How do you smooth/lerp "Get Input Motion State" outputs over time?

So, on an android smartphone build, I’ve set up Get Input Motion State blueprint node to control a camera’s position and rotation so that the camera points at 0,0,0 world location.

This means that you can tilt your phone and it will move the camera in relation to Get Input Motion State gravity output and it will always points at 0,0,0

However, the movement of the camera is so shakey. Is there a way to smooth/lerp this out over time?

I tried adding a lerp (vector) node to transform the current readout of the Get Input Motion State gravity vector to the camera’s previous frame’s location vector but this only smooths out to the last frame.

Is there a node/setup that I can use to take even more previous frames into account to smooth out the camera location changes even more? I understand this will result in increased lag, but I’m okay with that. However, if anyone have a better idea to smooth out/stabilize the camera changes in location over time, I’d like to know that too.

Please and thank you.

You could try adding a springarm and enabling lag:

214349-lag.png

Another solution is to interpolate manually, by updating the target and letting Tick do the rest:

Thank you very much! That did it!

I went with the interp route as it fits my project better.