june gloom on 6/10/2011 at 10:00
Quote Posted by Koki
And so is America.
You're an awful person and by extension so is Poland.
Renzatic on 8/10/2011 at 08:54
The other day, I came upon the realization that my life doesn't have nearly enough tedium in it. After realizing that, I came to the conclusion that that's pretty damn tragic, and I need to do something about that.
So I decided to make yet another old creepy house model. This time in glorious high poly detail. And by that, I mean I'll have to make board, every tile in the roof, every window, every crack in the wall, every little throwaway object by hand. I wanted tedium, and holy shit did I get it.
(
http://dl.dropbox.com/u/3018396/Farm_House_IRL.jpg) This is the inspiration. I love this old house. Wish I could buy it, but it's sitting on about 800 acres of fertile farmland, which costs well over a million bucks. Probably the reason why it's been abandoned for so long.
(
http://dl.dropbox.com/u/3018396/Farm_House_Modo.jpg) ...and this is the current work in progress. I started it yesterday, and...damn...it's time consuming. By the time I'm done, I'm gonna have a rough brick foundation around the bottom, more boards along the edge of the porch, all the tiles on the roof (I'm dreading that), and all other kinds of details.
(
http://dl.dropbox.com/u/3018396/Farm_House_Modo_2.jpg) Here's an even more dramatic shot.
Guess I'll be done with it in a month or so.
Forever420 on 8/10/2011 at 13:26
Quote Posted by Yakoob
Oh yea. I love MS for those kind of reasons. People shit on MS because they are the big bad corporation but honestly, they treat their devs so damn well. With pretty exhaustive documentation, amazing tools (VS, debugger, PIX), good cross-platform APIs (DX, XAudio, XNA, .Net) and even options for indies to develop on their proprietary console (via XNA), its is really nice developing for their platforms (especially compared to stuff like Wii or PS3 development).
They do treat devs decent, even indie gets easy to use toolkits.
Much better than the sham that apple is.
demagogue on 8/10/2011 at 14:37
Quote Posted by Renzatic
Guess I'll be done with it in a month or so.
Worth suggesting that when you're done, it'd be nice if you could rig it in a game engine with some nice atmospherics, a lot of that overgrowth, etc, to let people walk around it, soak it in, and be inspired. I just think if someone spends that much time working on a 3D model, it can really be put to use putting it in a space that really shows it off and makes it a real place.
Renzatic on 8/10/2011 at 21:30
That's actually my long term plan. I hit up Judith about a week ago, asking if he wanted to do a project in UE3, either for the start of a new game, or just to make something that looks cool, and learn a little bit in the process.
The house above is pretty much the first step of that. Hopefully, it'll help me address my biggest modeling weaknesses...mainly that I absolutely suck at efficient UVing, and can't bake high poly to low poly for crap because of it.
Renzatic on 9/10/2011 at 05:19
(
http://dl.dropbox.com/u/3018396/Farm_House_Clay.jpg) First clay render of the night. Well, actually the last. I just spent about 3 and a half hours laying down and organizing fake slate tiles.
That's pretty tragic. I didn't even get paid for it.
Judith on 9/10/2011 at 14:26
Wow, that house looks nice! Although I can see from the wireframe that you haven't been thinking about tiling the wall sections. Of course you can do all that in low poly version, but it's easier to take that into account from the beginning. In general it's better to plan your architectural stuff as a tileset and use game-scale values, even if it botches the proportions a bit, while you can use real-world proportions for other props.
Btw. Still working on that facade:
(
http://imageshack.us/photo/my-images/805/hzld.jpg/)
Inline Image:
http://img805.imageshack.us/img805/3019/hzld.th.jpgRenzatic on 10/10/2011 at 20:55
Quote Posted by Judith
Wow, that house looks nice! Although I can see from the wireframe that you haven't been thinking about tiling the wall sections. Of course you can do all that in low poly version, but it's easier to take that into account from the beginning. In general it's better to plan your architectural stuff as a tileset and use game-scale values, even if it botches the proportions a bit, while you can use real-world proportions for other props.
I dunno. Tileable chunks work alot better with large buildings and cityscapes than they do with unevenly detailed houses like this. The only way I could get it to work would be to make it tile in incredibly small portions, otherwise it'd end up looking too generic.
Right now, I'm thinking about just normalmapping the roof tiles, and leaving the main structure boards as they are for the final model. Save for a couple of pieces, the building itself doesn't have too much subsurfing going on, so I could export most of it as is and keep it at or below 4000 tris fairly easily. UE3 could handle it no problem. I just have to work in a few smoothing groups, and maybe edge bevel a couple of corners, and it'd be about good to go.
Whoa. Is that in the TDS engine, or in proper UE3? Whichever it was, you did a spectacular job on exporting the normalmaps. It looks almost indistinguishable from your HP source.
Wish I could do that :mad:
Bakerman on 14/10/2011 at 03:26
I've taken a break from working in Torque to play with a bit of procedural generation. You can see the fruits of my effort at (
http://www.ug.it.usyd.edu.au/~dbuc6168/Xn/). It's a little Python/WebGL app that generates a random arrangement of rooms connected by corridors, as you might see in a Roguelike. It's all rendered in 3D in glorious primary colours (which I may insert the option to change :p). The performance is pretty balls, which I would expect from 3D rendering
in a browser, but it's very cool nonetheless.
A crappy demo shot (in case your browser doesn't have WebGL, or you're too lazy to click the link):
(
http://imageshack.us/photo/my-images/818/43909239.jpg/)
Inline Image:
http://img818.imageshack.us/img818/6240/43909239.th.jpgI'm slowly working on upgrading the algorithm - at the moment it's very simple, and doesn't really generate anything too exciting-looking. I want to have a look at adding things like clustering, and more advanced pathfinding to connect doors on each level (at the moment, it's a simple BFS; I'm thinking a spanning tree would be more appropriate, but I also want to add the ability to reach levels that only have one door. In that vein, a check for overall connectivity of the level would be nice, but also extremely tedious with the simple data structures I've set up :p.
Also, I should do something about the wild proliferation of elevators. It's partly an algorithmic problem (elevators are very likely to fit in a given spot) and partly a case of me defining lots of elevator shapes relative to how often they should appear.
Also, I was very disappointed at the level generated by seed 451. Tak, tsk.
