AutoIntegrate Options Reference

If selected do not try to autodetect light and flat files based on FILTER keyword.

Selecting this enables manual adding of filter files for lights and flats.

If selected do not try to autodetect calibration files based on IMAGETYP keyword.

Use file name for recognizing filters and ignore FILTER keyword.

Add light files manually by selecting files for each filter.

If checked consider light files to be integrated files for AutoContinue.

It is useful for example when using integrated lights from WBPP as there is no need to rename images.

Add flat files manually by selecting files for each filter.

Add flat dark files manually by selecting files for each filter.

Select bayer pattern for debayering color/OSC/RAW/DSLR files.

Auto option tries to recognize debayer pattern from image metadata.

If images are already debayered choose none which does not do debayering.

Values: Auto, RGGB, BGGR, GBRG, GRBG, GRGB, GBGR, RGBG, BGRG, None

Extract channels from color/OSC/RAW/DSLR files.

Channel extraction is done right after debayering. After channels are extracted processing continues as mono processing with separate filter files.

Option LRGB extract lightness channels as L and color channels as separate R, G and B files.

Option HOS extract channels as RGB=HOS and option HSO extract channels RGB=HSO. Resulting channels can then be mixed as needed using PixMath expressions in Settings / Narrowband processing section.

Channel files have a channel name (_L, _R, etc.) at the end of the file name. Script can then automatically recognize files as filter files.

Values: None, LRGB, HSO, HOS

Create SuperBias from bias files.

For modern CMOS sensors you should not use superbias.

For CCD sensors you can use superbias.

Use bias files on light frames.

If there are master dark files, bias files are ignored on light frames unless explicitly enabled.

Use bias files on dark frames.

For modern CMOS sensors you should not use bias files on dark frames.

For CCD sensors you can use bias files on dark frames.

Files are master files.

When there are multiple master files, they are matched by resolution.

If checked darks are pre-calibrated with bias and not during ImageCalibration.

If checked darks are optimized when calibrating lights.

For modern CMOS cameras optimization should not be used.

For CCD cameras it can be used.

Files are master files.

When there are multiple master files in a group, they are matched by resolution.

Files are master files.

When there are multiple master files in a group, they are matched by resolution.

Files are master files.

When there are multiple master files in a group, they are matched by resolution.

If you have stars in your flats then checking this option will lower percentile clip values and should help remove the stars.

If selected then darks are used for flat calibration.

Use automatic output pedestal when calibrating lights.

Pedestal value added when calibrating lights if auto output pedestal is not checked. Value is from range 0 to 65535.

If checked, fix linear column defects by using linear defect detection algorithm from LinearDefectDetection.js script. Defect information is used by CosmeticCorrection to fix the defects.

If checked, fix linear row defects by using linear defect detection algorithm from LinearDefectDetection.js script. Defect information is used by CosmeticCorrection to fix the defects.

Do not run CosmeticCorrection on image files.

Can be useful when doing Comet align.

Hot Sigma and Cold Sigma values for CosmeticCorrection

Hot Sigma and Cold Sigma values for CosmeticCorrection

Do Canon banding reduction using the method in CanonBandingReduction script.

Banding reduction is run for each light file separately after cosmetic correction and possible debayering.

Protection for highlights.

Reduction amount. An amount less than 1.0 is often necessary with fainter banding.

Do not run SubframeSelector to get image weights.

When this option is used then the first image in the list is used as a reference image unless reference image is selected manually.

Use SSWEIGHT weight keyword during ImageIntegration.

Generic - Use both noise and stars for the weight calculation. Noise - More weight on image noise. Stars - More weight on stars. PSF Signal - Use PSF Signal value as is. PSF Signal scaled - PSF Signal value scaled by AutoIntegrate to 1-100. FWHM scaled - FWHM value scaled by AutoIntegrate to 1-100. Eccentricity scaled - Eccentricity value scaled by AutoIntegrate to 1-100. SNR scaled - SNR value scaled by AutoIntegrate to 1-100. Star count - Star count value.

All values are scaled so that bigger value is better.

