Download

Basic animation set

Sorry if this derail the conversation - look I have a very small team as well - yes we are using mocap as well I have made threads about this.

Look you can say a walk cycle is generic and not worried about copying what other games look like. That’s not what I was saying about unique requirements to the game.
No matter how many cookie cutter setups you have your gonna need custom anims, its just the way it is. Not always but more than likely. It has nothing to do with a walk cycle.

Does the character have something in there hands - is it big - can it potentially clip into the player - does this affect his walk cycle - does he interact with the environment.

Even if some cookie cutter gets you 75% of the way there, your gonna need some skills and understanding to complete your project. Thus if you are already going to have to do some customer stuff and you already have spent the time to get to that point, why wouldn’t you do it yourself?

Time… oh well game development takes time I agree anything we can do to cut down on that would be great. Do I think all the tools we have to use to do animation and combine everything is convoluted as ****, hell ya. There could defiantly be a ton easier ways to generate animations using tools starting with some soft of fill in the blanks rendering creation more natural than what is there right now with the frame by frame ****.

But then we are talking about making the tools to make animation better - not necessarily just saying here are your animations and handing someone a set. And again I didn’t say I was against the sets - go for it by all means. I just said I don’t get why until they make better tools otherwise your gonna have to do it yourself anyhow.

Regardless, most projects get hung up and blocked on animation WELL before ever needing custom anims. Usually around the point where you need full set of basic locomotion and interaction animations. If you’ve got a mocap setup and a team making animations for you, that’s awesome…maybe this thread isn’t really for you then, though? Not everyone’s got a team of animators and mocap rig in their back pocket, after all.

Exactly my thoughts.

If we build a good collection of community animations then this will probably include “pick up”, “sit”, “walk with big sword”, “talking hand gesture” etc. before too long anyway… I really don’t think there are that many completely unique situations. With a working locomotion set up to start off with, we can easily make little tweaks to bend the arms a bit further away from the body etc. in minutes rather than days. Plus, we can use blend spaces, kinematics and scripting in the engine to mix and match animations and adapt them in real time for the game situations.

And, as n00854180t says, we can worry about all that once we have the game well under way and playable. Most games get derailed long before the point where it matters if a hand pokes through a body once in a while. When we’ve got the game kickstarted, greenlight, sponsored, published whatever we can always hire an animation team :slight_smile:

Not to mention with a basic set like that, you can pull it into something like MotionBuilder and retarget it to new skeletons. That’s a lot of versatility.

For custom animations, crappy placeholder stuff works just fine. For your core motion and stuff, you want something that at least doesn’t make you sick to look at, and that you can show off to someone without them thinking your game is total garbage.

Assuming you can even get those locomotion animations set up, which many people get blocked on completely, especially one man programmer/scripter/level designers. Given that as a single developer you really need at least something basic that looks decent, with decent motion, to even recruit people to a project in a realistic fashion, I don’t think it’s a big deal to have generic anims.

Also you can always add some nice customization, or replace them later. That’s the point really. To just get on with making your game without getting stuck at the most basic things like animation.

First walk cycle is done:

here:

(just kidding :smiley: )

Haha that’s awesome!

I made a tutorial for the basics of using MakeHuman and Blender to create a character:

Please feel free to comment, edit and enhance!

It is a relief to know I’m not the only one who has big problems with this.
Well, one of the large walls that need to cross, and I think one of the more complicated for me because I am a programmer.
Every time I have to deal with animations I see more and more away my chances of finishing the project I am developing.

I also have this package when i was working with unity.
But I really can not adapt them to make them work in UE4.
If you managed to make it work, please, Could you point me the right direction to do this?

Check this out: http://forum.unity3d.com/threads/released-open-source-unitypackage-extractor-from-cmd-line.180080/

It will allow you to extract the raw FBX files from that pack.

I believe the Rigify export option overrides whatever rig preset you select, so setting it to HumanIK may be unnecessary. What you end up with when you export a rigified rig from MakeHuman is something with 460 odd bones, even if you then export only the deformation bones from Blender to FBX you’ll end up with messy bone names and probably more bones than required (and I don’t recall if the bones are all properly parented and connected for rag-dolling and/or full-body IK purposes). The approach I’ve taken (though it’s early days yet so I can’t say it’s the right one) is to get MakeHuman to export a custom rig (similar to the standard Unity Mecanim rig) and a Rigify rig. I then run a script to generate constraints between the custom rig and the Rigify rig (while fusing/ignoring twist bones) so that the custom rig follows the motion of the Rigify rig. I then retarget BVH motions onto it using the MakeWalk addon, and/or animate the Rigify rig using the IK/FK controls. Once I’m happy with the animation I bake it using the custom rig and export the result.

That’s the theory anyway, I haven’t gone through the entire process yet, currently I need to implement a script to crop the BVH motions so it’s easier to extract loops (cropping is already implemented in BVH Hacker, but I’d like to be able to do it all directly in Blender). Also, note that if you use MakeWalk to retarget BVH motions it will automatically downsample to Blender’s current frame rate, so you don’t have to do that in BVH Hacker.

I know there are too many bones, but I’m not an animator and I have no f***ing clue what I’m doing, or should do, actually :slight_smile:
Would you care to help me and others understand the process better? You seem to have in-depth knowledge about rigging and animating… :slight_smile:

It’s really nice to see everyone else posting in this thread… it’s clearly a much wider problem and lots of us programmers are struggling with the animation pipeline. Making models is something you can pick up fast with lots of good tutorials on youtube etc. but animating seems much more of a dark art.

