SimoPI

05-03-2010, 10:51 AM

I have this sphere implementation:

float x = (float) (cos(phi) * cos(theta));

float y = (float) sin(phi);

float z = (float) (cos(phi) * sin(theta));

taken from a source code.

I read: to calculate sphere tangents i must do the partial derivative of x, y, z (in respec to theta). So i should have:

tx = -cos(phi)*sin(theta)

ty = 0

tz = cos(phi)*cos(theta)

if i'm right. But reading the code, here is the adopted solution:

float tx = (float) (sin(phi) * sin(theta));

float ty = (float) -cos(phi);

float tz = (float) (sin(phi) * cos(theta));

Why?

Thank you for the answers and sorry for my english!

float x = (float) (cos(phi) * cos(theta));

float y = (float) sin(phi);

float z = (float) (cos(phi) * sin(theta));

taken from a source code.

I read: to calculate sphere tangents i must do the partial derivative of x, y, z (in respec to theta). So i should have:

tx = -cos(phi)*sin(theta)

ty = 0

tz = cos(phi)*cos(theta)

if i'm right. But reading the code, here is the adopted solution:

float tx = (float) (sin(phi) * sin(theta));

float ty = (float) -cos(phi);

float tz = (float) (sin(phi) * cos(theta));

Why?

Thank you for the answers and sorry for my english!