Values: Generic, Noise, Stars, PSF Signal, PSF Signal scaled, FWHM scaled, Eccentricity scaled, SNR scaled, Star count

Limit value for SSWEIGHT. If value for SSWEIGHT is below the limit it is not included in the set of processed images.

Note that the value is written to the FITS header using 10 digits. Smaller than 10 digit limit values should not be used if using the SSWEIGHT value that is written to FITS header.

Sort order for the filter and sort button.

Values: SSWEIGHT, FWHM, Eccentricity, PSFSignal, PSFPower, SNR, Stars, File name

Choose filter measure and value. FWHM and Eccentricity are filtered for too high values (min), and all others are filtered for too low values (max).

Limit value zero (0.0000) means that there is no filtering.

Values: None, FWHM, Eccentricity, PSFSignal, PSFPower, SNR, Stars

Choose filter measure and value. FWHM and Eccentricity are filtered for too high values (min), and all others are filtered for too low values (max).

Limit value zero (0.0000) means that there is no filtering.

Choose filter measure and value. FWHM and Eccentricity are filtered for too high values (min), and all others are filtered for too low values (max).

Limit value zero (0.0000) means that there is no filtering.

Values: None, FWHM, Eccentricity, PSFSignal, PSFPower, SNR, Stars

Choose filter measure and value. FWHM and Eccentricity are filtered for too high values (min), and all others are filtered for too low values (max).

Limit value zero (0.0000) means that there is no filtering.

Choose filter measure and value. FWHM and Eccentricity are filtered for too high values (min), and all others are filtered for too low values (max).

Limit value zero (0.0000) means that there is no filtering.

Values: None, FWHM, Eccentricity, PSFSignal, PSFPower, SNR, Stars

Choose filter measure and value. FWHM and Eccentricity are filtered for too high values (min), and all others are filtered for too low values (max).

Limit value zero (0.0000) means that there is no filtering.

Choose filter measure and value. FWHM and Eccentricity are filtered for too high values (min), and all others are filtered for too low values (max).

Limit value zero (0.0000) means that there is no filtering.

Values: None, FWHM, Eccentricity, PSFSignal, PSFPower, SNR, Stars

Choose filter measure and value. FWHM and Eccentricity are filtered for too high values (min), and all others are filtered for too low values (max).

Limit value zero (0.0000) means that there is no filtering.

Skipping outliers can be useful when processing very large data sets and manual filtering gets too complicated

Option 'SSWEIGHT' will filter out outliers based on the calculated SSWEIGHT value. It is an alternative to using a fixed Limit value.

All other options will filter out outliers based on individual values.

Skipping outliers can be useful when processing very large data sets and manual filtering gets too complicated

Option 'SSWEIGHT' will filter out outliers based on the calculated SSWEIGHT value. It is an alternative to using a fixed Limit value.

All other options will filter out outliers based on individual values.

Skipping outliers can be useful when processing very large data sets and manual filtering gets too complicated

Option 'SSWEIGHT' will filter out outliers based on the calculated SSWEIGHT value. It is an alternative to using a fixed Limit value.

All other options will filter out outliers based on individual values.

Skipping outliers can be useful when processing very large data sets and manual filtering gets too complicated

Option 'SSWEIGHT' will filter out outliers based on the calculated SSWEIGHT value. It is an alternative to using a fixed Limit value.

All other options will filter out outliers based on individual values.

Skipping outliers can be useful when processing very large data sets and manual filtering gets too complicated

Option 'SSWEIGHT' will filter out outliers based on the calculated SSWEIGHT value. It is an alternative to using a fixed Limit value.

All other options will filter out outliers based on individual values.

Skipping outliers can be useful when processing very large data sets and manual filtering gets too complicated

Option 'SSWEIGHT' will filter out outliers based on the calculated SSWEIGHT value. It is an alternative to using a fixed Limit value.

All other options will filter out outliers based on individual values.

Skipping outliers can be useful when processing very large data sets and manual filtering gets too complicated

Option 'SSWEIGHT' will filter out outliers based on the calculated SSWEIGHT value. It is an alternative to using a fixed Limit value.

All other options will filter out outliers based on individual values.

Different methods are available for detecting outliers.Two sigma filters out outliers that are two sigmas away from mean value.

