This tutorial is intended to provide an easy start to load, stack and apply deconvolution to the following:
To help, sample data has been provided for down-load. The data provided has been extracted from large SER files (typically 2000 frames). The lunar colour data was produced by cropping an area of a 5440x3645 video to reduce the file size from 1.5Gb to 130Mb and facilitate faster file download times.
Although the sample data has reduced quality compared with the original, the smaller data dramatically reduces the processing time, thus providing an enhanced learning experience.
Download the test data from [BiggSky_Quick_Start_Data] and save the data to a suitable folder. The downloaded ZIP file will need extracting. This is the expanded folder structure.
Data courtesy of David Richards, please give credit in any use of images or derivative work.”
NOTE: This tutorial requires BiggSky v0.4.7 or later. Please go to the Downloads page for the latest version.
BiggSky - Quick Start Guide – Lunar and Planetary Processing
Single Planetary Image (Jupiter)
Single Planetary Image (Saturn)
The downloaded data includes:
** Astrobin Top Pick Nomination | |
** Astrobin Top Pick | |
| |
| |
| |
|
An overview of the steps required to perform Stacking and Deconvolution:
Steps required for single Images:
| Steps required for multiple images:
|
** [Note: If the filename or path contains the word lunar, luna_ , moon, jup_, _jup, jup-, -jup, jupiter, sat_, _sat, sat-, -sat, saturn, mars_, mars-, _mars, -mars, mars, sol_, solar , sun_ (case insensitive) then the Target Type will be selected automatically.]
Prior to processing, confirm the Create Project and Save Options (under the Process tab) in BiggSky are as desired.
Check Create Project to create a settings file so the project can be easily reloaded. |
*** The software is distributed using the settings below. This document has been written using the settings below. It is recommended that, whilst learning the application, these settings are used.
To automatically save the output files in the same location as the source files, make sure Use Image Dir is checked. If Sub-Dir is checked the files will be saved in the BiggSky subfolder. For maximum flexibility with future processing, check 32-bit Data, and the desired file format. For maximum compatibility with other applications and to have the rendered view saved, check 16-bit View. |
These are the steps required when processing a single video file (see next table for additional steps when processing multiple files). First, start BiggSky.
From the Start tab in the upper centre panel click Open Image(s)… and select the downloaded data file (a SER file). | |
The Target Type window may appear. Select Target Type = Lunar or Planetary as appropriate. [Note: If the filename or path contains relevant keywords such as lunar, moon, jupiter or related abbreviations (case insensitive) then the Target Type will be selected automatically.] | |
In the right-hand panel, Click the Do It All button. For Lunar/Planetary targets, Do It All will run both Stacking and Deconvolution without user input. |
Wait for the processing to complete.
Additional steps are required when processing multiple images.
When asked to confirm the internal file names, click OK or change the names to suit. The default Dataset names are determined from the unique parts of the multiple filenames. The common filename prefix and suffix are also extracted. | |
When asked to confirm the Project Name, click OK or change the name to suit. The default Project Name contains the prefix, first and last dataset names, and suffix. |
In the right-hand panel, Click the Do It All button. For Lunar/Planetary targets, Do It All will run both Stacking and Deconvolution without user input. | |
When asked which Datasets to Process, click Select All and OK. |
Wait for the processing to complete.
BiggSky will apply sensible defaults relating to the image(s) and proceed, unattended, to analyse/stack the video file and apply deconvolution to the resulting stack. The result will be stored in a folder called BiggSky, located in the data folder.
The typical deconvolved final image may look too dark, due to the increase in contrast. For Lunar images the default rendering will have a Gamma correction of 1.5 applied. This can be customized on the Render tab. The brightness of the final image will depend on the equipment (telescope/camera) and the settings used in the capture software. There are no definitive rules for defining the appearance of the result, but suitable values can quickly be determined and are partially down to taste (for example – how bright should the final image be?).
The rest of this document describes further processing tasks which are optional and could be applied, as appropriate, to the final image.
What just happened? BiggSky applied sensible defaults to process a single lunar/planetary image.
In the right-hand panel, under the Stack tab:
| In the right-hand panel, under the Decon tab:
|
During the stacking process, BiggSky analyses the frames and produces a Quality graph.
