User's Manual
The command line User's Manual is available here

If you have any problems with DeepSkyStacker, please report them to the DeepSkyStacker mailing list hosted at groups.io:
https://groups.io/g/DeepSkyStacker

Introduction
Quick Start
Registering and Stacking Tab
Batch Stacking
Processing Tab
Create a Star Mask
The Picture List
Picture Preview (Stars and Comet editing)
RAW and FITS DDP Settings

Introduction
The purpose of this short manual is to explain the features and principles of DeepSkyStacker.

The user interface of DeepSkyStacker is quite simple and intuitive and all the main features and commands are readily available. The user interface is divided into two areas: To the left a set of tabs gives access to all the commands and features. To the right a screen depending of the current tab that is used to configure and view. To change the current tab you can click anywhere in the tab.   If you are a beginner in astrophotography and you are wondering what light, dark, flat and bias frames are, and how to create them there is a short explanation here.

Quick Start
You are just out of a busy night and you just have downloaded all your pictures (lights, darks, offsets and flats) to your PC.

The first step is to add all the pictures (accordingly to their type) to the list.
To do this you use the commands of the left area:
- add picture files
- add dark files
- add flat files
- add offset/bias files
- add dark flat files

Then you just launch the registering process, you check the "Stack after registering" box after entering the percentage of pictures that you wish to keep for the stacking process after the registering process.

If you are using raw files (which is highly recommended), you just check the Raw DDP settings and you start the registering process.
You can now go to sleep knowing that after a good night you will view the first results in the processing tab.
Note that even if your PC crashes after the stacking process, the result is automatically saved in a special file: AutoSave.tif.

Registering and Stacking Tab
The Registering and Stacking Tab contains all the commands and features pertaining to the registering and stacking processes.

The "Compute offsets" feature may be used to compute the offsets and rotation angles between the pictures without starting the stacking process. Once computed the dX, dY and angle columns of the pictures list are populated with the computed offsets and rotation angles.
Note: it is not necessary to compute the offsets manually prior to the stacking process since the offsets and rotation angles are automatically computed just before the stacking process anyway.

The "Check above a threshold..." may be used to check all the pictures for which the score is greater than or equal to the given threshold value.

Before the stacking process a screen recapitulates the different steps. It is important to know that the light frames will be automatically associated with their dark frames using ISO speed and exposure time. Flat frames and Offset/Bias frames will be automatically associated regarding only the ISO speed. In each case the best matching frames will be proposed and a warning will be issued if ISO speed and/or exposure time are not matching.   It is possible to temporarily "force" the ISO speed and exposure time by using the context menu of the files. From this dialog you can also change the stacking methods for each type of picture by clicking directly on the link or on the "Stacking Parameters" button.
A new dialog is then opened, from which you can modify the folder into which are created the temporary files needed by the stacking process and all the settings controlling the stacking process.   The first tab is used to set the stacking mode among which the possibility to use a Custom Rectangle. You should note that in all modes the offsets are computed from the reference frame, so it's not a true mosaic mode but a pseudo mosaic. This is also where you can choose the Drizzle option when stacking.   The drizzle option creates image 4 or 9 times bigger. For already large DSLR images it can create very large images and DeepSkyStacker  may need a lot of memory and disk space to process them.
The other light, dark, flat and bias tabs are used to modify the stacking method for each type of frame, and to activate the background calibration option for the light frames or the automatic detection and removal of hot pixels and bad columns detection and removal for the dark frames. If all your light frames are FITS files, and they have valid World Coordinate System (WCS) data in the FITS file headers, then the checkbox "Use FITS WCS Data" controls whether to use that data to determine the image offsets and rotation with respect to the reference frame. If it is checked, then the image offsets and rotation with respect to the reference frame will be derived from the WCS data which is very fast, whereas if it is not checked then the offsets and rotation will be calculated by comparing the location of the stars in the reference frame and the current image which is much slower.
The Alignment tab may be used to change the alignment method used to align light frames. You can choose Bilinear, Bisquared or Bicubic transformation or you can DeepSkyStacker select the best transformation based on the number of available stars.
This tab is used to control the creation of intermediate files. Two options may be used to create files containing intermediate images: The first option may be use to create a file containing the calibrated image for each light frame. The second option may be used to create a file containing the calibrated and registered image for each light frame. The created files have the name and location of the matching light frame with a .cal.tif or .cal.fts extension for the calibrated images and .reg.tif or .reg.fts extension for the registered and calibrated images. The files are always uncompressed TIFF or FITS files, monochrome or color, 8, 16 or 32 bit depending of the light frame bit depth.   When you check one of these options be sure to have enough free hard disk space to host all the created files.
This tab is to control the way the calibrated images are cleaned before stacking. The goal is to detect and clean remaining hot or cold pixels after calibration. Checking these options may seriously alter the resulting image so use them with caution. Two parameters are needed to define the behavior of the cosmetic cleaning. The first parameter is the size of the filter used to detect and fix the hot/cold pixels. The larger the filter, the longer the processing time and the smoother the resulting image. The second parameter is the threshold under which no correction is done (the pixel value remain the same). A lower threshold means that more correction are done. The last option may be checked to create an image showing the cleaned pixels for each light frame. In this image, cleaned hot pixels are white, cleaned cold pixels are dark gray and all the other pixels are light gray. The created files have the name and location of the matching light frame with a .cosmetic.tif or .cosmetic.fts extension as defined in the previous tab.
The Comet tab is available only when at least two light frames (including the reference light frame) have a comet registered. In this case the Comet tab is used to select the comet stacking mode between the 3 available modes. See the Comet Stacking topic in the Technical Details for information on how to use Comet Stacking.

