Anim BP - Mixing upper & lower body without montages

I have a player with idle animation, and two skills: fireball, and stomp.

The fireball can be used while moving, hence it’s an upper body only animation.
The stomp can’t be used while moving, and it’s a full body animation.

I want to use 2 state machines - one for moving/walking/around (locomotion), and the other for spells (it will include both the fireball and the stomp), but I don’t know how to blend them.

I thought it should be simple - isn’t there any function that gets a list of poses, and override previous bones based on existing keyframes (or not, if it doesn’t exist) ??

Hey there, the idea is to use a layered blend per bone in spine_01 in the animation graph, on the base you set the lower body state machine, on the other node you use the uppet body. You might need to also use Sync groups to keep at least your movement animations synched.

Yes, but if you want upper & lower body behavior without using montages that is the only way i know. There are things in the upper body state machine that won’t exist in the lower body.

I saw that option - but wouldn’t you need to then duplicate many of the nodes in lower and upper (everything that includes both) ?

(a) Why do you need split both if you use montages?
(b) I split my animations into different stages (before casting, casting, and after casting). The player can cancel before casting, casting can be repeated for some skills, and casting and before casting can be interrupted. Can you deal with those cases when using montages?

Is the montage the better way to do it? Montage feels like using animations out of the system, as if not the correct flow.
Also, what do you think about simply creating any animation combination needed and use a single state machine (with many blend spaces probably) ?

i really like to use montages if you want to move while punching, firing or other one off events. Since it’s disconnected from the animation graph i don’t need to worry about transition rules and all of that. It’s a personal preference i would say, i use a split body and use montages for most of the things and it works well for my game.

I had to split them because i bought a movement pack that only had a rifle pose for the upper body, so by splitting i can use the legs from the pack and have other weapon poses. Yes, you just call Stop Montage and he goes back to the animation graph.

What do you think about simply creating any animation combination needed and use a single state machine (with many blend spaces probably) ? Is that a better looking yet more expensive way? Or are there other reasons not do do it in a single state machine ?

If you can do it with just one state machine then that’s great, the simpler the better, it really depends on what you need. In my case it was the best solution, but it might not be for your case.

I don’t have the animations yet so I can go either way; the question is what are the pros and cons of each way?

Splitting into 2 state machines is more versatile but it will require synching and a bit of duplicated states. Just having one state machine is simpler but more limited if you want to do more upper body stuff. If they are just one offs (like punching, kicking, taunt, etc) i would go with one state machine and animmontages. if you require to have idles for different upper body parts (like having different weapons ) then unless you want to make full animations for everything, you’re better served with the split.