@faraday,
By “impulse,” Day, do you mean for instance a single pulse with steep sides
If so I can recall very unmathematically that in the case of a dsb signal sidebands trail off symmetrically both side of the carrier, maintaining a significant amplitude sfor a distance depending on the steepness of the pulse
Naturally pulse width figures in but I’m not sure just how
Repeated pulsations complicate the picture, with more significant sidebands at a spacing depending on their frequency
The ssb version represents one side or the other but of course without the carrier