Last updated on February 27th, 2022 at 09:06 pm
Overview of my Pixinsight workflow and how images are moved around within my workflow. I have established a folder structure which breaks down to the system into three objects (Camera Object, Telescope Object, Image Object).
This article will not going into the collection and purposes of Bias, Darks, and Flats Starizona has a great article on how these images will isolate the camera sensor noise and optical artifacts of your image from the signal. This article on LightVortexAstronomy has a detailed breakdown of the Pixinsight (PI) workflow to build those calibration images.
The Camera Object (ATIK_383L_Mono) in this example has three master libraries for Bias, Darks, Flats. They are included in the Camera Object since this is a cooled camera. If the is a non cooled camera (e.g. Canon XSi) I would move these files into the Image Object level since you will want to take these under the same ambient conditions your Image Object was captured under. This is also a mono camera so I do not need to Debayer the images. If were using a color camera I would add a folder at the Image Object to collect the Debayer output files.
I have broken my workflow down into 6 Main steps:
- Image Capture – this is where we capture the images of our Object
- First Pass Cull – removal of image sub frames which are bad (e.g. clouds, satellites, planes, blurry).
- Second Pass Cull – using Pixinsight we will pick the top subframes using the FWHM calculations
- Image Calibration – using our Bias or SuperBias, Darks, Flats from our Camera Object we calibrate the images to remove the sensor noise and optical chain artifacts (e.g. dusk mites). During this phase we will also register our images and perform a star detection register the images prior to the next steps where we will stack them.
- Stacking – Take all the hard work above and combine the images into a single image.
- Post Processing – Now once you have a single combined image with all your data you will use the image processing power of Pixinsight to bring out the images
Step 1 (Take your pictures)
My workflow is built around kStars and EKOS. This is an open source program which is built into Astroberry which is an embedded OS running on a Raspberry Pi4. I need to tell EKOS to put stuff into this folder structure correctly.
Step 2 (First pass remove Junk Images)
Make a Junk, approved, debayer, aligned folders
I use Blink process within Pixsight to take a first cut look at the images. This will allow you to pair images down for satellite and aircraft obstructions. Blink will move stuff around for you if you choose so I typically will move stuff the Junk folder for Image captures which are not usuable
Step 3 (Second pass ranking of Images)
Use subframe selector to select the best frames. Use Full Wave Half Maximum “FWHM”
Use the below web site to calculate the Image Scale of the Camera
Step 4 (
Perform the Imagecalibration process within Pixinsight to clean up you frames with the Bias, Darks, and Flat for your camera.
Use best image from the FWHM selection as reference image and select all other images to align. Set Star Detection as defined below
Step 5 (Stacking)
The rejection algorithm is dependent on the number of samples. Use the mouse tips helper for description
Step 6
At this point in time you have a single stacked image that has been corrected for Bias (Sensor Offset), Dark Current (Sensor Dark Current), and Flats (Optical Corrections) that is ready for Image Processing within PI.
- Crop with Dynamic Crop
- Dynamic Background Extraction
- Automatic Background Extraction
- Background Neutral Extraction
- Color Calibration or photometriccolorcalibration
- Create a mask and INvert
- MultiscaleLinearTransform Noise Reduction (See tutorial 8).
- Luminance and Chrominance
- SCNR removes some amount of green or any color
- HDRMultiscaleProcess
- LocalHistogramEquliazation
- CurvesTransformation
- Star Mask and Star bloat
- Saturation
Alternate Steps when using Color Camera Debayer (Add to Step 4)
Select Pure Raw for image format
