero if NDC.z is zero (and NDC.w is zero, which it shouldn't be). Yes, M2,2 is zero, but M2,3 is not.

Oh, yeah, and don't forget perspective division (divide result vector by it's w component after matrix multiplication).

Share this post


Link to post
Share on other sites
There is an app for that

[b]GluProject and gluUnProject code[/b]

http://www.opengl.org/wiki/GluProject_and_gluUnProject_code

Share this post


Link to post
Share on other sites
Ah thanks everyone! I see now that the problem wasn't that the zero was screwing it up. The problem was that I had forgotten to divide by the "w", which confused me, until I realized that the orders of magnitude differences between the GL code and my emulating code were due to my using 0.1 and 100.0 as clipping planes! Thanks for the tips,
-G

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this