1
Feature requests, suggestions / Use geometry to represent cameras and lights instead in 3DS scene instead?
« on: July 07, 2007, 05:38:42 PM »
Hi, I'm very new to Wintermute but experienced with 3D. I am interested in making a 2.5 adventure and I am working through the documentation.
Like many others, I am struggling with the issue that the scene geometry has to be in 3DS file format with cameras and lights included. Of that, the biggest problem is the camera and light information. I have several inexpensive, easy-to-use programs that can export 3DS geometry and textures but no lights or cameras. I have Blender and your WME/3DS export Plugin ( http://wiki.dead-code.org/wakka.php?wakka=BlenderHiddenGeometryExport ) but Blender is very hard to pick up and I'd much rather use the inexpensive modeling programs I have.
So this led me to think about the whole issue of how a 3D scene is created and used for a 2.5D experience and I want to ask if Wintermute can be made to only need camera and light position and orientation information in the scene geometry file and the rest of the information (Field of view/Focal length for cameras and colour, intensity, angle and range for lights) can be created and manipulated in Wintermute?
Currently, a user creates a detailed 3D scene in any package and renders out images to use as backgrounds and sprites. They then have to create a low-poly proxy scene that matches up to the detailed scene. Wintermute determines what the geometry is for in the scene by how it is labeled: areas the 3D actor can walk on start with the label walk_*, blocked off areas start with the label blk_*and so on.
What if, when building my proxy scene, I could align a simple cube to where the camera is and label it "camera_*" and put other cubes where the lights are and label them "light_*". Then if my exporter only supports geometry in 3DS export but not camera and lights, I have a low poly scene geometry file that still has the information about where I placed the camera and lights and how they were oriented. If Wintermute could then use this position and orientation information as a starting point for creating the rest of the scene camera and light information, I could do the rest of the adjusting within Wintermute (just like I can adjust camera field of view right now)
If Wintermute could do this, I wouldn't need to worry about learning Blender (and relying on a script that could be broken in the next release of blender or python) or buying some expensive 3D software that I can't afford, and anyone could use the modeling software that they like/have because all they need to do is export 3DS geometry.
So, is it possible for Wintermute to eventually do this?
Like many others, I am struggling with the issue that the scene geometry has to be in 3DS file format with cameras and lights included. Of that, the biggest problem is the camera and light information. I have several inexpensive, easy-to-use programs that can export 3DS geometry and textures but no lights or cameras. I have Blender and your WME/3DS export Plugin ( http://wiki.dead-code.org/wakka.php?wakka=BlenderHiddenGeometryExport ) but Blender is very hard to pick up and I'd much rather use the inexpensive modeling programs I have.
So this led me to think about the whole issue of how a 3D scene is created and used for a 2.5D experience and I want to ask if Wintermute can be made to only need camera and light position and orientation information in the scene geometry file and the rest of the information (Field of view/Focal length for cameras and colour, intensity, angle and range for lights) can be created and manipulated in Wintermute?
Currently, a user creates a detailed 3D scene in any package and renders out images to use as backgrounds and sprites. They then have to create a low-poly proxy scene that matches up to the detailed scene. Wintermute determines what the geometry is for in the scene by how it is labeled: areas the 3D actor can walk on start with the label walk_*, blocked off areas start with the label blk_*and so on.
What if, when building my proxy scene, I could align a simple cube to where the camera is and label it "camera_*" and put other cubes where the lights are and label them "light_*". Then if my exporter only supports geometry in 3DS export but not camera and lights, I have a low poly scene geometry file that still has the information about where I placed the camera and lights and how they were oriented. If Wintermute could then use this position and orientation information as a starting point for creating the rest of the scene camera and light information, I could do the rest of the adjusting within Wintermute (just like I can adjust camera field of view right now)
If Wintermute could do this, I wouldn't need to worry about learning Blender (and relying on a script that could be broken in the next release of blender or python) or buying some expensive 3D software that I can't afford, and anyone could use the modeling software that they like/have because all they need to do is export 3DS geometry.
So, is it possible for Wintermute to eventually do this?