Stacker Unlimited Directives

Job Directives	Values	Description
Name	text	Name of Job (automatically used to name XML file)
Description	text	Job Description
Stack Directives	Stack Parms array	Stack Directives (Job can contain any number of Stack Directives)
Run Directives	Run Parms	Automatic Run controls

Run Parms	Values	Description
Auto Start	yes/no	Automatically run Job when loaded
Minimize	yes/no	Automatically minimize window when loaded (use with Auto Start to run in background)
Auto Close	yes/no	Automatically terminate Stacker Unlimited upon completion of all stacks (be sure to save files)

Stack Parms	Values	Description
Name	text	Name of Stack (Shown in processor)
Description	text	Stack Description
Input Image Type	Monochrome Bayer Color AVI	Type of images to be processed. Only one image type per stack is permitted. Supported File Types · Monochrome: Image is a single array of unfiltered or single-filtered pixels. · Bayer: DSLR raw, One Shot Color raw. Image is a single array consisting of Bayer Matrix pixel groups. · Color: The image is 3 arrays [R,G,B]. · AVI: Uncompressed monochrome AVI video stream. (Color input is rendered as monochrome).
Bayer Directives	Bayer Parms	Input Bayer attributes and directives Required if [Input Image Type] = Bayer
Color Directives	Color Parms	Input Color (RGB) directives Required if [Input Image Type] = Color
Image Files	InFile Parms array	Specify input files and pre-processing directives At least one image file is required
Sub Image	yes/no	Define sub-image (active image area) Some sensors have masked areas (overscan) or bad edges. Lenses may have serious edge aberrations. Pixel not in the Sub Image Area are set to MissingValue.
Sub Image Area	Rectangle	Sub Image area {X=*,Y=*,Width=*,Height=*}
Alignment Directives	Align Parms	Registration/alignment directives (optional) Default = no registration
Selection Directives	Selection Parms	Image Selection Criteria Beta Warning: methods are not fully implemented
Combine Directives	Combine Parms	Combine, Normalize, Data Reject, Save Default = create combined image with no normalization or rejection
Number Of Threads	integer >0	Number of parallel processes Each process handles one frame from input to stacking. Setting this directive to 1 produces linear serial processing. Setting this directive to 2+ results in parallel processing, this can produce a slightly random order of frame processing. Use multiple threads to obtain faster performance. For maximum speed on a dual-core CPU specify 4-6 threads. Each thread consumes added memory, so avoid multiple threads if memory is constrained (e.g. large color images in 32 bit OS). Default = 1

Bayer Parms	Values	Description
Bayer Matrix	BGGR RGGB GBRG GRBG	Bayer Matrix filter colors specified from top left to bottom right. The Bayer matrix is incorporated into Profiles. Most raw DSLR files contain metadata that documents the matrix but it may be unknown for FITS and TIFF files, in which case use a Viewer to figure out the matrix. BG RG GB GR GR GB RG BG
Bayer Method	Interpolate Extract Drizzle RetainMatrix	Method of interpolating pixel colors from the Bayer matrix · Interpolate creates a full-size color image from each Bayer frame by filling-in all the missing colors using averages of near-by pixels of the same color. Bayer interpolation blurs signal and smoothes apparent noise. · Extract creates a half-sized color image consisting of like-colored pixels extracted from the Bayer matrix. The smaller color image consumes much less memory and is recommended for large images on 32 bit OS computers. · Drizzle creates a color frame where each pixel is only one color (red green or blue). When registering, use Nearest Neighbor to map each colored pixel to a single pixel location in the registered image. Drizzle produces maximal resolution and preserves original noise characteristics but generally requires at least 30 offset (dithered) frames or more to produce a good combined image. · RetainMatrix performs no transformation. This is used for tasks like creation of master darks and flats, which should remain in the Bayer format.
Retain DNG	Yes/No	Do not delete the DNG (Digital Negative) files All RAW DSLR images are converted to Adobe Digital Negative (DNG) as an intermediary file. No (default) = delete the *.dng file Yes = do not delete the *.dng file DNG files are placed in the same directory as the source file and have the same base file name with the “.dng” extension. By default, each DNG file is removed after conversion, thus only one DNG file is present at any time and none are left after the run. Setting this directive prevents the removal of the DNG files, thus a DNG file is created and retained for every raw DSLR input file. In subsequent runs, these retained DNG files are automatically recognized and used, thus improving performance.

