The first difficulty I discovered was that our first capture run had the markers spread out among several channels. Whenever the capture system lost track of a marker and then reacquired it that marker would be placed in a new channel. That meant that we had no idea how to track any given marker over the entire capture session. This was solved in two ways. First, we captured another run of data and used the vendor's software to manually reassign markers to the correct channel. This solved most of the problem, but there were inevitably parts of each capture that we missed. Those sections were detected by playing the raw data in Maya and manually entering into my processing program which channels were to be merged.