Game Development Community

Large Model Test

by Randy Hearn - Magnum · in Collada Test · 02/18/2009 (8:54 pm) · 2 replies

This is a rather large F16. First views shows Collada imported into Houdini... Second is Collada imported into T3D..

AS you can see in the second pictures lots of faces left off. Not sure if its becasue of the large size. It did call up in Houdini, but very few game engines have successfully imported the file.

NOTE: Nothing is setup in this file, strictly a straight Collada export from MAX.

i60.photobucket.com/albums/h15/kd4vcu/Game/F16-2.png
i60.photobucket.com/albums/h15/kd4vcu/Game/f16-1-1.png

I can't post the actual model for obvious reasons. If GG is interested in it let me know.
Randy Hearn - Magnum's profile picture
Torque Owner Randy Hearn - Magnum
Default Studio Name

About the author

Technical Product Designer (Mechanical Design) for Boeing for over 25 years working with CAD and PLM systems. I have a Associates Degree in Business and a partial B.S. in Game and Simulation, just couldn't see paying the costs for some of the classes.

Recent Threads

  • I am back..... And of course Need help getting started again.
  • C++ Edit source button not showing
  • GUI on objects
  • Vehicle Datablock - Nevermind
  • Active AI
    • NextNext Thread
    PrevPrevious Thread
    More from Collada Test


    Associate Chris Robertson
    #1
    02/18/2009 (9:55 pm)
    Hi Randy,

    I think what is happening here is that you are hitting the upper limit to the number of triangles-per-mesh that Torque (currently) supports. Torque uses 16 bit vertex indices to render triangles, which means you only get approx 21-22k triangles before things start going missing.

    I think there are plans in the works to handle this properly (ie. to split meshes up into chunks that can be processed using 16 bit indexing).

    Side note: if using a DTS exporter, it's only around 11k triangles since they are stored as signed 16 bit.


    Torque Owner Randy Hearn - Magnum
    Default Studio Name
    #2
    02/19/2009 (9:05 am)
    It is in that range, this is a test model and I am waiting for the reduced poly version that should end up in the 10-15K range. So that is what I suspected as well.

    But hey, it looks better than my last attempt:) At least I can work with this one...