The residua in the following plots have been computed as the absolute difference between projected track hit and actual hit using alignment parameters determined by Akio.

The first 6 plots are for disk1, the second set of three for disk2 and the last set of three for disk3.

For each disk I selected the residua vs x [cm] on the detector. The resiuda are also in cm.

With outliers the mean residual is around 1mm, cutting out the outlieres one gets down to around 300-400 micron, a similar value that Akio saw.

Here it looks like the outer two disks have residua which have several local maxima. Maybe the detector got moved? To check I also looked at a single run but one sees the same pattern.

The middle disk2 doesn't seem to have this problem, but the residuals are not centered around zero. So that seemed to hint to an incomplete alignment or a wrong implementation on my side or it is just the effect of the residua of the points used for tracking, come to think about it, I should probably do a refit with the point included...

First two plots are projections of the residua for disk2:

Disk 2

Disk3