Figure 1 shows the scan (in the opposite direction than previously), looking at the camera at the same time, I have noted the shape of the modes going through the OMC.
Found only two TEM00 modes (expected 3: carrier, lower side-band, upper sideband), that would mean that the upper side-band TEM00 is hiding underneath the carrier TEM10 mode, which would be 10 times brighter.
Hence the small a clean TEM00 is the lower side-band, and the brighter and jagged TEM00 is the carrier (contrary to what I have said in the previous entry).
It actually is not surprising for the carrier TEM00 to be jagged. The loop is on dark fringe, so we just see here the noise of the loop fluctuating around the proper dark fringe. This means that we could try to lock the OMC without a dark fringe offset, as there is substantial offset already (contrast defect?). But of course having some controlled dark fringe offset would make it easier and presumably lower the power fluctuations.