After the stacking process, the resulting image is saved in the AutoSave.tif file which is created in the folder of the reference light frame..
If an Autosave.tif file already exists in the folder, an Autosave001.tif (002, 003, ...) is created.

Batch Stacking
The Batch Stacking dialog is used to create resulting images from more than one stack.

Processing Tab
The processing tab contains the commands and features pertaining to post processing.

I would like to emphasise that DeepSkyStacker only allows simple stretch and colour balance adjustments  of the image to allow you to visualise the result of the stacking process.  However, if there's no background gradient, it may be possible to achieve a quite acceptable result using just the tools provided here. Any complex post processing should take place in other software specifically designed for that task. The processing tab contains a preview of the RGB histogram, a tab that allows a Mid-tone Transfer Function (MTF) image stretch to be applied, a tab that allows an Inverse Hyperbolic Size (ASinH) image stretch to be applied, and another tab that allows the Colour Balance to be adjusted. If you want to adjust the colour balance, you should first adjust the stretch parameters and Apply the stretch, then switch to the Colour Balance tab, make your adjustments and Apply them.

 

The Mid-tone Transfer Function Transformation or Stretch The MTF Transformation (or Stretch) is one of the most powerful image stretching tools available. The tool is presented in the form of three sliders for each primary colour which can be moved to transform the image.   If the "link" button (to the left of the Autostretch button) is set to "linked" then the sliders for the three colours are linked together, whereas if it is set to "un-linked", then the sliders for each colour are independent. The left hand slider changes applies to the shadow signal, the middle one is the midtone balance adjustment and the right hand slider adjusts the highlights.   The values of these sliders are displayed below the sliders on the left side of the tab as spin-boxes that are linked to the slider settings.   You can directly edit the values in the spin-boxes, or use the the Page-Up and Page-Down keys for coarse changes, or the cursor up and down keys for finer changes. If the sliders are linked, then the text to the left of the spin boxes is displayed in the normal "text" colour.   If they are not linked, then the colour of the text matches the primary colour for which adjustments are currently active. To the right of the spin-boxes are displayed the percentage of pixels that are clipped by the transformation, as ideally you want to minimise the amount of clipping created by the transform. The shadow (left hand) slider sets the black clipping point which allows you to darken the image background by cutting off the extreme left tail of the histogram, thus removing sensor noise and establishing a true black point.   Dragging this slider slightly to the right or using the Shadows spinbox to increase the setting will darken the image background.   Don't overdo this as otherwise you could lose faint nebulosity or shadow detail. The midtone balance slider controls the MTF function which maps input pixel values x (between 0 and 1) to output values y. It is formally expressed as:        (m - 1)x  y = -------------      (2m - 1)x - m  where m is the midtone balance slider value. m = 0.5: Leaves the midtones of the image unchanged (identity function). m < 0.5: Brightens the midtones (pushes the curve to the left). m > 0.5: Darkens the midtones (pushes the curve to the right). The right hand slider or Highlights spin box sets the white point. In addition to the manual adjustments described above there is an "Autostretch" button.   It behaves slightly differently depending on the "link" button (just to the left of the Autostretch button).   When "linked" is true, the image statistics used to calculate the stretch parameters to apply are computed as an average across all three channels, which results in the existing colour balance being preserved. When "linked" is false, then the statistics are computed on a per channel basis which means that the colour channels are stretched independently which results in an implicit background neutralisation that tends to remove colour casts. Finally it is possible to adjust the parameters used for calculating the autostretch by right clicking the Autostretch button.  These control the level of non-linearity applied by the auto stretch process. The Target Background parameter defines the median brightness of the background sky after the stretch.   The theroetical range is [0.0, 1.0] with a default of 0.125.   You'd probably set a value in the range of .125 to .25 for an acceptable result. The Shadow Clipping parameter controls how far below the median the black point should be set.  The value is in units of Median Absolute Deviation (MAD) which is a robust statistical measure of the variability of the data.   Unlike the standard deviation, which is highly sensitive to extreme values, the MAD measures the typical distance from the median and is completely resistant to outliers. If the Preview check box is ticked (the default setting), then adjusting the parameters (by editing the spin box fields directly, or by selecting the spin box or slider and using the Page-Up and Page-Down keys for coarse changes, or the cursor up and down keys for finer changes), then the displayed image and histogram will change to reflect the parameter changes.   However, to actually apply the changes you need to press the Apply button.

