Blending Non-linear image - Help with planetary Nebula Data

Christopher Foster

Adam,

I have some very nice narrowband data on M27. The OIII data is very strong and the dynamic range of the data is huge. I can see a very nice halo region of the nebula, but to do so blows out the core. This is different that with a target like M42 where the core saturates the image in the linear state. Here there is no saturation of the linear data. I have tried various methods of bringing out the halo, but everytime the core is blown out. I can stretch for the core and see the details, but the image is clearly under stretch for the background, stars and the nebula halo. Or I can stretch for the halo and see it nicely but lose all detail of the core. I want to blend the core (low stretch) with the halo (normal stretch). I have tried the blend script, pixelmath with a core mask, and HDRMST, and even HDRCombination and nothing I try really works. What would you suggest? I can send you the master light frames for Ha, SII and OIII if you would like.

thanks,

Chris FOster

Christopher Foster

I tried to just use a mask to select the core and then use Pixelmath the set the core to 0 for the halo image so I could then add the core back in. However, PI is ignoring the mask and setting the entire image to 0. why is PI ignoreing the mask????

BlockHead

Hi Chris,

At this moment...I have quite a few things on the stove top cooking (as you probably know). I need to finish these items first. I am hoping to create more videos this week.

Concerning Pixel Math-  read this forum entry that answers your question:

https://pixinsight.com/forum/index.php?threads/pixel-math-and-masked-images.1631/

-the Blockhead

BlockHead

On another note... 

If you use the GAME script... you can write zeros directly to the image. 

This is one of the things I ask Harmut to implement in his script.

-the Blockhead

Christopher Foster

I have tried a bunch of methods blending the data and all are very unsatisfactory. I am trying a very different approach. The issue here is very wide dynamic range but no with any sort of saturation. I am stretching now using only ArcSinhstretch to preserve color and then use HDRMultiscale transform on the RGB image to flatten the dynamic range. This is a very tough data set. I may need to punt. I know you are busy, but this is an intriguing problem.

BlockHead

I think what you are fighting against is that when you use any of the tools you have described (ArcSinStretch, HDRMT, MaskedStretch...etc etc) that ALL make faint and bright things more equal to one another. This is a decrease in contrast. This isn't about saturation- it is about the fact more and more values are nearly equal to one another. So the trick is... to put contrast into the highly stretched versions of the image(s). 

See? The contrast of your original OIII data is very high! The faint stuff is there...*but* you have to stretch the thing (making values more EQUAL) in order to see the outer halo. But let me repeat- the original OIII data has the highest contrast! Do you see where I am going? The blend you likely want is with your stretched image WITH the original image. You are trying to make stretched images work with one another..but I think there is another pre-step you can do to help yourself out.

-the Blockhead

Jeffrey Cymmer

I did some color imaging of M27 last year.  I saw Ron Drescher's image where he combined some narrowband and the final product is impressive.  I took some Ha and OIII images recently and, like Chris, have some good detail of the halo around the nebula.  Also, like Chris, I am having difficulty in bringing out the halo without blowing out the nebula.  In one of your recent videos I see you've been working on M27.  Do you plan on doing a video?

Jeff

BlockHead

Hi Jeff,

Indeed, I was thinking about it!!

I am also trying to get my courage up to do a video on Photometric Mosaic... 

Let me work on this...

-the Blockhead