When BiggSky switches from the Stack tab to the Decon tab, a blue square briefly appears. This represents the Region of Interest (ROI) used to perform the deconvolution calculations.
This is the final image as displayed in the application.
The final image might be too dark, but this can easily be amended by adjusting the gamma value
Change Gamma value, in the bottom middle panel on the Render tab:
|
Gamma values such as 2, 1.75 and 1.25 could be tried until a suitable value is found.
With Gamma = 1.5, the image now appears thus.
The final step – the orientation of Plato is not as normally displayed. This can be rectified by rotating the image through 180°.
Rotate image 180°:
|
The image now has the normal orientation.
Finally, save the amended image.
Save the image:
[Note: with the Save options displayed on the left, the final file will be found in the Frames_View_16b folder.] [Note: these are the minimum recommended settings to ensure compatibility with other applications.] |
The final image.
These are the final folder structures and contents (rotated image in the View_16b folder) :
These are the final images as displayed in the application.
When BiggSky switches from the Stack tab to the Decon tab, a blue square briefly appears. This represents the Region of Interest (ROI) used to perform the deconvolution calculations.
These are the final images, as displayed in the application.
Move between the 2 images as follows.
On the left-hand panel, near the top, the 2 datasets are listed. By clicking on either of these, the display will switch between them. Any actions taken will be applied to the displayed image. |
The two datasets can be combined to make a mosaic. Software to do this, Microsoft Image Composite Editor, can be found on the Wayback Machine at Image Composite Editor – Microsoft Research . [possible disclaimer about ‘download at own risk’]
When the software has been downloaded and installed, creating the mosaic requires a few simple steps.
The 2 files required can be found under the folder BiggSky in the Decon_View_16b folders
| The resultant 2-panel mosaic looks like this: |
At this point, the processing can be considered complete.
These are the final folder structures and contents:
This is the final image as displayed in the application.
Next the colour needs to be enhanced. Select the Color tab in the right-hand panel.
Alter the following in Color:
|
Because further processing has been applied, the image requires cropping to remove blurred unprocessed edges (a by-product of the deconvolution process) and saving to a file.
Finally, save the amended image.
Save the image:
[Note: with the Save options displayed on the left, the final file would be found in the Frames_View_16b folder.] [Note: the settings under Save Options are the minimum recommendation to ensure compatibility with other applications.] |
The final image.
These are the final folder structures and contents.
When BiggSky switches from the Stack tab to the Decon tab, a blue square briefly appears. This represents the Region of Interest (ROI) used to perform the deconvolution calculations.
This is the final image.
This is the final image as displayed in the application, processed with Denoise 100/100. The final image might look too smooth, but this can be amended by adjusting the denoise values. |
Adjusting the Denoise values:
Try adjusting the Denoise %:
|
With Denoise % set at 50/50, the image looks too noisy. | Try Denoise % set at 75/75 and click Run Decon. |
A comparison of the settings.
Once satisfied, the image can be cropped and saved. Ensure the Decon view is selected.
Finally, save the amended image.
Save the amended image:
[Note: with the Save options displayed on the left, the final file would be found in the Frames_View_16b folder.] |
The final image.
These are the final folder structures and contents:
This is the final image as displayed in the application (note the moon in the lower right-hand quadrant).
The Stack/Decon process will be repeated (unattended) for each image loaded.
The values for Denoise % and Capture Gamma can be adjusted if required (see the Single Planetary Image section).
These are the final folder structures and contents, with the folder being created within the Projects folder.
The processing steps for Stack + Decon of Saturn follow those for Single Jupiter above. Because of the rings, plus the lack of surface features from which to extract the PSF, Saturn is a special case.
Before performing deconvolution on a body such as Jupiter or Mars, BiggSky identifies a region of high contrast on which to base the deconvolution calculations. This Region Of Interest (ROI) is defined by the blue rectangle. | |
This would be the default ROI for Saturn. However, the rings distort the calculations. | |
To overcome this, the ROI must be adjusted to include all of the body and the rings. |
How to define the deconvolution ROI for Saturn.
Carry out the usual steps:
When the run has completed:
| |
| |
|
This is the final image as displayed in the application.
This is the saved image.
These are the final folder structures and contents with the final image stored in the folder Decon_View_16b.
Page