First of all, this video must be
very old. You can tell this from the fact alone that it was done by Steve, the general manager of Siglent North America back then. All application videos from 2016 until today have been done by Jason, an application engineer and not the GM himself.
I think this must be an early version of the SDS2000X FW and this is a far cry from today's capabilities and features. The FFT using up the full screen width without any axis labeling is a dead giveaway. Even though window functions have been available back then, we don't know which one has been used for the video. FFT lengt has been a fixed number of 1024. Because of this, a narrow RBW is impossibe to get even at very low sample rates. The low number of FFT points also explains why it could be a little faster.
The numbers on the soft buttons do not mean anything like RBW or similar, but simply the graphical zoom of the display.
Here you can see a review of this early FFT implementation, still on an even older SDS2304 (without X), but this was the same FFT before the redesign. It was also the time when Siglent scope FW used to be rather buggy
https://www.eevblog.com/forum/testgear/siglent-sds2000-new-v2-firmware/msg816867/#msg816867Later on, Siglent tried to implement the same FFT (with markers and marker table) as we have it today even on the bottom of the barrel SDS1000X-E series on the soon obsolete SDS2000X, but it was clear from the start that it could be a 64 kpts FFT at most. In the end it turned out that resources in this scope were only sufficient for a 16 kpts FFT and without the marker functions. This is still vastly better than what we had before the redesign.