We did many things, but the punch line is we seem to be at about or better than our tolerance.

We pushed the cube further to take out the gross misalignment, and then used an additional 50 gram trim mass.

After this, we put the table on the stops to fix the input aperture position and put the target on it. Jason's measurement showed that it is about 1mm off horizontally, which is very good.

Then we removed the aperture and backed off the stops so the entire thing floats freely and damped the vibration using BOSEM.

Jason was able to focus on the bottom periscope mirror. By panning the total station, he was able to measure the angle of the four edges of the mirror aperture:

Top: 1.319 mrad, Bottom: 1.363 mrad, Left: 1.993 mrad, Right: 1.867 mrad

Since it was absolutely impossible to eyeball the centering on the periscope mirror or on the table hole, I and Virginio calculated independently the centering on the periscope using Jason's numbers.

My script (preliminary) produced (-0.83mm, -0.41mm), which is good. Note that the requirement for the centering at the table hole (which is close enough to the periscope mirror) is 1mm.

Virginio's script (again preliminary) produced (-0.41, -0.14)mm, which is even better.

At this point we haven't reviewed the script of each other's, but it's very unlikely that both of us are grossly underestimating the centering error.

We asked Jason to look further down, but he can only go as far as the first steering mirror after the top periscope mirror, and there the view is already very dark even with Stinger flashlight.

There's no way he can look all the way down to the QPDs.

Also, the visible red laser from the total station becomes really faint on the ISC table and is unusable.

If we want to make another independent measurement to confirm the alignment, we need to use stronger IR laser.

For future, I think we need to make a thin target for the table hole. It should have a small 1mm-ish hole in the center which we can illuminate from the top, and scribed cross and bull's eye pattern which we can illuminate from the bottom.

If we make this out of thin aluminum disk. If we can make this less than 10 grams, judging from our experience of moving 50+ grams on the table, it would not tip the table that much.