Two objects, one image

[cmaier]

After imaging the Ring Nebula there's a galaxy(IC1296) nearby but much dimmer. How would one process the two objects so that the dimmer galaxy shows nicely along with the ring nebula? I assume I would use a mask, but wondered if there was a tutorial or any other help in doing that ...thanks

[zcdawson]

Good question, I will be watching this one. A tutorial would be nice.

[pfile]

i processed this some time ago for jerryyyyy -



if i recall i did exactly what you're talking about - stretched the heck out of it to get the outer shell, then on a clone only stretched enough to see the inner nebula, then used pixelmath with a mask to insert the core of the nebula into the blown-out image. i have to see if i have the project on disk somewhere to get the exact steps because it's been a couple of years.

rob

[cmaier]

"then used pixelmath with a mask to insert the core of the nebula into the blown-out image"  .  That's where you lost me, nice image....thanks for your help...

[msmythers]

Here is the real simple version of what Rob was saying. I have 2 images, one a cup the other butter. I make a mask of the butter using the range mask tool but you can use what ever method of masking you want. I apply the mask to the cup image. I then use pixel mask to add the butter to the cup image through the mask that is applied to the cup image. The result is an image with butter in the cup.


Mike

[cmaier]

I didn't see a "range mask tool" in PI  do you mean range selection tool?

[pfile]

yeah he means RangeSelection

rob

[msmythers]

Thanks Rob.

[pfile]

ok i found the project.

what i did is cloned the image and stretched one a lot, so that the outer shell of m57 was visible. in the other image i stretched so only the core was visible. then i created a mask from the "normal" stretch version by extracting the L*, then used the histogram transformation to clip the blacks and push up the whites in the mask. i also removed the first 3 pixel scales from the mask (with AtrousWavelets, which is depreciated now - you can use MMT instead) to smooth the mask.

then i applied that mask to the super-stretched image, which protects everything except the core. then i applied the pixelmath expression iif($T>0.3,regular_stretched_image,$T) to the masked super-stretch image. what that does is to look at every pixel of the super-stretched image and only replace pixels in the super-stretched image (the "target", or $T) if the pixels were > 0.3. this may not have been strictly necessary since the mask would prevent the dim stuff from being replaced. i came up with 0.3 by using the readout tool to look at the areas that were blown out vs. clean halo and chose a value that represented blown-out pixels.

rob