Skip to content

MORESANE vs Clean on imperfectly calibrated data #21

@cyriltasse

Description

@cyriltasse

I'm experiencing Moresane within SSD in DDFacet. It's working pretty well on simulation, and is super fast, but with real LOFAR data, a few sources are not properly deconvolved. Inspecting in more details, I think this happens when the true PSF is too different from the computed one, the place where we have to solve for a deconvolution problem with the incorect convolution function. So yeah, in terms of inverse problem, crap in crap out. But Clean happen to behave pretty well in those situations and produces artifacts of course. But MORESANE seem to fail and simply stops iterating. This is anoying cos the source may be bright, far from the target, but still contaminating it with its nasty sidelobes. I've played with the different parameters, but could not find any recipe solving the issue.

Is is somthing that anybody has seen before? Or came up with a solution?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions