Less faces vs more solid model?

I’m modeling an item for a game. The model is very low poly box.
http://img545.imageshack.us/img545/8429/ammopack.png

Now I’m asking you guys, should I use less faces and make some parts of the model “separate” (Like the handle on the side) or should I just make the whole model just one piece? The model uses triangles.

Well to answer your question a wireframe would be needed.
But I guess, like the handles at the side, they are currenly part of the mesh and plate where the handle is has a high triangle count due to it.
In that case you´d most likely split it to save geometry but that is also dependend on the engine you use and the requirements to a textured object.

However for this particular type of box I´d need 16 Triangles =)
I´d bake a normalmap on a simple cube to absolutely shrink the geometry count, the box itself is nice for it because it has really a rather flat surface. Also has the advantage that you can use a subsurf for the high poly version and make some ornaments and stuff. The handle would be all round… ah well its called ammopack… so no ornaments =D

You could just try reshaping a simple cube to fit your model and then bake the normals to it from your model and see what the result turns out like. Once you’ve seen the process in action you can then probably actually increase the poly count on your current model and then bake a quality normal map onto a slightly improved cube.

There are quite a few tutorials about baking normal maps in Blender. An ambient occlusion map could be made too and combined with your diffuse map in Gimp or another program.

lol, i would have never come up with this… wait… :smiley:

Haha, guess I should read stuff before posting. :stuck_out_tongue:

Well, at least it proofs the point.
Normalmap´s a great idea for that box :slight_smile:

Floating geometry is best when there is a part which protrudes from the object by a sizeable amount, such as a door handle. You don’t want to use floating geometry for every detail. If you seperate small details there will be z-fighting between the floating geometry and the geometry behind it where the two are near parallel.

you model looks good, I would use that for a close detail then as the “player” get further away it changes to a cube using a DoF(Depth of Field) you could look into that.

I mean’t to say LoD(Level of detail) not DoF sorry bout that :o thank you Josip Kladaric for pointing that out

What AdobeMan wanted to say is LOD (Level of Detail) not DOF (Depth of Field).

yes lol thank you for correcting me :stuck_out_tongue: i feel stupid hahaha