One sigma filters out outliers that are one sigmas away from mean value. This option filters more outliers than the two other options.

Interquartile range (IQR) measurement is based on median calculations. It should be relatively close to two sigma method.

Values: Two sigma, One sigma, IQR

If checked outliers are filtered using both min and max outlier threshold values.

By default FWHM and Eccentricity are filtered for too high values, and SNR and SSWEIGHT are filtered for too low values.

Sensitivity setting. Bigger value will detect fainter stars.

• Sensitivity: 0.70
• Noise reduction

If the star aligment fails you can try to change values. Here is one suggestion of values that might help: If you have very bad distortion then also increasing maximum distortion can help.

Maximum star distortion setting. Bigger value will detect more irregular stars.

• Sensitivity: 0.70
• Noise reduction

If the star aligment fails you can try to change values. Here is one suggestion of values that might help: If you have very bad distortion then also increasing maximum distortion can help.

Structure layers setting. Bigger value will detect more stars.

• Sensitivity: 0.70
• Noise reduction

If the star aligment fails you can try to change values. Here is one suggestion of values that might help: If you have very bad distortion then also increasing maximum distortion can help.

Noise reduction filter radius layers setting. Bigger value can help with very noisy images.

• Sensitivity: 0.70
• Noise reduction

If the star aligment fails you can try to change values. Here is one suggestion of values that might help: If you have very bad distortion then also increasing maximum distortion can help.

If checked, run CometAlign process using settings in the Preprocessing / CometAlignment settings section.

For more details see the help icon in CometAlignment settings section.

Coordinates for the first comet image.

Below is the suggested workflow with comet processing in AutoIntegrate:

• Run a normal workflow to get correct stars and background objects.
• Load star aligned *_r.xisf files as light files. Those can be found from the AutoOutput directory.
• Set a Window prefix to avoid overwriting files in the first step.
• Check Comet align in Settings / Image processing parameters section.
• Check star removal option (StarXTerminator or StarNet2) in Settings / Tools section.
• Check Remove stars from lights in Postprocessing / Star stretching and removing section.
• Check No CosmeticCorrection in Other / Other parameters section.
• Go to the Preprocessing / CometAlignment section.
• Fill in first and last comet position coordinates. To get the coordinates click the Preview button for the first or last image, go to preview, zoom to 1:1 view and click the comet nucleus with the left mouse button. Note that the first and last images are selected automatically based on image timestamps from the DATE-OBS keyword when images are loaded.
• Copy coordinates from the preview coordinates box and paste them to the comet coordinates box. There are arrow buttons in the preview to automatically copy coordinates.
• Use the Run button to process images.Comet alignment will automatically skip star alignment and SCNR. Since already star aligned images (*_r.xisf) are used then Star alignment could invalidate coordinates given here and thus it is not used.

Note that using starless images may cause problems for example with ImageIntegration or BlurXTerminator. With missing PSF error in ImageIntegration you can use an option ImageIntegration use ssweight. Sometimes on starless images PSF value can not be calculated. Then a manual value should be given or BlurXTerminator should not be used.

It is also possible to manually run the CometAlignment process. Below are the steps to use AutoIntegrate with manual comet alignment:

• Run normal workflow to get correct stars and background objects.
• Manually run the CometAlignment on star aligned *_r.xisf files. This will create *_ca.xisf files.
• Remove stars from *_ca.xisf files. StarXTerminator has a batch mode that makes this easier.
• Load comet aligned files into AutoIntegrate as lights files.
• Check Start from ImageIntegration in Other parameters.
• Use the Run button to process images.

Coordinates for the last comet image.

