PI deconvolution issues

[Astronut1982]

Hi,

I'm trying to use deconvolution for the first time & I have come across a problem that I don't understand.

I'm processing the horsehead & I've been experimenting on a preview to get the settings right before applying to the linear image. The problem is that I have found settings which work fine on the previews & I'm happy that there is no ringing around the stars but when I apply it to the full image, I'm getting dark rings around the stars?

Correct me if I'm wrong but shouldn't using the same settings on the full image as on the preview, produce the same results?

Has anyone else had this & if so, how can I fix this?

Have included pics of the effect on the preview & the full image to show the problem & which settings I'm using.

Thanks in advance.

Jeff

[NickK]

It could be that the previews are running their own autostretch?

[Astronut1982]

I posted on the pixinsight forum & one of the designers told me that deconvolution previews are slightly different & that my preview box was too small. I needed to pick a larger area with more dynamic areas.

Once it had done this, it worked ok.

Cheers,

Jeff

[blinky]

I never really have much success with Deconv can you post a before & after image for me to see?

[NickK]

I posted on the pixinsight forum & one of the designers told me that deconvolution previews are slightly different & that my preview box was too small. I needed to pick a larger area with more dynamic areas.

Once it had done this, it worked ok.

Cheers,

Jeff

I'm assuming that PI works out the dynamic range as part of the LR. If the preview area is too small then the dynamic range will not be correct. The result is that scaling during each LR iteration results causes undershooting (i.e. ringing) or overshooting (not so common).

[Astronut1982]

Yeah, that's along the lines of what Juan told me.

Jeff