Mateba.
mod
Alright, hope to get around to doing lowpoly/texture for this at some point. Modo render gives me boners. Just simple radiosity + blurred reflection. I might take some of the floaters and dig them out into the geo(the rectangle bits on the barrel for instance).
Replies
now back on earth is time for down splash
I agree that the handle could use some lovin.
Heres some wires.
Like the grip here for instance, would be a big pain to model, but pretty fast to whip up in PS.
Or this logo, its pretty easy to just copy/paste this into PS, and trace it, turn it into a normal and overlay it.
if you do model those things in, you can ask yourself if you are going to model in the scratches and dirt left on it.
Your doing a good job sir! like the Ivor handle better then wood.
Only thing that would be nice to add is the small gab where the thumb is on the grip.
And the render is just :poly101:
Nice gun sir. Can't wait to see it textured.
all systems go, prepare for downcount.
SupRore: lulz, if you were to see the wires for this little bit here you would have a heart attack.
Ok, had to add a few little details, for Per, or else he's going to cry about it.
I like this.
Sorry I know nothing of guns, that little safety switch (or whatever) looked like floating geometry in your wire but in this render it looks like you made it all one piece? If not the render does a good job with shadows & reflection.
I'm curious...what is it about Modo that is liked over max with polyboost?
As far as modo goes, its just what i'm comfortable with. I used lightwave for a really long time, and modo was just a nice natural transition from that. I've never really liked the way max works, it always seemed clunky to me. Anyway, i dont want to get into Modo Vs Max here, so just give it a try if you're curious. Its really an app you can learn in about 30 minutes.
I'll have to give Modo a try after that.
Polyboost definitely adds a level of interactivity that is missing from standalone max so I can see how you feel that way.
Now I just have to find some good tutorials for modo. Doesn't look like gnomon has any.
Anyway, looking good thus far. Floating geo if done right, you shouldn't be able to tell the difference...and Modo's render system deffinitely allows for you to be able to make it impossible to tell.
What happens to your modo in vista?
Nice looking model EQ
Great work EQ. Can you maybe post the wires for that three step part behind the cylinder? I'm curious how to do that.
I have a question about painting in normals with Photo shop. If you use Object Space normals, what do you use to convert a height map into normals? I thought nvidia doesn't support OS maps.
Yeah i dont really know dude, its a pretty simple render actually, all i did was turn on global illumination, and check the "blurry" tag for reflections in the materials. It just works out!
In some of the shots, some bits are pretty obvious, in that last shot, you can tell the 3 little circle bits are floating, they're brighter from not being occluded than what is below them.
Dole: Yeah, there are a couple things you can do, one that pior has mentioned doing is convert to TS, add your details and then convert back to OS (xnormal can convert).
Another thing is that, the front-facing side of your mesh will essentially be very similar to a TS map(as long as its mostly flat) and the reverse side will essentially be the same, but inverted. So you can use the same TS tricks here, and invert them on the opposite side, and it works out!
lol joking aside awesome work man cant wait to see the texture to this
http://dl.getdropbox.com/u/499159/dicks_render.lxo
If somebody could make something like that, gawd that would save so much time.
http://area.autodesk.com/index.php/forums/viewthread/10267/
I don't really model that way in max anymore but I could definitely have saved myself some headaches in the past. Oh well...max's crappy ability to handle stuff like this forced me to try and be a better sub-d cage reference modeler so in the end whatever.
I think what I was trying to get at was something like mental ray's "round corners" option...except a new method that could intelligently blend (fake) a bevel between a small piece of floating geometry (like your safety switch) and make it look like the sub-d's were worked into the surface it is laying on to appear as if it is all one piece even though it really isn't.
This dream cheating method of mine would also work on a curved surface. I don't know how that could be handled. Like if the floating piece was pushed into the surface of the other and then the user would have to select/define both parts in some way as intended? Pipe dream...would probably be a scripting nightmare...
3rd object?
Anywho I think you're thinking too small potatoes. I'm talking about a Kevin Johnstone-type of floating mesh (or many different shapes of floating meshes) pressed into a curved surface at different points all around...or multiple curves have you. Very complex sub-d's all around. Merging it all together with proper quads would take much longer than 10 seconds.
[edit] Also it seems like what you're asking for is very similar to how CAD applications work, from what i understand.
Yeah I think you've got the idea of what I'm talking about...except the cad application would create a nurbs surface leading to crazy unusable topology for sculpting...
I don't want to physically merge for bevels, just let the rendering look like it has...no matter how complex so long as the pieces intersect.
Anywho this probably has no place in your thread...I shall cut my bullshitting now...
You could create a library of floating elements, and then render displacement maps for all of them. With that done you could UV your highres mesh, and apply those displacements direction onto the uvs. Thus giving you "real" deformed geometry without any visible "floatyness".
Honestly to me, having objects floating is not a big deal, unless i'm making models for rendering, and not for normals.
Yeah I've tried that. Then I started experimenting with this using ZProject in zbrush. I'm just always on the pursuit for the best method that requires the least amount of time.
Anywho thanks.