|
|
|
|
#1 |
|
Banned
Join Date: May 2007
Reputation: 0
Posts: 204
|
I have a lot of fun making abstract stuff with brush geometry, and I've successfully made a dome that looks pretty good through use of 140 wedges, but that's really a lot of wedges. Anybody know a better way to make one more efficiently? Also, I think this could be a forum for anyone to show off cool stuff they've made with brushes.
|
|
|
|
|
|
#2 |
|
Guest
Posts: n/a
|
Quote:
http://forums.facepunchstudios.com/s...26&postcount=1
|
|
|
|
#3 |
![]() Join Date: Oct 2004
Reputation: 0
Posts: 279
|
models. A model would definitely be the most efficient, and it would probably look the best b/c you could make it way more detailed, but I've never made a model. i have no idea how to make models either.
I only use brushes, but then again, I've never tried to make a sphere. EDIT, zyklus, the problem with using displacements, is it won't turn out round. It will always be a bit lumpy depending on the original geometry that you're displacing. For instance, if you use a cube and displace it, you will end up with a round cube, but still a cube. If you use a tetrahedron or something with more faces, it might be a bit less noticeable, but you will still see it lumpy. Last edited by Latrosicarius: 06-03-2007 at 05:39 PM. |
|
|
|
|
|
#4 | |
|
Guest
Posts: n/a
|
Quote:
Technically, you can never have a completely round sphere in computer graphics. |
|
|
|
|
#5 | |
![]() Join Date: Oct 2004
Reputation: 0
Posts: 279
|
Quote:
I looked at your link. Nice job creating that displacement program. You actually got a cube to look completely spherical. However, to do so you had to use such a high subdivision power that the resulting "sphere" has an enormous amount of triangles. If it weren't for the fact that displacement brushes continue to look lumpy unless they have a ludicrously high amount of subdivisions, you could actually get a object to look just as spherical with MUCH less triangles. Even though displacements are rendered faster than normal brush geometry, I'd have a hard time believing that the shear amount of triangles in a "spherical" displacement doesn't counteract that efficiency negatively. Again, models are rendered very efficiently, and you can have a model that looks almost perfectly spherical, that has much less triangles, because it doesn't have the properties of resembling an original brush shape like a displacement. See the following pictures for comparison between a geodesic sphere capible of being modeled and inserted as a prop, vs your displacement. Notice that they both can appear very spherical. ![]() If you want, you could double or triple the triangle count on the geodesic sphere to make it look absolutely round, and you'd still have less polygons than the displacement. |
|
|
|
|
|
|
#6 |
|
Guest
Posts: n/a
|
Well, I was just showing off how detailed it could be. It doesn't have to be that many triangles (http://img337.imageshack.us/img337/5226/spheresmlo2.jpg). I had it this detailed for a reason, because in the case that Megiddo asked me to build, higher resolution would be the integral testing for some of his scripts. The link posted here, uses only 1 cube with 6 displacements. In the link above, I used 296 cubes with 384 displacements (btw, power 4 is the highest resolution a displacement can go; if you want to go any higher, you'll have to make multiple brushes) If you really want a poly-count, then a power-4 displacement has 512 triangles. 384 displacements with 512 triangles is 196,608 triangles. The only thing that this slows down is hammer
![]() EDIT: Well, I guess I can't argue; models would render faster
Last edited by zyklus: 06-04-2007 at 10:54 AM. |
|
|
|
#7 |
![]() Join Date: Jun 2007
Reputation: 5
Posts: 91
|
Blacksmith, it depends. If you're back in goldsrc land, you can use models, but they don't light very well. If you're resigning yourself to doing it with brushwork, then wedges are the only way to go
![]() I usually just stick to modelling anything complicated, rather than brush models. |
|
|
|
|
|
#8 |
![]() Join Date: Oct 2004
Reputation: 0
Posts: 279
|
^ well that link you just posted does seem much more reasonable in triangle count. I see that you were going for the max in that first displacement sphere pic
:P |
|
|
|
|
|
#9 |
|
Banned
Join Date: May 2007
Reputation: 0
Posts: 204
|
Really, the main thing I was goin for is trying to figure out how they made the dome for the dm_domebox_tsgk map in HL2: DM. Is there a console command that shows brushwork? Mat_wireframe shows it but with all of the clipping lines as well so I can't see a thing.
|
|
|
|
|
|
#10 |
![]() Join Date: Sep 2006
Reputation: 9
Posts: 2,691
|
Erm, if you give me a download, i'll decompile it. Or decompile it yourself with VMEX.
|
|
|
|
|
|
#11 |
|
Banned
Join Date: May 2007
Reputation: 0
Posts: 204
|
Actually, I tried decompiling it but it's "protected from decompiling". Maybe one of the VALVe team could help?
|
|
|
|
|
|
#12 |
|
Guest
Posts: n/a
|
mat_wireframe 3
shows the wireframe for visible objects only and not everything else. That might clean up your view enough. |
|
|
|
#13 |
|
Banned
Join Date: May 2007
Reputation: 0
Posts: 204
|
hah! works great. thanks!
|
|
|
|
![]() |
|
||||||
| Thread Tools | |
| Display Modes | |
|
|