Well, some kind of fm signal is captured, and it works like this.
I am preparing a webpage for the detection of all our measured 22GHz masers like W49N. The results over the period may to september 2022 were interesting.
When recording the profile, it was noted that the profiles are not constant; they even were not constant over a short time.
Large peaks appear and disappear in a short time. An explanation is that clouds are orbiting the protostar and when they are inline with the observer, extra masing takes place and the protostar masing signal is amplified.
when it surpasses the inline direction the signal drops again.
By monitoring a long time, an orbital period could be found.
Interesting now is that when a cloud is also rotating on its axis and then is passing the inline direction, the maser peak signal frequency will change up and down (or vv).
When this is monitored accurately, the rotation speed can be determined.
The measurement in the paper gives a deviation of 36kHz, and this gives a surface rotation speed 1700km/h. In comparison; the earth equator's surface rotates at a speed of 1600km/h.
With the SG telescope we do our best to also capture this effect, but it is difficult with a DRO based oscillator.
Solution could be a GPS based LO or perhaps a GPS based pilot tone insertion. Another solution may be comparing 'stationary' peaks to the monitored peak.