Color Parms	Values	Description
Extract Color	RGB Red Green Blue	Color(s) to extract from the color image · RGB extracts the full color image · Red, Green, or Blue creates a monochrome image consisting only of the color selected. This is useful for making low-noise images from filtered frames (e.g. H-a only illuminates the Red pixels). Single-color extraction conserves memory: extract and processing each color in a separate Stack then combining the three results in a final Stack (all operations can be contained in a single Job).
Ratio Red	number	Red filter ratio. All Red pixel are multiplied by Ratio Red
Ratio Green	number	Green filters ratio. All Green pixel are multiplied by Ratio Green
Ratio Blue	number	Blue filter ratio. All Blue pixel are multiplied by Ratio Blue
Bias Offset	number	Average Bias or pedestal. = average camera bias ADU (DN) for images that have not been dark subtracted = 0 for images dark subtracted via Stacker Unlimited or CCDStack = 100 for most 16 bit images that are dark-subtracted in other software

InFile Parms	Values	Description
		Help using this Form or Data Grid see [Input Image Type] for supported file formats
File Name	text	Image File Full Name fully qualified (e.g. c:\StackerUnlimited\images\filename.fit) or relative to the directory from which Stacker Unlimited starts (e.g. images\filename.fit) In Forms mode: click on this textbox to open a File Dialog, which can accept multiple file selections.
Assign To Color	Grayscale Red Green Blue RGB	Assign this image to the selected color · Grayscale results in a grayscale Stack image. · RGB applies to Bayer or Color images (Bayer or Color Directives must set [Extract Color] = RGB) · Red, Green, or Blue assigns this image to that color component of the Color Stack image (e.g. a red filtered image would typically be assigned to Red). In Forms mode: use the Synchronize button to copy this setting to all following files.
Pre Process Directives	Pre Process Parms	Dark Subtraction and/or Flat Field directives In Forms mode: use the Synchronize button to copy these settings to all following files.
AVI Directives	AVI Parms	AVI video stream directives

PreProcess Parms	Values	Description
Dark Subtract	yes/no	Dark subtract this image. Requires valid [Dark Frame File Name]
Dark Frame File Name	Text	Dark File Name fully qualified or relative to the directory from which Stacker Unlimited starts
Adaptive Dark Subtract	yes/no	Use Adaptive Dark to dark-subtract this image. · [Dark Subtract] must be true · Requires valid [Dark Frame File Name] · Requires valid [Bias File Name]
Bias Frame File Name	Text	Bias File Name fully qualified or relative to the directory from which Stacker Unlimited starts
Bias Level	number	Average bias ADU (including pedestal) of un-subtracted frame. [Bias Level] is used when: · Flatting without dark-subtracting · Adaptive Dark Subtract with no [Bias File Name]
Flat Field	yes/no	Flat Field this image. Requires valid [Flat Frame File Name]
Flat Frame File Name	Text	Flat File Name fully qualified or relative to the directory from which Stacker Unlimited starts
Flat Bias Level	number	Average non-photonic bias (including pedestal) If the flat has been dark subtracted then level = pedestal (usually 0 or 100) Else use the mean ADU of a bias frame
SaveImages	File Save	Directives for automatically saving the Pre Processed (calibrated) image

AVI Parms	Values	Description
AVI Start Frame	integer	Start frame number (base = 1) Default = 1
AVI End Frame	integer	Last frame number (base = 1) Default = last frame in file

