Pawel Kozlowski

Alright, so I'm not doing things completely wrong! 😝 I took a normalized lineout (horizontally at the 100 pixel index) of the symmetrized image before and after inversion and this...

I tried shifting the background level to zero, and that didn't really help. I also applied a Gaussian filter to smooth things out, and that seems to improve things substantially....

Looks like Abel inversion has an additional distance unit in the denominator compared to the input values. So we have ```[cm**-3 / pixels]``` and we want ```[cm**-3]``` so we do...

@DanHickstein Good to know about the `dr=` parameter! I don't see a `reg` parameter in the docs and I get errors when I try to use it as an input....

I'm on v0.8.0, and yet no reg parameter... how recently was this added?

@stggh as mentioned in a previous comment, I removed the background and it didn't seem to do much... Also that "background" is likely real signal, it's just lower density gas....

> Note, in general the signal onset follows the instrument function and so a sudden edge step is unphysical. So would a Gaussian be too "sharp" then? > Also, I...

> the edge data must be zero (in the real world) Not necessarily. That "edge" is the blast wave and it is expanding into a background gas that has been...

Yea, there is certainly some structure in the center - it's quite apparent in the 2D image. BTW, it's a laser-plasma experiment, not a discharge. ;) Do you guys know...

@stggh will this work on data where instead of having a straight rectangular patch of missing data, it's sort of jagged and cutting into the image by varying lengths? Or...