Below is the suggested workflow with comet processing in AutoIntegrate:

• Run a normal workflow to get correct stars and background objects.
• Load star aligned *_r.xisf files as light files. Those can be found from the AutoOutput directory.
• Set a Window prefix to avoid overwriting files in the first step.
• Check Comet align in Settings / Image processing parameters section.
• Check star removal option (StarXTerminator or StarNet2) in Settings / Tools section.
• Check Remove stars from lights in Postprocessing / Star stretching and removing section.
• Check No CosmeticCorrection in Other / Other parameters section.
• Go to the Preprocessing / CometAlignment section.
• Fill in first and last comet position coordinates. To get the coordinates click the Preview button for the first or last image, go to preview, zoom to 1:1 view and click the comet nucleus with the left mouse button. Note that the first and last images are selected automatically based on image timestamps from the DATE-OBS keyword when images are loaded.
• Copy coordinates from the preview coordinates box and paste them to the comet coordinates box. There are arrow buttons in the preview to automatically copy coordinates.
• Use the Run button to process images.Comet alignment will automatically skip star alignment and SCNR. Since already star aligned images (*_r.xisf) are used then Star alignment could invalidate coordinates given here and thus it is not used.

Note that using starless images may cause problems for example with ImageIntegration or BlurXTerminator. With missing PSF error in ImageIntegration you can use an option ImageIntegration use ssweight. Sometimes on starless images PSF value can not be calculated. Then a manual value should be given or BlurXTerminator should not be used.

It is also possible to manually run the CometAlignment process. Below are the steps to use AutoIntegrate with manual comet alignment:

• Run normal workflow to get correct stars and background objects.
• Manually run the CometAlignment on star aligned *_r.xisf files. This will create *_ca.xisf files.
• Remove stars from *_ca.xisf files. StarXTerminator has a batch mode that makes this easier.
• Load comet aligned files into AutoIntegrate as lights files.
• Check Start from ImageIntegration in Other parameters.
• Use the Run button to process images.

Do binning for each light file. Binning is done first on calibrated light files before any other operations.With Color option binning is done only for color channel files.With L and Color option binning is done for both luminance and color channel files.Binning uses IntegerResample process and should help to reduce noise at the cost of decreased resolution.

Values: None, Color, L and color

Resample factor for binning.

Do binning for each light file. Binning is done first on calibrated light files before any other operations.With Color option binning is done only for color channel files.With L and Color option binning is done for both luminance and color channel files.Binning uses IntegerResample process and should help to reduce noise at the cost of decreased resolution.

Crop all channels to area covered by all images

Use Drizzle integration

Drizzle scale 1 does not change the image size but may help with fine details like stars in the image.

Drizzle scale 2 doubles the image resolution and may help with small details in the image.

Consider using drizzle when selecting SPCC for color calibration.

For Drizzle settings see Integration / Drizzle section.

Use Drizzle integration

Drizzle scale 1 does not change the image size but may help with fine details like stars in the image.

Drizzle scale 2 doubles the image resolution and may help with small details in the image.

Consider using drizzle when selecting SPCC for color calibration.

For Drizzle settings see Integration / Drizzle section.

Drop shrink value for drizzle.

Number of consecutive bad pixels allowed before detecting crop edge.

Use rejection_low from ImageIntegration instead of integrated data to calculate crop amount.

Limit value for detecting crop edges. Values below the limit are considered to be inside the cropped area.

This value is used only if rejection low is selected.

Warning limit value in percentages. If image is cropped more than the warning limit percentage a warning message is printed at the end of processing.

After a warning you should manually check the cropping area.

Drizzle drop kernel function.

Values: Square, Circular, Gaussian

Use fast mode for drizzle integration.

Use local normalization data for ImageIntegration

For local normalization settings see section Integration / LocalNormalization

If checked, use FastIntegration process instead of ImageIntegration process when integrating light images.

In Integration tab there are some settings for FastIntegration.

Do not use clipping in ImageIntegration

Rejection normalization. This is value is ignored if local normalization is used.

Values: Additive, Adaptive, None

Pixel combination operation

Values: Average, Median, Minimum, Maximum

Auto1

• Percentile clipping for 1-7 images
• Sigma clipping otherwise.

Auto2

• Percentile clipping for 1-7 images
• Sigma clipping for 8 - 19 images
• Linear fit clipping for 20 - 24 images
• ESD clipping for more than 25 images

Percentile - Percentile clip

Sigma - Sigma clipping. If you are not happy with the rejection result you can lower the High sigma for example to 2.8.

Winsorised - Winsorised sigma clipping. To remove satellite trails with Winsorised sigma use lower high sigma value like 2.2 or even 1.8.

