Apply amplitude fluctutations at a fixed rate

[HGSilveri]

Now that I see it, the implementation is perhaps a bit weird, since in the case of a Sequence containing only one super-long pulse, the value of noise_base that will be applied will be the same all along - there will be no "fluctuation" actually.
I would suggest to define a rate at which fluctuations should happen (and perhaps they should fluctuate at each ns ?), and sample them at each run (as it's done at the moment).

Originally posted by @a-corni in #710 (comment)