Foreign Art Collection

[dei]

I browsed by a CC-Art-Site recently: http://opengameart.org/ and thought why do we create everything by ourselves? And why don't we make a collections of usable (CreativeCommons or things like it) 3D and 2D art-links for Terasology? So I'm starting this now (Didn't find something in the forums).

For example I found a really beautiful minecart there: http://opengameart.org/content/mine-cart-set

 

[overdhose]

another head for the miniions : http://opengameart.org/content/cute-dragon

an army of head - only miniions. I had actually been looking around myself to find some 3d models which I could use, however I find it very time consuming to actually find something usable, but I might be taking a wrong approach. And while some objects might be reusable like crates etc, I doubt it will be a source for pc / npc models, we are already experiencing how hard it is to decide on a unified theme / style.

Oops almost forgot : many of the free models have a high poly count, A lot higher then the low poly models proposed so far

 

[dei]

I'm writing this from a developer's viewpoint, so I'm not very good in aestetics. I only think a predefined nearly accurate model is better for a intermediate miniion-model than the example-monkey from blender .

This is right: There aren't as many CC-licensed model-sources around. But contacting the artist of restrictive licensed material he will eventually contribute his model for our game or it even could end in a good collaboration with him.

 

[Cervator]

That's a pretty neat site / idea, I love the concept as a developer thinking about libraries. While it would probably work for placeholder art, I don't know if using a site like that would be a solid idea in general for serious games? Would seem pretty awkward to see the exact same dragon in two completely separate games, tho I suppose that might be unlikely

Poly count and style are other issues, but I like the idea of looking for appropriate stuff and contacting the artist(s) if any might be interested in contributing. And yeah, using any suitable stuff like that for placeholder art instead of BlenderMonkey - can't argue with that.

I wonder, would it be a possibility be to just lower the level of detail on CC art like that, for instance the rail + mine cart? Or does that take just as much work as just creating something from scratch? Although again we get back to being stuck solidifying our target art style.

 

[woodspeople]

Idea: block-built art. Say instead of getting artists to contribute and commit to one graphic style (hard to do since each artist rightly prides him/herself on unique style), instead concentrate on making that workshop thing where you can shrink down an area of say 9x9x9 blocks to a small sub-block model? Then we can ALL be artists, and THAT becomes the unified graphic style. I'm not a professional artist but I am an amateur one, and I'll bet I could churn out some cool looking creatures with a block workshop. I'll bet lots of people could say that. I mean no offense to the serious artists here, but this a bottleneck that could be removed by building a different-shaped bottle.

 

[glasz]

Thats basically going for the voxel character approach of CubeWorld :

I'm not sure where you see a bottleneck, the production of assets is not in itself a problem at this point. Choosing a graphic style is not either, we just need to make choices. The voxel approach is a possible choice. But characters and creatures are better aprreciated when they are animated, wich raise the questions of what software does voxel-based animation...

 

[Cervator]

That's a very good point. It has been mentioned in passing before, but I haven't made the connection to the artistic style.

Yeah, it would be a closer match for CubeWorld, and might not be appropriate for all models, but I could imagine building A'nW's dwarf one block at a time in-game - tho texturing might be trickier. Not to mention it would need to be crazy tall for a full model like that, I was more thinking about architecture and miniaturization from block shrinking. Animation would probably also be lacking for in-game created models, unless we can do something smart with skeleton-flagged blocks or just find a way to embed Blender...

However, we may indeed want to consider what user created content capacity we can get to (I hope for a lot, would love to see some creature or moving gadget ability - like design your own train cart) when thinking about the art style. So a plus for "more blocky" probably.

I recall ironchefpython desiring an in-game mod editor down the road for simple text-based stuff and I could see modifying properties of blocks and such pretty easily. Anything building actual objects would be a great match. SecondLife's tools might offer some inspiration there

 

[overdhose]

well the character editor makes me think of another game, I believe it was 3d dot hero
while I never played the game, I have seen tutorials about how to make a sonic model etc for us e in the game. From what I remember, you had a couple stances you needed to create (step forward, step backward, jump and attack) or something of the kind. I had also seen a vid of cubeworld (at least as far as I remember, can't immediatly find it) where you had to find this giant turtle who had a spin attack etc. But again, noob as I am, I wonder how resource intensive such an approach is, not that I want to sound to pessimistic, I would love to see a similar approach in Terasology.

PS : found it :

 

[glasz]

However, we may indeed want to consider what user created content capacity we can get to (I hope for a lot, would love to see some creature or moving gadget ability - like design your own train cart) when thinking about the art style. So a plus for "more blocky" probably.

