Image resizing after DBE and stacking images with different gradients

[Bobinius]

Hi everybody,

First post here. I have a problem with the background extraction process. After having applied the process, the new generated image is actually a cropped and larger version of the original one. I did  not modify the resolution during the settings and I tried with downsample at 1. Is there a default setting that reduces the size?

Second question: I have two series of subs with different exposures and ISO of M42. One has a gradient that is more pronounced a bit difficult to extract. Should I stack them in linear mode after having applied DBE for each and StarAlignment and then redo DBE on the final image or perform DBE on the stacked image before any treatment?

Thanx

B

[pfile]

well... i'm not aware of DBE ever changing the size of the target image. it could be that the new window opens with a larger size and the image zoomed out a little, but the dimensions of the actual image should be the same. the only control that DBE has that affects the size of anything is the downsampling factor for the extracted background model, but this model is applied automatically to the image as long as you have told DBE to subtract or divide it (which is not the default behavior - by default it just shows you the extracted background.)

as for #2, applying DBE to each sub is kind of painful as it can not be automated. i would try doing DBE on the stacked image before moving on to background neutralization and color calibration.

rob

[Bobinius]

Thanx, you are right. The image is only zoomed, if I reduce it and overlap it on the original it fits.

For the second problem, I not trying to manually do DBE for each sub. I have two sets of subs with different exposures which I stack. I obtain two integrated images #1 and #2, do I perform DBE before combining them or after? Does it make a difference?

[pfile]

ok good. sorry i misunderstood on #2.

since they are different exposures, is your goal to recover the blown-out core of M42? if so you will probably be using HDRComposition to do the merge, and only a very small amount of the shorter-exposure image will participate in the final image. because of that it probably doesn't matter if you do DBE before the HDRC. but on the other hand, it should not hurt to remove the gradients in both images first, so that you're dealing only with the real DSO signal during the merge.

if you have a bunch of stacks you want to integrate together in an 2nd phase, i think DBE before stacking is also a good idea; ImageIntegration is going to compute noise in each input image to come up with a weighting, and if there's a bunch of high-SNR sky signal in your subs the noise calculation is probably going to be too optimistic. in other words, stacks with high LP will probably be heavier higher than they should be.

rob

[Bobinius]

Great, very clear reply!