Having access to neuronal recordings is essential for any network inference algorithm. However, the term ”recording” covers many different schemes which vary significantly in technology, type of output and abundance of data. This report aims to briefly overview some of the methods currently used and compare them from different aspects. In addition, we will take a look at some methods to map the connectome (i.e., finding the ground truth). Finally, we suggest the overall approach that currently seems most appropriate for neural inverse problems and inference algorithms.