I don't use the old 8568B very much and it's under a desk in another room but I quickly fired it up and set it to 72MHz and a 10kHz span with 10Hz RBW as per the other plots.
Note how I've increased the RF level to 0dBm but you still can't see any noise. This is because the true divider noise level of the synthesiser in that analyser is below the range of the log amp for the display when 10Hz RBW is selected.
I've had to turn off the graticule because it is very fuzzy when the analyser is still cold. The CRT is getting very old now...
It's in a totally different class to anything on this thread and is a classic lab grade analyser rather than a technicians analyser
Only the latest analysers that still cost a fortune can beat it and even then they have to resort to trickery with the PLL to reduce the noise level.
This 8568B analyser cost me £125 about 6 years ago. A tenth of the cost of the Rigol