A little in game voxel editor to create your own custom items would be cool indeed, and no need for anims in that case.

 

[woodspeople]

The bottleneck is, if you have to be a bona-fide artist to make art to be used in-game, there may be a long wait to get a nice long list of animals/creatures/minions/NPCs etc. And wanting them all to have the same style is an even smaller bottleneck, since no two artists have the same style, naturally. My thinking is, it fits a sandbox game-for-and-by-the-people to make it possible to build art, not just wait for it to be handed down by the experts. I am unfamiliar with CubeWorld, but we are already deep into mashups here so mashing in one thing isn't any worse than mashing in another... Cervator adding textures could be done with a painter-tool, and we have metouto's great textures to start with ... I'm all about shortcuts and getting things out there in the hands of the general populace ASAP.... is where I'm coming from.

 

[woodspeople]

On animations: maybe the plus of being able to make your own creatures balances out the minus of the animations not looking all that great. Right now the bobbing monkey head minions don't have any animation and they still manage to be pretty cool just in how they move around and respond to what you do. Also there may be possibilities for procedural animation: like, in the workshop you can "paint" blocks with a limited set of animation behaviors: this block bobs up and down, this block turns side to side, this block turns up and down. Also maybe some compromise is possible, where the voxel-based workshop can make simple procedurally-animated creatures, but real artists can do more with it by making labeled animation sequences using it? And/or import them? Ramping up while letting in the masses ... ? Just thinking out loud...

 

[glasz]

To be an artist is not about expertise, its about being interested in doing that kind of work, and willing to spend the time necessary to do the work until its satisfying. People who are given tools to create art easily often produce crap, not because they are lacking the expertise, but because they are not that interested, not to the point of going beyond the disapointing first try.

On animations : i'm not saying its not possible, just saying that that part might be trickier.

 

[Cervator]

The shortcut angle again seems a valid point. "Casual" artists like that might not see their work very far, or add animations, but I get the idea they'd still enjoy being afforded the option. And if you could re-import particularly interesting existing creations into the editor or easily "fork" them to then make a better version, and then a better version yet, etc...

In-game created animations definitely would be a long ways down the road, if ever, but it isn't like we currently have normal animations. It'll be a while before we get there, maybe August/Septemberish. SL does seem to support user-created animations, and not exclusively out-of-game: http://wiki.secondlife.com/wiki/How_to_create_animations

CubeWorld is reminding me about that idea to shrink the blocks slightly to gain a higher level of detail than the traditional cubic meter. I wonder how that game handles the huge number of blocks - if it isn't just from saving on colorizing voxels rather than texturing? Pinging begla or Immortius for ideas - IIRC the sheer number of vertices is a big part of what eats memory, rather than textures, maybe CW merges them in large numbers?. If we could pull off smaller blocks (would probably require introducing vertical chunks too) it would also make the in-game creation of objects easier. Or at least they would appear smaller and more detailed

Nice turtle fight, I like the impact to the terrain. Another thing that's easier with smaller blocks.

Also: Don't you love how we just totally fly off on a collective tangent, topic-wise?

 

[Immortius]

An optimization that can be made with identical (same texture and lighting settings) adjacent square surface (so if you have a bunch of grass blocks next to each other in a flat surface) is you can handle rectangular groups of those squares as one large rectangle (2 triangles) instead of 2 triangles per square. Needs a bit of shader support to maintain the texturing, but doable.

The limitation with that is we'ld need to avoid using vertex lighting effects (possibly replace the shadowed corners with SSAO). Also the world would need to predominantly be composed of cubes to get the full benefit from this - enough to support such granularity anyhow.

There is also the possibility of handling distant chucks as point clouds - 1 vert per square, down from 6 - that someone mentioned before?

 

[overdhose]

I like the impact to the terrain.

just wanted to point out : I don't see it destroying terrain. I think you should see the top layer as grass, and his attack just destroys the grass on top, I don't think it destroys dirt though

 

[glasz]

Here i think the troll dig holes in the ground.

 

[Cervator]

I was a little uncertain at first too, I think the trick is that it evenly destroys exactly one layer of cubes regardless of terrain elevation changes. So you don't get the very obvious "break" with already sloping terrain and it just looks like grass went poof. You can see it when it triggers on a flat area tho, it makes a nice round indentation.

 

[Immortius]

I think it is only the grass cubes. But grass is terrain too. So you both are correct.