Since I have questions involving both compositing, animating, modeling and rigging, I really don’t know which section of forum to post the thread in, and this section seems to be the most generic one to me, I just put it in here.
I am a apprentice Blender user. As in “learning how to model in 3d environment”. And my goal is that I can make game-compatible models. And today I ask my questions here in the hope of getting enlightened. Here are my questions:
I am currently working on a tank model. (I uploaded my blend file here). I followed a tutorial on nystic and managed to make the track to animate when I move the tank forward. But the point is that I want to use this model in video games. And I am aware that when I make the animation for the tank to work with game engines, the tank must stays at its position, and only the track animated. And when I put the tank in the game engine, the game engine plays the animation and translate the tank’s position forward, creating the illusion of the tank moving forward with the track movement synchronized. So how do I only animate the track without having to move the tank in Blender? Right now I can only make the track move in Blender by moving the tank.
As you guys can see, in the model, the tank is not a single mesh, but rather dozens of mesh. I want to ask if it is a good idea, and should I keep it that way or nail the tank down to 3-4 meshes only, the turret, the chassis, the gun and the tracks. I ask this for the sake of low-poly modeling.
Is there any tutorial about composite basic? I know that the composite has nothing to do with game engines and video games model, but I just want to know more about Blender.
The same goes to texturing and UV. Is there some kind of “UV and Texture for dummy” somewhere on the net? I tried to find some easily understandable tutorial and I even tried truckloads of tutorial myself, but I still can’t get a grab on it.
About the armature. After laying down a skeleton for my model, for example a stickman, I made the model to deform just right. And with it I mean when the hand bone is rotated 30 degrees, the hand rotates 30 degrees. But then I found out I forgot the feet bones, so I add 2 of it. And after apply auto weighting, the skeleton just… screwed up. The mesh always deforms more than I rotate the bone. What is the cause? Which parameters affect this process? How can I fix it?
That’s it for now. Thanks for your time
Those are a lot of questions! To get a good response, post one question at a time.
texturing and UV
Texture is a 2D image. Flat. 3D object has surfaces and volumes. It has depth. But the surfaces of 3D object are made of flat polygons. What UV system does is to relate each polygon of a 3D objects face to a given texture using UV coordinate. Same as XY coordinate system, they just call it UV coordinate system:
Do you see how the modeler snipped faces off the 3D model and located it on 2D image space? It’s all about “flattening” 3D object in to 2D space. Some feature like the head is difficult to “flatten” and so it can end up in a strange shape.
Thanks for the help with UV and texturing. But how does one know which part of the texture file to lay on the UV? Do people do it manually or is there some sort of tool that lets you paint an image on a mesh?
Also, thanks to daren for that. I tried it and it’s much more under control
But how does one know which part of the texture file to lay on the UV? Do people do it manually or is there some sort of tool that lets you paint an image on a mesh?
Yes you can do all that with Blender. Which method you use depends on what you have.
When you have good image to UV map over the object, you can just lay UV map right over it. Look at this tutorial to get an idea on that; this is with older Blender version but the process is the same for the current version of Blender:
Others may disagree, but I think the best approach is to think in terms of: one part, one mesh. If two differing shapes are actually joined together in real life, then join them and deal with the topological challenges that come along with that. But I’m not talking about ‘bolted together’ or ever ‘welded together;’ think of a pair of scissors to see what I mean, differing shapes all part of the same object.
If, for instance, with your tank, the turret has a hatch that opens and the hatch has a handle that turns and the handle is mounted on a plate and the plate is bolted onto the hatch with four bolts, leave them all as separate objects, especially if you’re planning to pose those parts.
To make selection simpler, you have the option of drag-n-drop parenting in the Outliner. Make the bolts children of the plate, the handle also a child of the plate, the plate a child of the hatch, the hatch a child of the turret and the turret a child of the tank body. Then if you have to move the whole thing around for any reason, select the tank body and you’re ready to translate.
Once you have all the parts built, if it’s too ‘heavy’ vertex-wise, then decide where you’ll cut back. The bolts holding the plate onto the hatch might be better as a normal map, for instance. Also, any surface that won’t show can be deleted and the resulting edges ‘tied off’ with tight edge loops to keep them from warping too badly. If you don’t know what I mean by ‘tied off’ let me know; it’s not a standard term and I don’t know if there is a term for it. Basically, it’s a way of forcing sharp edges in preparation for subsurfacing an object.
But overall, modeling each and every part that will show in its entirety is a great exercise. Going further and modeling all the parts of the internal combustion engine would be an even greater exercise, but perhaps more than you might want to bite off.:spin:
