Jump to content

  • Log In with Google      Sign In   
  • Create Account

jenny_wui

Member Since 12 Jan 2013
Offline Last Active Mar 21 2016 03:11 PM

Posts I've Made

In Topic: Some clarification needed regarding geometry shader

21 March 2016 - 02:35 PM

It seems to me some thing goes with the version. OpenGL 4.5 allows vertex shader output to be bypassed by the geometry shader (even when geometry shader is linked) and fragment shader can work with directly from the geometry shader though no output from geometry shader to fragment shader is made. But with OpenGL 4.3 it shows linking error as I mentioned above. Any way, I'd like to hear from others regarding this whether I'm right or not.
Thank you!


In Topic: Some clarification needed regarding geometry shader

17 March 2016 - 12:15 PM

I've encountered some unusual problem. Geometry shader is not set to NULL. In one platform when I run the program, it shows the following linking error:

 

Linking program
The fragment shader uses varying Position, but previous shader does not write to it.
The fragment shader uses varying Normal, but previous shader does not write to it.
 
 
In another platform it works fine, without showing any error.
 
I'm very much confused.

In Topic: Some clarification needed regarding geometry shader

17 March 2016 - 09:40 AM

I think I could not make you understand properly. I'm using geometry shader and it is linked to the program, but can the data  be sent from the vertex shader to the fragment shader by-passing the geometry shader? Is it any way possible?

 

 


In Topic: Some clarification about inversing rotation only around particular axes

01 July 2014 - 06:44 PM

Thanks for the suggestion and sorry for my poor knowledge. Actually I again want to write down some observation. As I said I have an opencv matrix. I would like to convert it to OpenGL matrix. As opencv is row-major, in opengl, I need to store in column-major order. I did that.

Next, Opencv y and z axis need to be reversed too. In order to get back the proper-oriented matrix in OpenGL, I multiplied the matrix with the following unit matrix:

|1 0 0 0|
|0 -1 0 0|
|0 0 -1 0|
|0 0 0 1|

I get the matrix with proper orientation ( I assume), but still with y and z axis vectors reversed. I know if I just inverse the sign of y and z-axis vectors, it will give me the desired result.

My question, what is the standard step of converting an opencv matrix to opengl matrix.
If anyone has done similar work, please let me know.

In Topic: Some clarification about inversing rotation only around particular axes

30 June 2014 - 06:38 PM

Thank you. I am trying to rotate an object by multiplying with a rotation matrix, but can't find it get right. That's why I was asking whether there is any difference getting transformation from basis vectors.

Actually I have opencv rotation matrix which I need to use for my OpenGL programming.I swapped the rows for columns to get rotation matrix to be used for OpenGL. Now I find the rotation around x axis is inversed. How can I make it correct? I know it is possible to make it correct by manipulating the basis vectors, but I need to know where I am wrong. Please let me know.

How to get the correct rotation matrix from OpenCV to OpenGL? I have posted this topic before but without any clarification.

Please shed some light if you have answer.

PARTNERS