Ask Unreal Anything: Animation | July 13, 2022 at 11AM EDT

[TomSarkanen]

@BenGTraega

Animation blueprint seems to be incredibly powerful to reduce what is actively being calculated at any given time. There is a sneek peek into how this was done in Fortnite on the supporting document page but I was curious if there were any further plans to do an Unreal Talk on the subject? Common things I’m running into that I would love more detail in practice on:

• How do we smoothly blend between linking and unlinking pose data via piping in previous playing linked data and base level animation graph?
• Can we pass data variables like speed/actor rotation etc from the main graph to the linked graph or is the intended workflow to reset up common casting calls in each graph?

Thanks!

The Lyra sample provides a pretty comprehensive example of how to dynamically link and encapsulate logic into your animation graphs at runtime alongside typical gameplay concerns. There is an upcoming live stream scheduled around the Fall where this is due to be looked at in detail.

As to the two sub questions:

• Smoothly blending between linked graphs is only achieved at the moment using either manual blending (flip-flipping slots or a custom node) or via inertial blending. The blend time for the blend is specified on the graphs properties.
• There are a number of ways to pass variables to sub-graphs. The original implementation (which still functions in UE5) allowed you to expose variables on your linked graphs/layers as pins on the node. These days the recommended workflow is to ‘pull’ data in your linked anim instances via Property Access, either from your ‘main’ anim instance or directly from the rest of the gameplay framework.