Once upon a time I was quite passionate about game development. About four years ago at the height of my game development aspirations, I was involved in a number of activities. I worked with a team of students at my college to build a game using XNA. I wrote a basic cross platform 3d engine using OpenGL/GLUT, and I wrote a simple tank game for Android using OpenGL ES 2.0.

For better or worse, since that time I've been swept up in the winds of corporate profit, and for the past few years I've been mostly preoccupied with building corporate applications and websites and I've done very well by it.

Even so, I've been trying to build games on the side, and one of the major limiting factors has always been the acquisition of quality resources (models, textures, sprites, audio) to support the game itself.

Basically what I am trying to say is that given an individual whose greatest strength is the programming, and is mediocre at best at developing graphical and media assets, what do you think some techniques I could use in conjunction with a 3d scanner to help me produce realistic or at least decent 3d models and/or 2d renders of scanned objects?

The catalyst for this thread is that I recently came across a kickstarter for a relatively affordable high quality scanner called 'Robocular'. Aside from being excited about the project as a huge nerd, I am quite frankly intrigued by the possibilities introduced by having the ability to cheaply manipulate a medium and subsequently scan it as a model. Although a device like this seems most suited to scanning terrain assets, I can't help but wonder whether it may be feasible to utilize multiple scanner runs to produce model animations, or perhaps scan individual body parts and cobble them together, applying animations later by manipulating the pieces in Blender. Or perhaps simply using a physical object as the initial base for the model manipulation.

Either way it seems way to potentially save a huge amount of time compared to how long it would have taken me to produce a model myself. I know there are a lot of free or affordable resource packs online, but one problem I've always had with them is that they don't necessarily combine with each other well. I.E. That cheap pack of humanoid models just doesn't quite work with the vehicle pack I purchased, or the free weapon pack I found, etc. It's often the case that the art style or level of detail from one pack to another just doesn't match up... But like I said I haven't been very active in the game dev scene for a while, so I'm not sure if a 3d scanner is necessarily going to provide as much of a shortcut as simply purchasing model packs...

It's just that the concept is so cool and open ended that it's difficult not to be excited about it.

Basically I was hoping to spark a discussion on the practical applications of such a tool. Given the high precision of the scanner and the fact that it supports color, it seems like you could do a lot with it.

Some of the initial ideas I had are:

  • Play Doh or clay + Paint

  • Wooden models + Paint

  • Soap models + Paint

  • Construction Paper or paper mache' + glue/tape

  • Purchasing toys as base models, then augmenting the toy with clay or play doh to produce something unique

  • Casting objects in wax, then carving details in the wax shape

I'm confident that some of these techniques could be significantly faster than using the actual 3d modeling tools, at least to start off. Anyone have some killer ideas on how to utilize a 3d scanner to create assets?

  • \$\begingroup\$ Discussions and questions that would generate a list of equally-viable answers aren't a good fit for this site; see the help center for more. You can also find a list of places where this kind of question would be more appropriate on that page. \$\endgroup\$ – user1430 Jan 14 '14 at 14:23
  • \$\begingroup\$ I don't think that's necessarily true. For example, in the early days of 2d image manipulation, all discussion regarding the application of 2d image manipulation techniques could be regarded as opinion based... However as the technologies and user base matured, specific methods and 2d image manipulation tools have become dominant and accepted as effective. \$\endgroup\$ – M. Smith Jan 14 '14 at 16:32
  • \$\begingroup\$ I was on the fence about using "too broad" or "opinon-based" as the close reason, but it doesn't really matter. You are asking to "spark a discussion," which is precisely not what this site is for (see the help center). I think this is an interesting topic, and worthy of a lot of interesting discussion, but the role of this site specifically does not include fostering that kind of discussion (which I why I pointed out the list of places this question would be better asked). StackExchange is a Q&A site for specific questions with specific answers, not a general discussion forum. \$\endgroup\$ – user1430 Jan 14 '14 at 16:38
  • \$\begingroup\$ I suppose so. Even if there are methods of utilizing 3d scanners that are superior in a quantifiable wayl because no one is doing it , there's no way to accept an answer as being 'right'. I'll have to move this discussion to another arena, thanks! \$\endgroup\$ – M. Smith Jan 14 '14 at 16:43

Well, your idea sounds great, but unfortuanately am afraid it's only great in theory , the reasons am saying this are:

  • First of all you need actual real life models of each asset in order to produce their 3D counterparts

  • I highly doubt that this is actually cheaper and easier than modelling and drawing or even buying (or getting for free ) your assets, you also need to buy the 3D scanner and maybe other things.

  • So unless your game only have limited assets it will be much cheaper to buy or use (or even make) ready made models (example from TurboSquid).

The above are only the initial costs and we didn't even start developing the game, on the technical side you have other obstacles:

  • When your models are 3D scanned, they are not particularly ready for in game use. 3D scanners will output very complex meshes with millions of polygions with alot of redundant polygons, you need to use fairly complex algorithms in order to decimate and simplify the meshes.
  • When you are done with scanning and simplification, you to generate UV coordinates for texturing, which is usually part of the asset creation process.
  • 3D scanned meshed still need to be Rigged and Skinned for actual in game animation. I don't see other way to do this.

In conclusion I would say that using 3D scanned models is just another way to create 3D models with no apparent advantages. You will end up with mostly the same process just with different initial path.

  • \$\begingroup\$ Great answer. Though technically I was asking specifically about techniques to use - which assumes an established intent to use a 3d scanner. Either way it helps though, thanks for the reply. \$\endgroup\$ – M. Smith Jan 14 '14 at 16:40

Not the answer you're looking for? Browse other questions tagged or ask your own question.