I’ve been careful about using stuff from Unity because I don’t want to get into legal tangles. Do you think the license allows Unity stuff to UE4? It does say you can use the assets in games, and I can’t find anything where it says the games must be made in Unity, but there are lots of other rules on copying and things like that so I wonder if they might try to claim we’re not meant to do that. Would be interesting to know if this is an OK thing to do.

Hey, I wonder… would it be possible to make MakeHuman export models with the same rig as the blue manikin in UE4 and right away be able to use all the motions and controllers from the samples? That might just solve our problem!

Looks like the UE4 blue dude has some extra twist bones (that have no weights it seems, so they don’t deform the mesh… don’t know what they’re used for in-game), and some IK bones (though again I don’t know if they’re actually used for IK in-game). MakeHuman doesn’t create those extra bones out of the box, so you either have to create them manually, automate it via a Blender script, or just ignore them. Then if you can import the original FBX animations into Blender you may be able to retarget them onto the custom MakeHuman rig, though I’m unsure exactly how you’d go about it. I’ve also run into a problem with importing the original blue dude FBX animations into Blender, I’m using Blender 2.70a and when I try to import the Idle.FBX for example it doesn’t seem to do anything (no error or message of any kind anywhere I can see). Maybe Blender 2.71 can manage it, but I’m on vacation with a crappy connection so that may have to wait a few days. In short, I don’t think there’s currently a straightforward way to do what you’re talking about.

I have been wondering this as well. Also I have been wondering if Unity has the whole animation pipeline solved a lot better? Or is it just all the premade stuff in the asset store makes it seem that way?? I’m a noob with animation so I honestly cant tell but they do have a lot of stuff in their videos that make it seem like its possibly easier in unity to share and adjust animations in engine.

Developers who sell their creations in the unity store are not obligated to sell the asset only in the asset store.
In other words is non-exclusive.
basically as it is said in this thread.
http://forum.unity3d.com/threads/selling-on-the-unity-asset-store-earn-70-each-sale-non-exclusive.78039/

You can simply send a mail to the developer and ask him or ask for his permission to use the asset out of unity.

n00854180t thanks for the info!
I can import the animation inside UE4, But the problem is associated with the way Unity handles his humanoid rig in the workflow (well, i think)

I’m neither a rigger nor an animator, but I’ve previously experimented with various approaches to getting a Rigify rig into Unity3d, and wasn’t really happy with any of them. According to the author of the Rigify addon it was never meant for generating game-ready rigs, and the best workflow he found for using Rigify with games was to duplicate the deformation bones into a separate armature via a script and bake the animation to that armature. So, my approach is based on that idea, except that instead of duplicating bones I use MakeHuman to export two armatures, one that is game-ready and one that is Rigified, then bind them together in Blender via script. You can find the Blender script (can be installed like any addon via User Preferences in Blender), game-ready rig preset for MakeHuman, and instructions in my repo.

Dude your hilarious, I have no Team when it comes to animation as of about 2 years ago I knew nothing about it at all. I am handling all Skeleton Creation, Motion Capture, Retargeting, Mesh Weight Painting & Animation for our Project. I spent 1k out of my own pocket for our animation studio. I still just don’t think you all are seeing the full picture because you do not understand how to do it all yourself yet.

That’s great you crank out 100’s of animation and attempt to customize them to each as best you can. Once someone gets to the whole I need this animation stage the hurdles they are going to have to overcome to be able to create what they need plus still use what your providing is enormous. The rig alone is very problematic, just trying to get everyone to agree on a Rig is going to be hard. That’s like a huge future planning sticking point is the skeleton that will be used. I can not stress this enough.

Look I not trying to sit here and say its a bad you got a great goal, its just not practical or focused on the right direction. If some AAA studio hasn’t come out with some sort of all encompassing get you started guide for this I doubt the indie community could do it justice. Plus others have tried and failed look at Mixamo.

Just handing someone a fish will not stop the problem, teaching them to fish is a better approach. But this is the one thing I think the community or AAA studio’s could do. Make better tools to streamline, uniform and simplify the above discusses process’s. Do I believe they could really improve the tools and not make it so difficult to learn or time consuming = DEFINATLY! Why do you think UE4 has that Maya Skeleton plugin, they are trying. Problem here is its not a universal tool its just for maya users…

But hey what do I know prove me wrong, if you do it will alter the developer scene for the better.

Cheers

Some really awesome news on that front, I’ve just seen that Epic have donated 10k euros to Blender to help improve the FBX pipeline! https://twitter.com/tonroosendaal/status/485431565447868416
Hopefully this should mean that we’ll soon have an easier ride getting things between Blender and UE4. Also shows Epic are committed to helping the community they’re building which is really awesome :slight_smile:

What you’re saying is true, but I think it’s you that’s not seeing the full picture here. The point is that when you’re starting out making a game you don’t necessarily need every single animation working right away, but you do need something. Once the prototype is working and you can see it plays well enough you can then use it to attract talent and funding etc. and then you can swap out the animations with better ones.

Of course there’s no substitute for learning how to do things yourself, but by your logic things like Blueprints and Playmaker are a waste of time because once you need something more complex you’re going to need to know how to program, so it’s a waste of time trying to make a game without knowing how to program. That’s clearly not the case - as long as you know and respect the limitations of the tools you’re using there’s plenty of scope to create awesome games without doing everything by hand.
Many people just want to make a simple game, and would be quite happy to use, say, a basic FPS motion set and work within its limits. If you don’t have the motion for “bend over and pick up” just code a nice particle dissolve when you click the object etc.

Bump.
This thread is awesome i have started reading it on September but the last post was in July. What happened ? Have you found a solution to all of that? This thread should not stop.