Bad case of Newton Rings ?

[tworrall1]

Hi

Eating into my 45 day licence and being bugged by an image "defect" the image attached is just to allow you to view the defect
noticed these on my fllats. and finished processed images

They appear as light and dark rings on stacked flats and despite all the corrections ie DBE etc end up tainting the final image

Equipment SBIG 8300 , Skywatcher 200PDS Starlight Express filter wheel with 36mm LRGB unmounted filters


Hope help is out there :-(



Trev

[georg.viehoever]

The "defect" (?) is barely visible in the image. Maybe you do the following:
- load an image showing the effect into PI
- start the STF tool
- press the A button to automatically get a stretched screen representation
- create a screenshot and post it here
Georg

[tworrall1]

Hi

Sorry about that , I have stretched the image with Picasa , see what you think


Trev

[georg.viehoever]

To me this looks like a normal flat frame...

[Juan Conejero]

Hi Trev,

This looks like posterization. Posterization is always caused by insufficient numerical precision to represent a smooth gradient, and may be a real problem or just a screen rendition artifact. Smooth gradients such as DBE background models or master flat frames require a 32-bit format to be represented correctly (sometimes even more).

For example, if you store a DBE background model in 16-bit integer format, most likely this will lead to posterization, since 65536 discrete values are not sufficient to describe the extremely smooth gradients as apparent continuous functions.

On the other hand, even if the master flat frame is a 32-bit image and it represents its gradients with enough precision, its screen rendition will be posterized since the screen is an 8-bit device. Of course, the same happens when you convert a 32-bit master flat frame into an 8-bit JPEG or PNG image, as the case of the stretched version you have attached before.

Hope this helps.