The ASinH Transformation or Stretch The ASinH Transformation amplifies the darker areas of the image.   The parameters are the stretch factor (or beta), the black point value, and whether or not to use "Human-weighted luminance". For monochrome images, pixel values are modified using the following function:    pixel =  (original - blackpoint) × asinh(original × stretch)             --------------------------------------------------                        original × asinh(stretch)    For color images, the function becomes:      pixel = (original - blackpoint) × asinh(rgb_original × stretch)            -------------------------------------------------------                    rgb_original × asinh(stretch) The stretch factor (otherwise known as beta) can take values between 0.0 and 1000.0.   Larger values result in a greater degree of image stretching being applied. The "black point" value has a range of [0.0000, 0.20000].   The value of 0.2 implies that the blackpoint is set to 20% of the maximum possible ADU value.    Values below the black point are treated as being black. As rgb_original is an average of the three colour channels, one or two channel values will be greater than rgb_original and could therefore clip.  To mitigate this problem the RGB blend algorithm (which was developed by the authors of the GHSAstro tool to adjust the "gain" applied to the channels) is used to ensure that such clipping does not occur. When the Human-weighted luminance option is not ticked, rgb_original is the mean of the three pixel values; when it is set, the weighting changes to 0.2126 for the red value, 0.7152 for the green value and 0.0722 for the blue value, which gets results closer to human perceptual color balance. If the Preview check box is ticked (the default setting), then adjusting the parameters (by editing the spin box fields directly, or by selecting the spin box or slider and using the Page-Up and Page-Down keys for coarse changes, or the cursor up and down keys for finer changes), then the displayed image and histogram will change to reflect the parameter changes.   However, to actually apply the changes you need to press the Apply button. You should note that the stretch is "multiplicative", so if  you Apply a stretch of 200.0, and then Apply a stretch of 10.0 the effective image stretch is 2000.0.

Colour Balance Adjustment The Colour Balance tab allows adjustment of the image colour balance.   You should make this adjustment after Applying any image stretch. If the Preview check box is ticked (the default setting), then adjusting the parameters (by selecting the appropriate slider and using the Page-Up and Page-Down keys for coarse changes, or the cursor keys for finer changes), then the displayed image and histogram will change to reflect the Colour Balance changes.   However, as with the image stretch, to actually apply the changes you need to press the Apply button.

Saturation Adjustment The Saturation tab allows adjustment of the image colour saturation and vibrance. Typically you would want to use that after Applying an MTF Stretch, as unlike the ASinH Stretch, that transformation tends to desaturate colour images. The Saturation shift slider controls the overall colour saturation allowing you to either decrease it slightly or to increase it.   The Vibrance slider control allows you to selectively boost the intensity of muted, less saturated colours in the image. If the Preview check box is ticked (the default setting), then adjusting the parameters (by selecting the appropriate slider and using the Page-Up and Page-Down keys for coarse changes, or the cursor keys for finer changes), then the displayed image and histogram will change to reflect the changes you have made.   However, as with the image stretch, to actually apply the changes you need to press the Apply button.

Undo/Redo adjustment parameters

On the bottom of each tab a set of buttons is available:
- to apply the last modifications to the picture
- to undo the last applied modifications
- to redo the last undone modifications
- to reset all the values

Loading a picture

DeepSkyStacker can load  TIFF images (8, 16 and 32 bit; monochrome or RGB) or FITS images (8, 16, 32 and 64 bit; monochrome or RGB).
At the end of the stacking process the image created by the stacking process is loaded automatically .


