Square star “centroids”

[Theunissen]

I noticed the following before, I wasn’t entirely sure about it, now I am. When I compare a stack, for example NGC7000 made with APP and the stack made with PI the stars (integration result) in the APP stack have a more square appearance. The attached image shows the APP stack left, and the PI stack right (only auto stretched). Of course, this is pixel peeping, but this effect is very noticeable in my wide field results.

Is it possible that de Adaptive Airy Disc (Debayer Algo) is cause of the above?

https://www.astropixelprocessor.com/community/rfcs-request-for-changes/square-star-centroids/

[Haverkamp]

Marc,

I wouldn’t think so, according to my documentation, AAD will give rounder stars and beter colors then VNG, or AHD debayer. (and have tested this frequently)

Especially VNG debayer (PI) will make square stars, 45 degrees rotated.

And I don’t think that the two images that you show aren’t a fair comparison. You need to show the data normalized for a fair comparison, at least that’s my take.

If you normalize, then you can say somehing about colors, sharpness and roundness. In this example you show two images with very different stretches, so what can or would you deduce from that?

Mabula

[Theunissen]

Mabula, I just noticed it (before).

To make comparison somewhat easier, this is the M31 image (crop) processed by you … and of course, i’m pixel peeping … if I enlarge the image I can see square “centroids”. I’m more interested in the cause (and hence solution) then anything else …

Marc

[Haverkamp]

Okay, but in this case we have no comparison so that’s even harder for me to give a reaction to.

Just send me both linear integrations of the NGC7000 data ?

I’ll normalize both stacks and will then have a good look.

Mabula

[Theunissen]

I’ll send you the links in a few minutes (after slow upload). Maybe we should continue this discussion on your forum?

[Theunissen]

Both are send and 1 PI stack (the original of last year :-) ).

[Author]

Posts