(January 16, 2016 at 3:19 pm)Alex K Wrote: Your density profile looks very nice, poca Good stuff!
(January 16, 2016 at 3:19 pm)Alex K Wrote: I don't get several things: what is the spectrogram's x axis if you say that for each probe frequency you look at its maximum. Maximum in what? I think I misunderstood something.Ah... right...
What is a spectrogram if not a succession of DFFTs?
For each probing frequency, we get a DFFT, so the X-axis of the spectrogram is the probing frequencies.
The maximum I refer to is that of each such DFFT, which is, in principle, the beat frequency resulting from the beating of the reference wave and the other wave that went in the plasma and came back. This beat frequency has a straightforward relation to the wave's group delay, so I plot it already converted.
(January 16, 2016 at 3:19 pm)Alex K Wrote: A more technical question, In your dispersion formula, I miss a complex variable prescription what to do with the pole of the greens functiony thingy in the integrand. Or does that just give you something finite automatically?Oh... heck... I got mixed up with the names there... damn me and this copying that formula from someone else's paper!!
w_p, in that case, is not the plasma frequency, like I said earlier, but the probing frequency.
The integral, which is calculated as a sum, stops at the last point before the one relative to itself.... actually, it's the d_phi that matters, so it's more like half a point before... We use the algorithm described in section III of this paper, but with a slight correction on the A_i,j, because that thing is numerically unstable.
Since we never reach the pole, we never run into any trouble!