Saving the resulting picture

DeepSkyStacker can save the resulting image as 16 or 32 bit TIFF or FITS  files using two methods: Applying the adjustments:  This saves the image with any changes you have made using the "Apply" button.   You should use this if you plan to do further adjustments in other software. Not applying the adjustments:   This saves the image without any changes that you have made using the "Apply" button.   You should use this if you plan to do all post processing in other software.

For more information about the saving methods you can read the faq.

---

Create a Star Mask
Using a Star Mask is often very useful during post-processing to use different processing techniques for the stars and for the rest of the image.
The Create a Star Mask feature may be used to create a variety of star masks that can be used later in any software you are using to post process (Photoshop, PixInsight, ...).

DeepSkyStacker will only creating star mask files. It does not use them in any way.
You can view this feature as a tool intended to provide quickly and simply a star mask file based on real star detection.

What is a star mask? A star mask is a synthetic grayscale image on which each bright enough star is replaced by a round white area which is fading to black following a curve. You can see an example of a star mask on the right. Mouse over the text to see the image or the star mask.

Star Field

Star Mask Creation To create a Star Mask you must load an image in the processing tab.   Then just click on the Create Star Mask... menu on the right and the dialog on the right will open. From there you can change the shape of the stars in the mask, the star detection threshold and a few other parameters. Once you are done, just click on the OK button and you will be asked to select a file (either a TIFF or a FITS file) in which the star mask will be saved. The saved file is always a grayscale 16 bit image. It is not automatically loaded by DeepSkyStacker. The star mask is always created from the image as you see it in DeepSkyStacker which means with the current adjustments applied.

---

The Picture List
The picture list, common to Registering and Stacking processes contains all the pictures used by the registering and stacking processes.
- Light frames
- Dark frames
- Flat frames
- bias/offsets frames

Only the checked pictures are used by the registering and stacking processes.

The list contains the following columns:

Checkbox	checked if the picture should be used
Path	Path to the picture file
File	Name of the picture file
Type	Picture type: Light Dark Flat Dark Flat Bias
Score	Picture score computed by the registering process or N/A for dark, flat and bias/offsets frames. When you have selected a frame as the reference frame using the context menu, the score is preceded by (*).
dX	X offset (in pixels) computed for the stacking process
dY	Y offset (in pixels) computed for the stacking process
Angle	Rotation angle (in °) computed for the stacking process
Date/Time	Date and Time of the file
Size	Size of the picture in pixel (Width x Height)
CFA	Yes if the picture is resulting from a Color Filter Array (CFA) which is the case of all the RAW pictures using a Bayer matrix.
Depth	Type (gray levels or RGB) and number of bits per channel of the picture.
Infos	Extra information about the picture (RAW with DSLR information, TIFF or Other)
ISO Speed	ISO Speed of the picture
Exposure	Exposure time of the picture
FWHM	FWHM: Full Width Half Maximum value. (in pixel) This is the average for all the detected stars
#Stars	Number of detected stars. When a comet position is set a +(C) is appended to the number of stars
Sky Background	Sky background value for registered image in %

You can sort the list on any column by clicking on the column header (and by clicking again to reverse the sort order).

You can change the columns order and each column width. Everything is saved after each use of DeepSkyStacker. The column widths order will be reloaded the next time DeepSkyStacker is used.

You can preview a picture from the list by selecting it.
If the selected picture is a RAW file, the RAW DDP settings will be used to decode the RAW file before previewing it.

You can select multiple files in the list. A context menu is available (mouse right button) to manually select the reference frame of the stack, change the file type, check or uncheck the files, remove the files from the picture list or view, edit the properties of the files copy the list content to the clipboard or definitively erase files from the disk.   If you select a reference frame that is not checked, the offsets will be computed from this reference frame even if the reference frame will not be stacked.

The properties dialog may be used to change the picture type, and temporarily (only for the current session) the ISO speed and exposure time of one or more pictures. When changing the ISO speed and exposure time, no change is made to the real files. The values are only stored in memory for the duration of the current session.

---

Picture Preview
The picture preview is used to quickly view a picture from the picture list (Registering and Stacking Tabs) or to view the picture being processed (Processing Tab).

You can zoom in and zoom out by pointing the mouse and using the wheel. This is the only way to pan and to zoom.

