Game Development Community

4 Questions About Torque Engine.. please help

by Erifeht · in General Discussion · 11/26/2006 (3:14 pm) · 11 replies

I have 4 general questions about the game engine, any help would be greatly appreciated.

1. What is the difference between TGE (Game) and TSE (Shader)?

2. Is there a limit to the size of the "world" model or terrain? If so, what limits you?

3. Does TGE use world coordinate systems (UCS) similar to ACAD/3D Studio? In other words, if a model is built in a specific location in 3D Studio, will it import into TGE at the same location, or do you have to move/relocate it?

4. How accurate are the shadows cast by a Sun object? If a sun path is built, will the shadows cast similarly to raytraced (or vector based) shadows? For example, will an object's shadow move, change in length and represent the projected outline of the object on the ground?

Since Torque won't answer these question via phone I am desperate to find the answers here. Thanks in advance.

#1
11/26/2006 (3:27 pm)
1 - Two largest visible differences are the graphics pipeline and the terrain engine. TGE uses an older version of both and has certain limitations. TSE (aka TGEA) uses a pixel and vertex shader graphics pipeline and uses the Atlas terrain.

2 - TGE, using the older terrain engine, you're basically limited to 2 kilometer x 2 kilometer worlds. TGEA / TSE uses the Atlas engine which, although there is a limit, it's beyond what you'll ever need.

3 - You build the model in your 3d program at 0, 0, then when you place it in Torque, you have to move it to the specific location that you need it to be.

4 - Only if you use dynamic shadows, which as far as I know doesn't work on all model types (i.e. buildings that you can walk inside), but dynamic shadows are computationally expensive and should be used sparingly. Your mileage may vary depending on optimizations you might do and your target GPU.

Hope that helps...
#2
11/27/2006 (8:04 am)
You mention that there is a limit with TSE regarding the world size. This is an important concern for me, is there a specific maximum size?

Another important aspect is World Coords. Is it possible to develop some type of API which will allow TGE to work with world coords?
#3
11/27/2006 (8:26 am)
The maximum size is the size of the physical media it is on. Not that you would want to take up a couple of tetrabytes of someones disk array...

What do you mean by world coords? You mean as exported from, say, Max? You would need to build it into the exporter or at least export mission data. Of course, Max is not the best choice to use for all aspects of world modeling (since you would need to export to map through Game Level Builder and then import it into the world-builder and place it accordingly. So, if GLB exports the model exactly, you could export the mission file and load it accordingly once the conversion process has finished.

I'm not sure how messy it would be, though.
#4
12/04/2006 (12:39 pm)
Thank you for the prompt and helpful replies. In response to the last post...

Assuming that we have to import our models manually starting at 0,0. Once the building is in place can it later be replaced using a reference to it's coordinates? ie: using place holders for items that will be added later.

An additional question:

1. If I am building a city full of buildings that are NON-Enterable should I be using .DTS models? keeping in mind that these buildings will vary in size from 2 stories to 20+ and in some cases these buildings will have a considerable amount of texture and other various details.
#5
12/05/2006 (2:48 am)
Hi Eri

To your first question, as I am not a programmer, take this with a grain of salt. But I do think you can make a script/code to check the coord of an item and then use those to place the new model.

If the Buildings are not enterable and you do not need very precise collisions, I would model those Buildings as .dts. Because DTS can have more detail, smoothing and better uv texturing.
#6
12/05/2006 (11:05 am)
Should I be concerned about DTS files being heavier in terms of processing? I will have a large number of buildings present.

Thanks.

p.s I am currently looking for some programmers and designers who are familiar with torque to assist in the development of a new project. If anyonen is interested please contact me Erifeht@gmail.com and we can discuss it further detail.
#7
12/05/2006 (11:19 am)
Should I be concerned about DTS files being heavier in terms of processing? I will have a large number of buildings present.

Thanks.

p.s I am currently looking for some programmers and designers who are familiar with torque to assist in the development of a new project. If anyonen is interested please contact me Erifeht@gmail.com and we can discuss it further detail.
#8
12/05/2006 (3:03 pm)
The collision mesh is what would be cpu intensive. If you stick to standard box meshes it will be perfectly fine and you will get better framerates as opposed to .dif structures. Also lighting time will be lowered drastically.
#9
12/05/2006 (3:19 pm)
Quote:
If you stick to standard box meshes it will be perfectly fine and you will get better framerates as opposed to .dif structures.

Not sure what you mean by this, but taken at it's face, this is almost never true. At worst case, dif shapes render at the same speed or slightly better than dts, and best case are orders of magnitude more efficient to render and cull.
#10
12/05/2006 (3:39 pm)
I think he is referencing the collision system as slowing down attainable framerates.

using dts collision is not as robust as the diff collision.
so with many more objects using dts you will see a difference in performance compared to say running those same objects as diff

dts probably render's faster than diff.
#11
12/05/2006 (3:48 pm)
Quote:
dts probably render's faster than diff.

No, they don't. That's my point.