This feature is in active development but is not ready to be shipped yet (even in beta). However, if you send us your project file and tell us exactly what you want to do, we might be able to implement a quick fix just for you. Let me know over PM or e-email if you want.
What’s available at the moment
At the moment, I’m afraid that the only solution I can think of is to import a blend map (a texture) as an image and then play with Texture resize, Texture add and other kinds of nodes.
I don’t know how your blend map and your whole setup looks like, but I am assuming you used the warp & blend tools in Canvas setup window. If so, you can export the softedge blend map from Projector’s layout as shown below:
The screenshots are saved to: C:\ProgramData\Lightact\Lightact3\saves
From there you could use the layout below as a guideline on how to adapt your blend mask dynamically:
A short explanation:
Gradient node serves a simulation of your blend mask
Float constructor serves as a simulation of your screen tracking data. You’ll need to transform the screen tracking data into a float that will properly represent how blend mask should adapt based on the location of the screen.
Float clamp node is there just so that we don’t go over 960px, which is the projector’s texture in the sample project I am using.
Vec2 joiner creates a vec2 from 2 floats
Texture Crop or Pad in combination with Texture Blend (multiply mode) effectively stretches the gradient so that the blend mask appears bigger or smaller.
Again, this might not serve your purpose, but please do let me know if it solves your problem or not.