The Montage video on youtube sets up a Montage for this sequence:
StartPunch → LeftPunch → RightPunch → StopPunch.
What I want to play is
StartPunch → LeftPunch[i] → RightPunch[j] → StopPunch (repeat) where each of the individual LeftPunch[I] and RightPunch[j] animations are not the same length. I’m trying to understand Montages but I think they aren’t what I want since the timeline in them is static. In other words, the sections start and end at fixed times in the timeline.
Right now I have something close implemented in separate animation states (StartPunch, LeftPunch, RightPunch, StopPunch) but without enough variation. In each state I have the animation sequences for that state instanced into the blueprint and then I use a “Blend pose by int” node to select from them. This functions well but isn’t powerful enough. Suppose I’m making StreetFighter and each character has a different set of punch animations. I want the actual animation sequence to be computed based on what the current character is and what state it is and to select from a set of animations for variation. How do I do that? Does the final result of the calculation need to be an index that feeds into “Blend Pose by Int”? I’d much rather that my logic results in which animation to play.
I am a veteran programmer so this would be easy in C++ if I knew how to indirectly refer to an anim but I’m trying to stick to Blueprint. I’ve written my own blend nodes in C++ and I was familiar with unreal script and Unreal 3 blend trees. I’m still trying to get my head around Montages.