Setting DBE points blind?

[iksose7]

Hi guys,

This is my first post here. I've only been using PI for a few weeks, i'm still in my trial period but love what i have seen so far and will be buying the commercial license as soon as i can.

Now, i have already processed a couple images with PI. One turned out successful in my opinion, one did not. But that was only due to me being heavy handed with noise reduction, i learned my lesson there.

On my last image i just used ABE to get started and that was fine. From what i have learned so far, background extraction is the first thing that should be applied to a image? Then colour calibration, deconvolution ect? Please correct me if i am wrong!

But this image has a lot more going on in it so i would like to make use of DBE but when i preform the first auto stretch my image is essentially all red so i have no idea where to place my DBE points to avoid nebulosity. I was hoping someone could enlighten me on what to do here? I'm guessing its not very efficient to apply DBE to a image that has already had ABE applied or is that what needs to be done?

I have attached a screen cap showing the auto stretched image and the image with ABE applied in the hopes it will help.

Thanks
Callum

[Carlos Milovic]

Hi Callum
First of all, try decoupling the channels in the STF window, so the stretched display is calculated for each channel independly. This should remove the color cast.
Another option is to clone the image, modify it (I like to perform a histogram equalization for better inspection for DBE) and create the samples on the clone. Then, create a new icon from this DBE instance, in the background, cancel it and close the image. Focus the original image, and double click the DBE icon, or import the instance to an already opened DBE window (drag and drop on the bottom bar). Now you just have to set the other parameters, and reuse the sample locations.

[sreilly]

Several things first Callum. There are some thing that must be done and many that may or may not be helpful to an image. BTW, your screen shot didn't come through at least on my end. But back to the point, depending on what camera you are using One Shot Color, dSLR, dedicated mono CCD and so on will alter what needs to be done when. For dSLR and one shot color cameras search for a processing sequence that works for that. For mono CCD you will find another work flow and both are suggestive. There are steps you need to take for all. As far as the STF stretch being all red, un-select (click) the chain icon in the upper left of the STF display.

Image calibration needs to be done before anything else.

Cosmetic Correction may be helpful next

Star Alignment is necessary for accurate stacking of images

Deconvolution might be helpful, maybe not

Combining images for master frames is necessary using the proper combination method depending on the number of subs

Cropping the master frame especially if the edges of the combined images overlap, especially if you dither your guiding

Applying DBE if your skies give you background gradients (Search for help on using DBE examples) Look here http://harrysastroshed.com/pixinsight/pixinsight%20video%20html/pixinsighthomenewbie.htmlhttp://harrysastroshed.com/pixinsight/pixinsight%20video%20html/pixinsighthomenewbie.html

Background Neutralization is done usually

Color Calibration of course

SCNR if needed to remove the greenies from image if needed

Noise reduction if needed. Keep in mind that NR is needed less if there is a good deal of data. I prefer to deal with noise near the end of my image processing but I may need to revisit this.

At this point I look to stretch the data using Histogram or Masked Stretch. I prefer Masked Stretch myself.

From that point on it pretty much depends on what you think the image needs. Remember the undo button works well so try a setting and if it looks wonky, undo and try something else. What works on one image may not do so well on the next. A lot depends on the quality of the data as well as the amount of data. And it's easy to overcook an image trying to make it something it isn't. Try different settings and processes and save as you go along. Then open them to compare against each other. Using either ABE or DBE needs to be done according to what is going on in the image. You need to check out Harry's video on DBE and get a good feeling for what and how DBE works. There are other examples out there as well. Get use to using Google to find examples of using processes in Pix Insight. A lot of people have posted on their own websites and blogs on how to do things using PI.

[iksose7]

Steve and Carlos,

Thank you both for taking the time to help me out. De-coupling the STF is exactly what needed to be done!

I dont have the time at the moment to start processing the image properly but when i do i will re-visit Harrys DBE video so i can see exactly how it is done. I'm just hoping my PI trial doesnt run out before i get the chance to process because due to recent astro purchases, the full version wont be affordable for a few months. Not looking forward to going back to PS after using this awesome software!

One question Steve. I use batch processing at the moment, though i will eventually dive into using different stacking methods. I notice you put deconvolution before combining images. Just to be clear, deconvolution is done to the master frame and not before hand? I have been using it after ColCal. Just want to make sure that list is not in order or have i missed something?

Thanks
Callum

[sreilly]

I don't use Deconvolution often but yes, it is done on the master luminance frame. If you use it on the color data I suspect it  needs to be the individual color masters before combining.

-Steve