Averaged - Averaged sigma clipping. If you are not happy with the rejection result you can lower the High sigma for example to 2.8.

Linear - Linear fit clipping

ESD - Extreme Studentized Deviate clipping. If you are not happy with the rejection result you can try higher ESD Significance like 0.2 and lower Low relaxation like 1.0.

None - No rejection. Useful for example with a blown out comet core.

Values: Auto1, Auto2, Percentile, Sigma, Averaged sigma, Winsorised sigma, Linear fit, ESD, None

Percentile low clipping factor.

Percentile high clipping factor.

Sigma low clipping factor.

Sigma high clipping factor.

If you are not happy with the rejection result you can lower the High sigma for example to 2.8.

To remove satellite trails with Winsorised sigma use lower high sigma value like 2.2 or even 1.8.

Cutoff point for Winsorised sigma clipping.

Tolerance of low values for linear fit low clipping.

Tolerance of high values for linear fit low clipping.

ESD outliers.

ESD significance.

If you are not happy with the rejection result you can try higher ESD Significance like 0.2 and lower Low relaxation like 1.0.

Enable large scale rejection of high pixels.

Large scale rejection of high pixels may help for example with satellite trails.

Enable large scale rejection of low pixels.

During local normalization use multiscale analysis instead of PSF flux evaluation.

Using multiscale analysis may help if you get errors like PSFScaleEstimator::EstimateScale(): Internal error: No reference image has been defined.

Increase the value if you have a lot of drift between images.

Increase the value if you have a dense starfield. Smaller values can be used for sparse starfields.

Alignment error tolerance. You can try increasing the value if alignment fails.

Run SubframeSelector only to a subset of images when using FastIntegration. SubframeSelector is used only to find the reference image.

When this option is used then the first 32 images in the list are used to find the reference image unless reference image is selected manually.

Do not run CosmeticCorrection when using FastIntegration.

With a very large number of files it usually should not be necessary.

Use gradient correction on channel images (L,R,G,B,H,S,O) separately before channels are combined.

With color/OSC images this does the same thing as Gradient correction on combined images.

Use gradient correction on all light images. It is run very early in the processing before cosmetic correction.

Use gradient correction on L and combined RGB images while image is still in linear mode.

Use gradient correction on L and RGB images after they have been stretched to non-linear mode.

Note that this option should not be used with GradientCorrection process.

Use GraXpert instead of GradientCorrection process to correct gradients in images.By default no gradient correction is done. To use GraXpert for gradient correction you need to also check one of the gradient correction options in the Settings / Image processing parameters section.

GraXpert always uses the AI background model. In the Tools / GraXpert section it is possible to set some settings.

NOTE! A path to GraXpert file must be set in the Tools / GraXpert section before it can be used.

NOTE2! You need to manually start GraXpert once to ensure that the correct AI model is loaded into your computer.

Run AutomaticBackgroundExtractor (ABE) to correct gradients in images.By default no gradient correction is done. To use ABE for gradient correction you need to also check one of the gradient correction options in the Settings / Image processing parameters section.

Settings for ABE are in Postprocessing / ABE settings section.

Use DynamicBackgroundExtraction (DBE) to correct gradients in images.By default no gradient correction is done. To use DBE for gradient correction you need to also check one of the gradient correction options in the Settings / Image processing parameters section.

Sample points are automatically generated for DBE. Settings for DBE are in Postprocessing / DBE settings section.

Use MultiscaleGradientCorrection instead of GradientCorrection process to correct gradients in images.By default no gradient correction is done. To use MultiscaleGradientCorrection for gradient correction you need to also check one of the gradient correction options in the Settings / Image processing parameters section.

Settings for MultiscaleGradientCorrection are in Postprocessing / Gradient correction section.

Note that you need to set up MARS database settings using the PixInsight MultiscaleGradientCorrection process before using this option.

When MultiscaleGradientCorrection is selected, image solving and SpectrophotometricFluxCalibration are run automatically for the image.

MultiscaleGradientCorrection may fail if the image is not part of the sky area in the MARS database. In that case the script reverts to another gradient correction method. If other gradient correction methods are checked then they are selected in the following order: GraXpert, ABE, DBE, GradientCorrection.

