I have a object that I want to first simulate, using dynamics, and then animate, using keyframes. But I want the transition from dynamics to keyframe-animation to be smooth.
I think the "blendTwoAttr"-node is what I need, but I cannot figure out the right workflow.
When I create it, it makes 10 blendTwoAttr -nodes, probably one for each item in the channelbox. If thats the case I wonder how to connect dynamics to input 1, and keyframed translation and rotation to input 2.
To me it seems impossible to know which channels each of the "blendTwoAttr"-nodes are operating on.
Do I need several objects, one simulated, one animated, and one that receives both these as input 1 and 2 on a "blendTwoAttr"-node? I can still not find out how to easily make the connections, and be sure they are right.
If anyone could point me in some useful direction I would be very thankful, Best regards, T.
The blendTwoAttr node blends the values of the input(0) and input(1) attributes of the node using a blending function specified by the attributesBlender attribute. The value of the output attribute of this blend node is computed by: output = (1 - attributesBlender) * input(0) + attributesBlender * input(1)
And thats just describing the principle, not the practical usage. I still wonder how to set it up, if it affects one or more channels at the time, if it is practical to use with creatures consisting of many objects. It does not seem so to me. I seems pretty awkward, but anything that can point me in some useful direction is most welcome.
Oks let me explain you have two inputs and on control input whent the controll is 1 input 1 is fully in charge when its 0 input 0 is in charge. when its 0.5 the avarege of both is the output.
So what you do now is key atribute 1 and then key atribute 2 and whan you want to change from 1 to 2 then change the value of the blender from 0 to 2. Now what theese walues controll is up to you realy but there are multiple of uses for this. The same goes to the multiply divide node and +-average. They all alow you to take multiple inputs and grab them into 1 output. You could even use them in conjuction with a switch node.
PS the manual is quite cryptic on practical usage of anything. Its reasonable to asume thet the users know or atleast should know what they are doing.
so what you need to do is this connect all simulated inputs into input 0 and all the keyed inputs to 1 then connect all blenders to a custom channel that i call for convinience simul/keyed when its 0 the simulated is fully in charge whan its 1 the keyed is. And anything in betveen is in between.
But Im not sure though: Do I need one simulated object, one keyed object and one object that is driven, via the blendTwoAttr-node, by these two?
Can I connect all translation and rotation channels from the simulated object to input0 and all translation and rotation channels on the keyed object to input1, all on one blendTwoAttr-node?
Or do I need one node for each channel?
And how do I practically proceed? Using the connection editor or the hypergraph?