Sprite Tool Kickstarter - Spriter

[Hearteater]

I just wanted to throw out a mention for this tool called Spriter.  I am in no way affiliated with this team, but it looks like a really useful tool for 2D game developers, and at $25 for the full version (via the Kickstarter) it's insanely cheap.  I am just starting to work with the tool itself, and although I need to write my own library to play back the animation files in XNA, it looks like I'll at actually have some source to work from.

[x4000]

Pretty impressive, to say the least!  It wouldn't work with absolutely every kind of animation, and that sort of thing does come with an increased draw-call cost, but with proper texture atlasing and even using more complex vertices than just textured quads, you could get some ridiculous performance out of this.  If I were a betting man, I'd say that the first implementations of this sort of thing will still perform better than traditional sprite stuff, but at only 30% or less of the actual performance you could get if you really integrated it right on the engine side.

It's not something I could use for AVWW because of the nature of how I create those images, but this is definitely a tool that makes me salivate in a lot of ways.

[eRe4s3r]

This isn't actually all that new. Ubisoft has developed their own ART tool for this, where you draw sprites and deformation based animation is done instead of spriteframe animation and generally I have seen a lot of attempts to make a really good procedural animation system for sprites (all failed) hence they all used keyframed deformation points as animation platform.

Because as you say, there is certain important animations that you need to draw by hand, like a simple turning around of the character and what if 2 characters interact with each other (lets just say, a HUG ?) or pick pocketing.

Now this is totally superior to standard sprite animation though, at least for characters and landscape, for effects.. not so much. Their procedural motion system is likely just an IK system adapted to deform sprites gracefully. But even that would be pretty nifty

[Hearteater]

What is really nice is Spriter's format is basically just an XML file (plus images) that describes all the animations.  But beyond that, it lets you add seemingly arbitrary things like collision boxes, action points, and sound effect start/stop points, that Spriter doesn't do anything with but dutifully includes in the XML.  So on the game side, you can process that information on a per-frame basis to generate additional effects.

I should have some experience with it over the next two weeks so hopefully it works out well.  I'll be replacing basic sprite frame animations so my drawing does get more complex, but the bottleneck for our team is art, not code, so anything that lets art assets get generated more efficiently is a huge plus.

[x4000]

but the bottleneck for our team is art, not code, so anything that lets art assets get generated more efficiently is a huge plus.

Not to go off topic, but what are you working on?

[Hearteater]

A turn-based tactical RPG with procedurally generated worlds, but a character that persists across worlds.  A cross between FFT, Diablo, Disgaea, and AI Wars if you will.  Your character searches out a world to explore (creates it like you create a game in AI Wars), then you explore it.  Everything you collect stays with your character to explore the next world.

But we are in the very early stages and being only a two-man team both of whom work full-time, we've got a lot on our plates.  But the preliminary coding work I've gotten through is very promising.  I've knocked out the first milestone of basic combat map display in XNA (using 2.5D basically) and sprite animation.  Next up was going to be a more robust animation system with XML scripting, but after we saw Spriter we've decided to see if that will meet our needs.

[keith.lamothe]

A turn-based tactical RPG with procedurally generated worlds, but a character that persists across worlds.

That sounds like it could become a favorite of mine

[Hearteater]

It will be my first fully complete game project.  We've really been working to keep on top of things and not let it slip into the background.  I have to say you guys at Arcen have inspired both of us to finally get serious with a game.  It took us awhile to pick which project we wanted to work on, but this one really caught our imagination and was still within the scope of something I think we can accomplish given a year-ish of part-time work from both of us.

[x4000]

Very awesome to hear!  That really sounds great all around, and reminds me a lot of how I got started with my own stuff.  Best of luck to you.

[keith.lamothe]

Indeed, I hope it goes well   Let me know if you run into tough problems or whatever.  Not that I'd know better in general, but I may have already done it before

[Hearteater]

I'm sure I'll come across something odd at some point that I'll try and float by you guys.  No matter how well you plan that always seems to happen when coding.  Generally when you least expect it.

[x4000]

You're telling me!  The number of those in AVWW has been really high, and this is my third released game.

[keith.lamothe]

Yea, you don't plan to prevent gotchas, you plan to prevent the easier gotchas so you can focus on the hard ones.