Align Parms	Values	Description
AlignmentMethod	StarMatch OLS StarShift None	Method of Alignment · StarMatch – automatic star detection and matching. Images must be nearly the same scale and contain several well formed stars. · OLS (Ordinary Least Squares regression). Currently [x,y] shift only. Use small area around a bright object for fast processing. OLS aligns planets, moon, comets, and terrestrial scenes. · StarShift – [x,y] shift based on centroid(s) of specified stars · None – default (use for darks and flats)
Resample Method	NearestNeighbor Bilinear BicubicBSpline None	Algorithm for resampling Registered image · NearestNeighbor - Pixels from the source image are transformed to the nearest point in the resampled image. Nearest Neighbor produces a one-to-one pixel correspondence and thus noise statistics are preserved in the resampled image. And because Nearest Neighbor does not combine source-image pixels there is no smearing of hot or cold pixels in the resampled image, which can be an advantages for data rejection. Nearest Neighbor should be used for Bayer drizzle extraction. Nearest Neighbor should be avoided for small stacks (fewer than 20 images), especially for under-sampled images. · Bilinear - Pixels in the resampled image are imputed from a small matrix of pixels in the source image. Bi-linear interpolation produces sub-pixel accuracy and works very well on non-magnifying and non-rotating transforms. · BicubicBSpline - Bicubic B-Spline produces sub-pixel accuracy and works well for rotation. This method produces the smoothest result with the fewest artifacts but also imparts a slight blur to undersampled images. Works well on small stacks (few images). · None – image is not resampled
Reference Color	Red Green Blue RGB	Color used for alignment applies to Bayer and Color images Beta-I - RGB is not implemented (Green is used by default)
Align Star Match Directives	Align Star Match Parms	Directives for automatic star matching alignment (optional)
Align OLS Directives	Align OLS Parms	Directives for OLS alignment (required for OLS) Beta-I only performs [x,y] integer shifts
Align Star Shift Directives	Align Star Shift Parms	Directives for Star Shift alignment (required for Star Shift) [x,y] shift based on centroid(s) of specified stars
Save Images	File Save	Directives for automatically saving Registered images

Align Star Match Parms	Values	Description
Reference Image File	text	Alignment Reference Image (optional) If this directive is blank or invalid then the first image in [Image Files] is used for reference and only 1 thread will be used to process the first image (subsequent images will utilize [Number Of Threads] specified).
Auto Settings	yes/no	Yes = Automatically calculate the settings below for each frame No = apply all settings below to each frame
Star Detection Sub Area	rectangle	Restrict star detection to the named area (only applied when Auto Settings = no)
Max ADU	number	Stars having ADU > [Max ADU] are ignored Set this to the saturation ADU of the camera Default = 50,000 (applicable for 16 bit camera) (only applied when Auto Settings = no)
Threshold	number	Stars having ADU < [Threshold] are ignored Default = 500 (only applied when Auto Settings = no)
Star Seperation Radius	integer	Distance in pixels between selected stars Default = 100 (only applied when Auto Settings = no)

Align OLS Parms	Values	Description
Reference Image File	text	Alignment Reference Image (optional) If this directive is blank or invalid then the first image in [Image Files] is used for reference and only 1 thread will be used to process the first image (subsequent images will utilize [Number Of Threads] specified).
Reference Area	Rectangle	This area in the reference file is used to find the most similar area in each aligned image example: {X=250,Y=300,Width=20,Height=20} (Use CCDStack to drag-out the desired area then copy and paste the Rectangle data from the Information window) Smaller areas execute faster than larger areas
OLS Compression	integer > 0	Speeds performance by compressing (binning) the affected areas Compression = 1 uses no compression Compression > 1 bins the first pass by that amount then finishes the alignment with no compression, so compression has no effect on accuracy unless over-compressed areas result in false matches (i.e. oversampled images can sustain high compression but undersampled images should use small compression) Default = 3 (for well sampled images)
Max Offset	integer	Specify maximum pixel shift Max Offset must be at least as large as the actual [x,y] offsets of the aligned images. If the actual offset exceeds this amount then the registration will be faulty and unpredictable. Small maximum offsets execute much more quickly than do large offsets

