In January, in alog 46521, I measured directly the angle that PRM could move in yaw before clipping on the CW1 baffle (on the left in the images of CW1), and the angle to move the REFL beam in the -X direction, toward centing the beam on CW1.  At that time, it was +200urad to clipping on the +X side of the CW1 baffle aperture, and -1000urad to clearly see the REFl beam near the center of CW1.

I've attached some video of the camera that looks at CW1, and the video shows that the intensity of the IR light changes, ie.e flashes that are birghter.  The baffle aperture is 16mm.  CW1 is 25mm.  The input beam is off center, +X direction, approx. 2mm, and this puts the REFL beam approximately 2mm from the edge of the baffle aperture.  There are are also 5-6 beams that come back through the IO Input Faraday, and since the input beam is off center, we can see at least 3 of these that would ideally be block before reaching CW1, but are seen clearly on CW1 and it's baffle.

There are 3 stray beams that are easy to identify:

1. There is a stray beam that is easily seen when at 2W, below and to the left of the input beam, that is clearly at the baffle aperature, so part of the beam gets through the baffle, and part of the beam is blocked.
2. There is a stray beam that is up and to rhe right of the main beam, and is clearly at the baffle aperature, so part of the beam gets through the baffle, and part of the beam is blocked.
3. There is a beam to the right of the main beam, and this beam sits on the thin edge of CW1, part of the beam on the optic, and part of the beam goes past the optic and is blocked by the baffle.

There is a fringe patter that is above and the the left of the main beam, seen on the edge of the baffle.  This is not a typical stray beam, it was not present when the IO Faraday was aligned at air.  One source for this beam could be a double reflection coming from the optics in the Faraday Rotator.  I have also seen this fringe pattern on AOE2, where REFL comes back through PRM, and is headed to HAM1/ISCT1.

If we assume that the current input optics and input beam from the IMC are the same as in January, then it seems unlikely that REFL itself is clipping on the CW1 baffle, however, with only 200urad of rotation before the REFL clips, it seems like it can't be ruled out.

More likely, the beams that are part way on the CW1 baffle aperture are mvoing on and off of the baffle, and creating the flashes seen on the camera.

While the stray beams are not particularly strong, it seems possible that when the beams move over the edge of the baffle aperture, the scatter from the stray beams, especially those on the left side of the camera image, would scatter directly into the main beam.

I've advocated for centering the beam on CW1 and have never been closer, and of course, this scattering may be deemed unimportant, but given the persistant noise we see on POP and REFL, I thought it important to bring this flashing to a wider audiance.