Assets - Continous mesh vs seperate

[Guppy]

With regards to modeling assert what is faster for the leadwerks engine - multiple meshes or a continous mesh ( with sligthly higher poly count ) ?

 

A highly simplified example would be a "T" asset it could be moddled either as two separate "beams" or as 4 "connected boxes"

 

Normally I'd not even consider the separate meshes a viable option constantly thinking "this need to be able to be unwound into gl_triangle_strip's but after watching

 

 

(it's very long video relevant part is around 12 minutes in ) he basically builds a wooden bookcase type thing out of individual boards which is admittedly very fast - but every fiber of my being is screaming for a retopo.

 

So which is it?

 

( kinda hoping it's just my dated knowledge of opengl showing and that the difference if any is neglible )

 

ps. I realize he is using a different engine and all, but the question is still relevant I think

[tournamentdan]

There are two problems with intersecting mesh. How you plan to use your models will decide how major those problems are. Problem 1. If you plan to have your intersecting model bend with animation. It will not look very good where the mesh intersect. Problem 2. It depends on how many times you plan to instance the model. Say if you make a vegetation model like grass. You are going to instance that model a lot. Especially if you have a large world. Then you will have a great deal of wasted polygons. But if your model is a shelf you probably will not have a bunch drawn. P.S. I am not sure if Occlusion culling will help with underlying mesh. But if it did. Problem 2 would not matter.

[Josh]

I don't think it matters, as long as you don't have faces that are on the same plane inside each other. (This would cause z-fighting between the two faces.)

 

Generally, when you export you want to collapse everything into a single mesh. If each beam in that shelf were a separate object, it would take dozens of rendering passes, but it will only take one if it's collapsed to a single object.

[Guppy]

Thanks for the swift replies

 

@tournamentdan

There wouldn't really be any waste polygons as anything that is fully hidden would be removed during modeling

 

 

@Josh

Great, that will make things a bit easier - and 'collapse to single object' sounds like something the blender integration could do automatically

[Flexman]

Generally, when you export you want to collapse everything into a single mesh. If each beam in that shelf were a separate object, it would take dozens of rendering passes, but it will only take one if it's collapsed to a single object.

 

What if an object with different parts only used a single material?

[beo6]

That is exactly one issue i had with an external model i tried to import.

The model showed up as a tree of different parts and i couldn't apply the single texture correctly to the model in the editor.

If i applied the texture to the parts seperately it looked just wrong.

 

Not sure how it is supposed to work.

[Josh]

What if an object with different parts only used a single material?

That still means an extra draw call.