Align Star Shift Parms	Values	Description
Reference Stars	Location array	Location of reference star(s) in reference image (required). Example: {X=123.55,Y=255.66} multiple stars are averaged to obtain [x,y] shifts (Use CCDStack to select the desired star(s) then copy and paste the Centroid data from the Information window)
Star Grab Width	number	Star Shift grabs nearby stars within an area around the star Specify a large grab width (20) for well isolated reference star(s) and smaller width for crowded star field. This is a Beta-I directive that should not be tampered with unless necessary Default = 20

Selection Parms	Values	Description
Select FWHM	yes/no	Select based on FWHM of specified stars. This is not exclusive with Star Zone Ratios (both criteria may be used)
FWHM Maximum	number	Maximum allowable FWHM (measured in pixels). Frames with FWHM > Maximum are rejected.
Select Encircled Flux Ratio	yes/no	Select based on encircled flux ratios of specified stars. This is not exclusive with Star Zone Ratios (both criteria may be used)
Ratio Minimum	number	Minimum allowable ratio of inner / outer
Inner Radius	number	Number of pixels from center of star-center pixel A circle is computed from the radius and any pixels within the circle are used (the center of the pixel is used to determine its exact location). For example, radius = 0 results in only 1 pixel 1 results in 5 pixels 1.5 results in 7 pixels etc.
Outer Radius	number	Number of pixels from center of star central pixel. Outer Radius must be larger than Inner Radius.
Stars	Location array	Location of selection star(s). Example: {X=123.55,Y=255.66} Results of multiple stars are averaged. (Use CCDStack to select the desired star(s) then copy and paste the Centroid data from the Information window)

Combine Parms	Values	Description
Create Combined Image	yes/no	Create combined (stacked) image This may be set to “no” for input image viewing (“video mode”) or calibration only, etc. Default = yes
Normalize	yes/no	Normalize each component image to reference image Reference image is the first image in Image Files Use to normalization to compensate for different sky conditions, exp times, aperture, etc. Normalization is usually not required for AVI or short sessions using equal exp times Beware of normalizing differently filtered images, as this will change the color offsets and ratios. If normalization of multi-filtered images is desired then it may be preferable to construct RGB images using 3 or 4 stacks (steps) - one stack per color plus a stack to combine colors (or combine in other software). Default = no
Data Reject Max N	integer	The number of maximum (highest ADU) values to reject from each pixel-stack Data rejection consumes memory and high settings may exceed the available memory for large color images on restricted computers (e.g. 32 bit OS), in which case consider lowering these settings and/or processing one color at a time.
Data Reject Min N	integer	The number of minimum (lowest ADU) values to reject from each pixel-stack Data rejection consumes memory and high settings may exceed the available memory for large color images on restricted computers (e.g. 32 bit OS), in which case consider lowering these settings and/or processing one color at a time.
Save Image	File Save	Automatically save the Combined image. Note that it is also possible to save the final image in the Stack Viewer (if used) via the main form Save button.

File Save	Values	Description
Save	yes/no	Save File(s)
Save Path	Text	Path of output file(s) (optional) Only need to specify Path if it is different from source image’s path
Save File Name	Text	Saved file name (optional) If full path is included then it overrides [Save Path] If this is blank then the source file name is combined with [Save Suffix] to create the saved file name
Save Suffix	Text	Suffix appended to saved file name (optional) Use when not specifying [Save File Name] to avoid overwriting the source file Example: set Pre Processed (calibrated) Suffix = “_cal”: input file = image001.fit output file = image001_cal.fit
Save Type	Output file types	Save file as 16 bit TIFF, 16 bit FITS or 32 bit FITS 16 bit TIFF (mono or color) 16 bit FITS unsigned integers (mono or color) 32 bit FITS floating point (mono or color)
Save Multiplier	number	Multiply every pixel value in saved image file Use to optimize dynamic range of cameras with less than 16 bit resolution For example, the highest value for a 12 bit frame is 4,095. A stacked image made from many 12 bit frames may contain significant information that is lost (truncated) when converting floating point precision to integer for TIFF or 16 bit FITS. Most or all of that information can be recovered by multiplying each floating-point pixel by 16 prior to converting to integer.