Star and Comet editing

When a light frame is selected in the registering and stacking tab it is possible: To edit the star positions: Stars detected by DeepSkyStacker will be coloured green.   You can select extra stars or deselect the ones DeepSkyStacker detected by left clicking with the mouse. To edit the comet position To create and edit (move/resize) a custom rectangle You can select one of the above modes using the toolbar in the bottom right corner of the image. The last item of the toolbar is used to save the changes. Comet editing note: The comet center may not be automatically detected by DeepSkyStacker. To force the comet to a given  position hold down the Shift key while positioning the comet center.

When used from the processing tab, the picture preview may be used to select a part of the image by drawing a rectangle directly on the picture. This rectangle is used on one hand to apply the adjustments only to this rectangle and on the other hand, optionally, to save only a part of the image.   If no rectangle is defined the adjustments are applied to the full image. In all the cases, when the image is saved to disk the adjustments are applied to the full image.

---

DDP (Digital Development Process) settings
The DDP settings dialog gives access to all the parameters used while developing a image file to transform it into a picture.

Image Adjustment (or Colour Adjustment) The parameters are: The overall image brightness which applies to the green channel The brightness of the Red and Blue channels relative to the Green channel. Image Adjustment (or Color Adjustment) is applied after the image has been read and scaled appropriately for the type of image (Raw, FITS or TIFF), and after all calibration frames have been applied.    If you use these adjustments, as you increase the values beyond the default value of 1.0, more and more pixels will become saturated. The brightness (or luminance) adjustment applies equally to all three channels.  The values for the Red and Blue scale factors are relative to the Green channel whose scale factor is the brightness value. If you wish to scale 12 bit TIFF or FITS files to 16 bits, then this can be done by setting a brightness value of 16. This can also be useful if you want to set the channel balance for FITS or CFA format TIFF files based on the documentation for your camera.

RAW Files The parameters are: White balance Interpolation mode By default DeepSkyStacker will use Daylight White Balance, unless either "No White Balance Processing" or "Use Camera White Balance" is selected.   If you select "No White Balance Processing", then no white balance scaling is performed at all, which is desirable if you are using software such as StarTools to post process your stacked images. In most other cases we recommend that you use Daylight White Balance, as this option minimizes the effort required to achieve a correct color balance during post-processing.   Original manufacture Astrophotography DSLRs   For specialised cameras such as the Canon EOS 20Da / 60Da / Ra and the Nikon D801A, you should set the camera to use Daylight White Balance and select "Use Camera White Balance" in DeepSkyStacker.   The Daylight White Balance coefficients in the raw file will result in correct colour rendition in this case.   Choosing "Use Daylight White Balance" for an EOS 60Da (for example), will result in a distinct pink colour cast.   The reason for this is that the raw file reports that the camera is an EOS 60D (not a 60Da) whch means that LibRaw uses the white balance coefficients for the EOS 60D which aren't correct for an EOS 60Da.   DSLRs modified for Astrophotography (H-Alpha modification)   You should probably select "Use Daylight White Balance" for this type of camera, even though it will result in a distinct pink colour cast which you will need to remove in post processing.   There's no way that DSS can solve this automatically.   Auto White Balance   The option to choose Use Auto White Balance is no longer supported because it resulted in each frame getting its own unique white balance which messed up dark and flat frame processing.  However, if you really want to use Auto White Balance, set your camera to Auto White Balance and select "Use Camera White Balance".   The white balance adjustment values are extracted from the RAW files by LibRaw Copyright © 2008-2019 LibRaw LLC.  LibRaw is based upon the original DCRaw decoder for RAW files written by Dave Coffin.   More information about the interpolation mode and the specific behavior of the Bayer matrix optimization, is available in the technical details .

 

FITS Files If the check box "Single plane 16 bit FITS files are from ..." is not selected, DeepSkyStacker will attempt to use automatic detection of the  CFA matrix based upon FITS keywords such as BAYERPAT, COLORTYP, and MOSAIC (for Meade DSI colour cameras).   If the FITS file doesn't contain these keywords or the incorrect pattern is specified in the file, you may select the check box and choose the camera/pattern to be used as in earlier releases of DeepSkyStacker. The first step is to select the DSLR or the CCD camera used to take the images from the list. This is necessary to use the correct Bayer Filter Pattern that will be used to decode the image transforming them from B&W to color images. The settings control the values used for: - FITS DATAMIN and DATAMAX values - interpolation mode More information about the interpolation mode and the specific behavior of the Bayer matrix optimization is available in the technical details.