Vignetting, Flats and Gradient

[JCIdaho]

Hi SGL!

I took this pic of M86 and friends last night. This is just auto-stretched in PixInsight, no background extraction or playing with the levels.

 

Note that there's a light haze emanating from the left hand side leading to a clear overall horizontal gradient, and there's a hazy round donut in there which is harder to see, plus vignetting in the corners.

If I play with the histogram, dynamic background extraction, I can get this end result, although if you look hard for it, you can still find the donut.

 

 

I'm going a bit blind looking at these images now! Here's another one, but without using flats, just to see what the effect the flats have on vignetting:

 

 

Firstly, should I be concerned about this Vignetting? Is it something I can live with by improving on making better flats and processing to mitigate?

For making flats, I'm using a Galaxy Tablet with an app called LightBox, with a white t-shirt over the scope, and the AsiAir to capture. My exposure lengths are 1.4 seconds and I used 20 for the above images.

I suspect though, the flats aren't the issue, and the vignetting is being caused by my setup. I noticed that the vignetting largely appeared when I added the M42M-M42F-2L, that gave me the correct 55mm back-focus as needed, to proper flatten the images.

I'm using:

StellaMira 90mm ED Triplet f/6
StellaMira 2" 0.8x Reducer / Field Flattener
Optolong L-Pro Light Pollution Broadband Filter
ZWO ASI 2600MC-PRO (IMX571)

 

Any thoughts / advice appreciated!

 

 

[rickwayne]

What  do your flats look like when you stretch them?  Can you see any of the same artifacts?

Personally I would not fash myself about it so long as it calibrates out. However you're having to manipulate in post to get rid of it, that's going to present a challenge when you image something with nebulosity filling most of the field.

Moderate vignetting, in particular, bothers me little, since flats can eliminate it. General gradients, eh, APP and PI have good tools for that. Weird circle-things...ok, that's getting uncomfortable.

I often work with linear data using a really hard preview stretch, to exacerbate problems like this so that they're easy to characterize.