I was toying around with a concept, to see if it would be viable given my current blender knowledge of near ZERO (only used Blender to make flat cutouts for Bomber Cat’s backgrounds, no actual 3D modeling.)
I did this robot thing. My goal was to do something simple and quick, see if I was able to do it fast enough to justify a very plain looking 3D project.
It did take me a few hours, followed some tutorials, made use of mirroring, had some hell accommodating the UV mapping (plain unwrapping was hell, nowhere near the simple “project from view” I used for my simple cutouts before.
After being done, I decided to do same extremely simplistic texturing in Paint.net (a free app that’s barely more functional than the plain Paint.exe, basically it’s Paint with layers.) I was horribly disappointed with the look and about to call it failure.
Decided to give a try to Ambient Occlusion, a technique I have heard a lot but never properly understood. Blender seems to make it easy on me, once I got a… Ambient Occlusion mapping? I took that gray scale texture and used it to darken my original color copy and TADA… the difference is huge…
Sure, it’s still plain but wow, I actually sort of think I can make something with this now! There are still some artifacts showing all over the center, think due to the mirroring. I guess once I’m ready to apply the mirror modifier things I can re-do the AO map.
Anyways, just wanted to express my own shock! Can’t believe I don’t see AO mentioned more often on 3D discussions!
I know its plain looking but… hey its my first 3D model!
Give yourself some credit. I’ve seen fully trained college art students that didn’t put in the effort to make something this good. Tell me this thought what kind of art background do you have?
Few years in high school as a comic book artist wannabie (horrible horrible stuff I still keep around in a portfolio)
2 years of arts study (1995-1997)
15+ years of SQL and VB development. (not art but thats the bulk of my life)
8 months developing Bomber Cat (formerly Misu Misu Kaboom.) There is a lot of 2D art there, mostly made with vector graphic software (Xara Photo Designer, badly named product since it’s not photo editing software, but GREAT vector design software.)
I did blender work for that game, but as noted, only flat cutouts. I have attempted modeling in blender before, but didn’t get past toying around with mirroring and extruding. This is my first finished model and my UV experience was all from those cutouts for Bomber Cat (mostly used in palm trees and rocks to avoid overloading the game with transparency.)
1 week developing Chicken Panic (pending approval, very simple high score game.) Only work I ever done in serious 2D animation, with the chicken walking animation.
Other than that, a LOT of doodling with pens in stickies. It is actually dangerous to give me a sticky stack; it will be used up in a few days if there is a pen nearby.
Even though I have about 6 yrs experience in Lightwave 3D, I was able to pick up Blender in a matter of days, like 3 days, after I finished these tutorials, now I can model anything I want in Blender.
Don’t be afraid to tackle 3D modeling, it really is easy, takes a little bit of patience… After two months of starting to learn Lightwave 3D 6 yrs ago, I was able to model my head and it wasn’t hard, USE BACKDROPS
Unfortunately, page restricted at work so cant visit that one during downtime. I’ll check it out from home, though.
BTW, two questions:
I understand all game engines work with triangles, but noticed Blender works with quads by default. Is there a way to force Blender into always using triangles? Although I guess that would make extruding harder…
Is there a way to see in Blender how many triangles and/or quads are being used in a model you are working on?
You don’t have to work with triangles the model is going to end up looking the same like in blender just faces will have a diagonal line across them which basically splits a face and makes a triangle.
Cool, thanks. I figured it would be converted (at least by Unity importing) but figured I’d be able to optimize polygon count if I was able to force triangles here or there.
So, my current drone thingy is at 164 faces/169 vertex. Does that sound too high for something of that level of detail?
Not really, but by the looks of your model, it could be further reduced and optimized.
If you’re not too concerned about texturing, you could try to seperate the mesh into several different meshes in the object. This is more efficent than a singular object that has alot of edges connecting to parts of the mesh.
Right now I am using a few meshes in that robot. There is one doing the boxy body, another doing the cannon, another one for the red “wing” thingy that holds the cannon, another one for the tiny tube that goes through it, and finally one you can’t see at the top of the head, like a hat (where a chopper-like helix was intended to “plug”)
That helix was never modeled though.
BTW, just jumped into edit mode and the value now says Fa: 1-82 instead, i am guessing 1 is the number of selected faces and 82 the total. Selecting all makes it 82-82… I take it the 164 is due to the mirroring modifier… I guess that’s where I’m losing a few faces, most faces in the box that touch the center are being duped when they would ideally just extend through to the other side.
Is there a way around that, or it’s an acceptable loss that I can’t do anything about anyways?
Mirror is splitting your object in half that’s why you got 164poly when you would have 82 if there was no mirror,so just apply the mirror modifier and try to delete the edges splitting it but you will maybe lose the UV coordinates and end up screwing up the whole model .BTW First before doing it duplicate the original model and work on copy.
Oh well, was just a learning experiment. I should be able to handle the UV much faster this time around now that I know how to easily reposition vectors.
One crazy thing: I have been trying to move more than one vector at a time in the UV editing mode. If I select one vector, I can right click to move it easy. If i pick multiple, i can easily use S to scale them, or R to rotate them. But I cant seem to move more than one vector at a time from the UV editor. Only way I managed to do it is by sliding the mouse over the X and Y values of the UV Vertex display box in the GUI.
So: is there a way to just move all highlighted vectors in the UV editing mode?
164 is a very low amount of polygons… is this going into a mobile game?? if so, if that’s like the main character then make it as detailed as you want, but you can optimize later. When you’re done modeling, apply the mirror modifier then go back to edit mode and you can select two polygons and merge them. Also, i think there’s a decimator modifier that reduces polys.
if you select multiple vertices, edges, or polygons. Hit “G”, that’s the move tool and you can move them to where you want them then click the left mouse button… Also, if you don’t like using the right mouse click for operations you can change that in user preferences… again, watch the tutorials i posted earlier, the teacher goes over everything you’ve brought up in this thread
Cool, thanks. I’ll be taking a look at the vids as soon as I get home later today. Slow day at the office so been trying to poke around with Blender.
Right now this thing is a raw concept of what I may use as main character in a future game. But just based of what I learned doing Bomber Cat, I think it’s wise to keep that as a temporary asset for now and work on the rest of the game. By the time I’m closer to finishing the project I’ll be much better with the tools and able to do a much better looking main character.
My current line of questions on the poly count is mostly to have an idea if I’m being optimal or wasteful given level of visual detail. I hear one of the biggest challenges making models is to make them look as you want without ending with way too many polygons.
So… other than Ambient Occlusion… what other techniques are there that I can “bake in” to make a model look cool without much work?
