Jump to content
  • Advertisement


This topic is now archived and is closed to further replies.


This is really driving me nuts

This topic is 5807 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

I have a function that generates texture coordinates, and every so often, it messes things up. About every 50th face that coordinates are generated for, I get floating point errors, i think. The flaw is very noticeable when viweing the face with in 3d, and when I save the floating point numbers to a file when i see one of the errors, this is output: "-1#IND" instead of an actual number. What the hell causes "-1#IND" to occur? Here is the function in question:
void CFace::GenerateTextureCoordinates()
    D3DXVECTOR3 vS( 0.0f, 0.0f, 0.0f ), vT( 0.0f, 0.0f, 0.0f ), vRef( 0.0f, 0.0f, 0.0f );
    m_vNormal = Normal( m_Verts[0].m_vPos,
                        m_Verts[2].m_vPos );

    if( ( m_vNormal.x == 0.0f && m_vNormal.y ==  1.0f && m_vNormal.z == 0.0f ) ||
        ( m_vNormal.x == 0.0f && m_vNormal.y == -1.0f && m_vNormal.z == 0.0f ) )
        vRef.x = 1.0f;
        vRef.y = -1.0f;

    vS = Normalize( GetCrossProduct( m_vNormal, vRef ) );
    vT = Normalize( GetCrossProduct( m_vNormal, vS ) );

    for( int y = 0; y < 4; y++ )
        // The final values here are occasionally producing -1#IND

        m_Verts[y].m_fU = ( m_Verts[y].m_vPos.x * vS.x ) +
                          ( m_Verts[y].m_vPos.y * vS.y ) +
                          ( m_Verts[y].m_vPos.z * vS.z );
        // The final values here are occasionally producing -1#IND

        m_Verts[y].m_fV = ( m_Verts[y].m_vPos.x * vT.x ) +
                          ( m_Verts[y].m_vPos.y * vT.y ) +
                          ( m_Verts[y].m_vPos.z * vT.z );

[edited by - Mulligan on January 19, 2003 4:27:37 PM]

Share this post

Link to post
Share on other sites
-1#ind iv a divison bu zero , division by zero is indetermined, check your dotproduct and cros products, when you are normalizing so be carfeull dividing everything by zero, you should add a check , if the length of the vector is zero , just add a little epsilon ( 0.0000001f ) this will prevent dividing by zero lenght vectors , byez and sorry for my english

Share this post

Link to post
Share on other sites

  • Advertisement

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

GameDev.net is your game development community. Create an account for your GameDev Portfolio and participate in the largest developer community in the games industry.

Sign me up!