Maximum "meaningful" precision of recorded data

Hello everyone,

To first give some insight into what I am doing I am recording the differences in GlobalTime between interactions in spheres in the order of 7um in radius and unsurprisingly these time differences are quite small and I require a large number of sig figs to take these differences.

I am wondering if anyone is able to give me some insight into the maximum precision/number of significant figures with which data can be recorded whilst still giving meaningful results?

Kind regards