Module |
|
Header |
/Engine/Source/Runtime/Core/Public/Math/TransformNonVectorized.h |
Include |
#include "Math/TransformNonVectorized.h" |
void AccumulateWithAdditiveScale
(
const FTransform & Atom,
float BlendWeight
)
Accumulates another transform with this one, with a blending weight
Let SourceAtom = Atom * BlendWeight Rotation is accumulated multiplicatively (Rotation = SourceAtom.Rotation * Rotation). Translation is accumulated additively (Translation += SourceAtom.Translation) Scale3D is accumulated assuming incoming scale is additive scale (Scale3D *= (1 + SourceAtom.Scale3D))
When we create additive, we create additive scale based on [TargetScale/SourceScale -1] because that way when you apply weight of 0.3, you don't shrink. We only saves the % of grow/shrink when we apply that back to it, we add back the 1, so that it goes back to it. This solves issue where you blend two additives with 0.3, you don't come back to 0.6 scale, but 1 scale at the end because [1 + [1-1]*0.3 + [1-1]*0.3] becomes 1, so you don't shrink by applying additive scale
Note: Rotation will not be normalized! Will have to be done manually.
Parameter |
Description |
---|---|
Atom |
The other transform to accumulate into this one |
BlendWeight |
The weight to multiply Atom by before it is accumulated. |