Animation node API
Posted: 22 Nov 2012, 16:34
In a current project we are using RadarTOUCH with a Touch Translation node, a script and a Timeline Control node to get smooth and continuous gesture control over the transitions from state to state in a keyframe animation.
Since the states in our animation are not equally spaced, we couldn't use Tick Attraction of the Touch Translation node to make motions with momentum stop at each state. Instead, at each state, we set the limits of the Touch Translation node to match the beginning time and end time of the current transition. Animation nodes don't expose the times of the previous/current/next state, so the only way we could achieve this was to actually add begin/end properties to the animation and animate these as a step function. It works well, but by doing so, we duplicate the information already contained in the timeline.
Therefore, feature request #1: Let Animation nodes expose the time of the previous, current and next states, when possible. (I realise that the "next" state is not obvious for a non-linear animation, but still, it would have been tremendously useful for us if we could simply get the times of the current and surrounding states in a linear sense, just like the next/prev events behave.)
Secondly, we wanted to have control by keyboard as backup. Normally, it would be an easy feat to bind the next/prev events of the animation to some keyboard input, but since we already bound the control property of the animation to our timeline control, we can no longer use next/prev events. Instead, we had to emulate the next/prev events in our script, which became rather bulky as a result.
Therefore, feature request #2: Allow a mixture of timeline and event-based control of animations, possibly with an input property to specify which control form should be active.
Since the states in our animation are not equally spaced, we couldn't use Tick Attraction of the Touch Translation node to make motions with momentum stop at each state. Instead, at each state, we set the limits of the Touch Translation node to match the beginning time and end time of the current transition. Animation nodes don't expose the times of the previous/current/next state, so the only way we could achieve this was to actually add begin/end properties to the animation and animate these as a step function. It works well, but by doing so, we duplicate the information already contained in the timeline.
Therefore, feature request #1: Let Animation nodes expose the time of the previous, current and next states, when possible. (I realise that the "next" state is not obvious for a non-linear animation, but still, it would have been tremendously useful for us if we could simply get the times of the current and surrounding states in a linear sense, just like the next/prev events behave.)
Secondly, we wanted to have control by keyboard as backup. Normally, it would be an easy feat to bind the next/prev events of the animation to some keyboard input, but since we already bound the control property of the animation to our timeline control, we can no longer use next/prev events. Instead, we had to emulate the next/prev events in our script, which became rather bulky as a result.
Therefore, feature request #2: Allow a mixture of timeline and event-based control of animations, possibly with an input property to specify which control form should be active.