How to use DBE in Pixinsight with mono image

[Spookyer]

Hi all, I am struggling a bit with DBE when processing LRGB images. I have the integrated images for each of the four channels. Without any colour in the images how do I effectively use the DBE tool? I cannot see what is pollution in the background and after I run the tool I cannot effectively see how much pollution I have removed.

It was pretty easy to see the red or green yellow gradients with my DSLR colour images but now with mono images I am at a bit of a loss as to how one uses this tool?

Any tips appreciated.

Brett

[White Rabbit]

Check out Harrys tutorials on pixinsight.

http://www.harrysastroshed.com/pixinsight/pixinsight%20video%20html/pixinsighthomenewbie.html

Cheers
Sandy

[RickS]

Do a huge stretch of a copy of the mono image so you can see any faint nebulosity. Place DBE sample points carefully. Make a process icon copy of the DBE instance. Apply to the original image.

Also, look carefully at the gradient that is extracted. It should be nice and smooth. If there is any fine detail then that's part of your image and not the gradient and you need to adjust the DBE. Usually a small number of well chosen sample points gives best results.

Cheers,
Rick.

[gregbradley]

What a great find and link - thanks!

Greg.

[White Rabbit]

Yeah, his tutorials are great. Just in case you don't notice the link is to the beginners tutorials but he has more advanced tutorials on his site.

http://www.harrysastroshed.com/pixinsight/pixinsight%20video%20html/Pixinsighthome.html

I bought and paid for Warren Keller tutorials, and Harrys walks all over them. I found with the ip4ap tutes you couldn't really follow along because there was always a process that would be explained in a future chapter wtf!! And because you can only stream them from the web (although there are ways around that) and no notes to follow a long to you may as well stick to harrys. Not that warrens are bad I just found the a little frustrating at times.

Thanks
Sandt

[Spookyer]

Yes, I have watched Harrys tutorials many times. They are great. Unfortunately with Pixinsight I have found that processes often only apply to a specific type of image. I follow the instructions with my images and sometimes they go okay and other times you get absolute garbage out even with spending significant time tweaking settings. For example his DBE example is with a colour image, you can see the gradients to place the marks because of the colour, on a black and white image I am not sure if I am looking at nebulosity or light pollution when I am marking, it all just looks like cloud.

Colour seems to be my main problem with Pixinsight. I am not even getting close most of the time. I spent 4 hours today on the Swan nebula following various tutorials and have given up in the end, can't get even close to reasonable colours. My lum image looks okay I think but all goes to hell when I try to do anything with the RGB channels.

The colour calibration tool seems to do nothing for me. In Harrys tute he clicks on a nice white galaxy to use as a reference, that is great but what if there is nothing obviously white in the image? How do you calibrate with the tool in the absence of a white reference?

Overall I find Pixinsight probably the most frustrating piece of software I have used. Just when you think you have learned something and found something that worked the next image it doesn't work and you are back to square one.

It not only has a steep learning curve but also a very long one. Normally you pick up speed with software as you continue to use it but I find it seems to be continuous heavy going.

When it works it can work well and is probably the most powerful software around but the wins don't see to come often enough for the time invested. I am wondering if I should be looking at some software that is easier to use and might produce better output for time invested?

Anyway end my sob story.

[Rod771]

Rick is right on the money Brett. Make a clone copy of the image and apply a heavy histogram stretch to it. Then you'll be able to see where to place the samples.

Here is another great tutorial for removing gradients while preserving faint background details using DBE.

DBE tutorial (http://blog.deepskycolors.com/archivo/2010/05/28/removing-gradients-while-preserving-ve.html)

[RickS]

Hi Brett,

In terms of your general complaint, it doesn't matter what software you use to process - every astro image is different and requires different tweaks and techniques. That said, the PI interface and workflow doesn't suit everyone.

I use a variety of techniques for colour calibration. If you don't have a galaxy to use as a white reference you can try using the whole image with structure detection enabled. This will use the combination of all the stars in the image as the white reference. This can give decent results so long as you're not imaging through dust. Another simple technique is just to line up the R, G and B peaks in the histogram (or try an unlinked STF stretch.) If you're really keen another option is to use eXcalibrator (http://bf-astro.com/excalibrator/excalibrator.htm). It gives good results but requires plate solving and a bit of mucking around.

Cheers,
Rick.

[Spookyer]

Thanks for the support guys. My post does sound a bit like a whine now I read it! :sadeyes:

I think I was frustrated with my lack of progress.

Thanks for the tips, I will give them a go.

Brett

[Octane]

Brett,

Keep at PixInsight. You just have to remember that you can't destroy anything. Just go nuts. Go through all the things that look interesting and relevant to your image and try them out. Try different steps and different values. I started doing this after Rogelio's presentation at AAIC in 2011. I got the software and started messing about. Within two days, I'd made my first mosaic. :D

Since then, the software has improved with new tools and functionality. I'm just getting back on the saddle again after a lengthy absense, and, unfinished projects that were started 3 years ago.

Keep at it!

H