I realize this is non-trivial. Would it be possible for the observation to contain the predicted GPS position of the satellite at the mid-point of the observation? The software and database used to track the NORBY LoRa cubesat (see https://t.me/s/FOSSASAT_DATA) reports the position and I have found that to be very useful in plotting the range and directional pattern of my antennas. See my blog posting for sample map.
Using the TLE of a satellite you can calculate the sub-satellite geographic coordinate at all times of interest (e.g. in python using this example). For each SatNOGS observation the associated TLE is stored so one could create a script which calculates the geographic coordinates for the reception times of each frame and create a map with them.
But imo the resulting map is of rather limited use for evaluating the directional pattern of antennas, as satellites are located in three-dimensional space: In contrast to ground-based transmitters, a certain point on the map will correspond to multiple pointings of the antenna.
- the satellite was at a high altitude above the sub-satellite point --> satellite is seen by the antenna at a large elevation; or
- the satellite was at a low altitude above the same sub-satellite point --> satellite is seen by the antenna at a low elevation.
The radiation pattern of a antennas is dependent of both azimuth and elevation, thus the polar plots.
That said: I always love to see more visualizations! So if you are interested in trying to implement this, reach out!