If checked the background model is created.

Model scale.

Higher values generate smoother models.

Model smoothness.

Run multiple iterations until difference between two models is small enough.

Prevent overcorrecting on image structures.

Decreasing this value prevents overcorrecting dimmer structures.

Increasing this value prevents overcorrecting significant structures.

If checked use a simplified model that is extracted before multiscale model.

Model degree for simplified model.

Gradient model scale.

Values: 128, 192, 256, 384, 512, 768, 1024, 1536, 2048, 3072, 4096, 6144, 8192

Scale factor for all channels.

Structure separation for MultiscaleGradientCorrection.

Function degree can be changed if ABE results are not good enough.

Correction method for ABE.

Values: Subtraction, Division

If checked sets the normalize flag. Normalizing is more likely to keep the original color balance.

If checked background neutralization is run before DBE on color images.

If checked ABE with degree one is run before DBE.

Number of sample points placed for each row and column.

If checked sets the normalize flag. Normalizing is more likely to keep the original color balance.

Minimum sample weight to be included in the samples.

Selects how the sample weight is calculated from the sample region statistics.

Strict penalizes the sample using the full noise ratio of the region. Only very flat regions get a high weight.

Balanced penalizes only the noise that exceeds the noise level of the image. A region that is pure background noise gets full weight. This is the default.

Background noise works like Balanced but uses the measured noise level of the selected sample regions as a reference instead of the whole image noise level. It also uses the image noise level when detecting samples that are brighter than the background.

Values: Strict, Balanced, Background noise

Selects how the distance of the sample from the image edge affects the sample weight.

Neutral does not change the weight based on the distance from the edge. This is the default.

Favor center reduces the weight of samples close to the edge. This helps to prevent edge artifacts from affecting the background model.

Favor edges increases the weight of samples close to the edge. This helps the background model to better fit edge vignetting.

Values: Neutral, Favor center, Favor edges

Path to GraXpert executable.

NOTE! Parameter is automatically saved the persistent module settings. The value is automatically restored when the script starts.

Note that this parameter is not reset or saved to Json file.

Correction method for GraXpert.

Values: Subtraction, Division

Smoothing for GraXpert gradient correction.

Gives the limit value as percentage of shadows that is used for shadow smoothing. Smoothing is done before gradient correction.

Usually values below 50 work best. Possible values are between 0 and 100. Zero values does not do smoothing.

Smoothening should be used only in extreme cases with very uneven background because a lot of shadow detail may get lost.

Do not run color calibration. Color calibration is run by default on RGB data.

Do not try to find background area.

Run color calibration using SpectrophotometricColorCalibration (SPCC). This requires image solving which is done automatically on Integration_RGB image if it is not already done.

If image does not have correct coordinates or focal length embedded they can be given in Postprocessing / Image solving section.

SPCC settings can be updated at Postprocessing / Color calibration section.

Solve image by running ImageSolver script.

Note that if Color calibration using SPCC is selected image is solved automatically with checking this box.

If the image does not have correct coordinates or focal length embedded they can be given in Postprocessing / Image solving section.

Consider also using Drizzle with scale 1 or 2 when using SPCC.

Run BackgroundNeutralization before ColorCalibration.

By default the script tries to automatically detect an area with a true background. If it finds one, an image with name AutoBackgroundModel is created.

AutoBackgroundModel image has a preview that shows the background area.If AutoBackgroundModel image already exists then the background preview coordinates from there are used. The preview in AutoBackgroundModel image can be edited to use a different area as a background reference.

Automatic background detect can be turned off using option No auto background in Other tab.

Give target type.

If target type is given then image stretching settings are selected automatically.

If no target type is given then current settings are used. They should work reasonably fine in many cases.

Galaxy works well when target is a lot brighter than the background.

Nebula works well when target fills the whole image or is not much brighter than the background.

When non-default target type is selected then stretching option is disabled.

Values: Default, Galaxy, Nebula, Star cluster

Target name (optional).

Target RA DEC in decimal hours and degrees.

Should be given if target image does not contain coordinates.

Search button can be used to search coordinates.

Focal length in millimeters. Empty value uses image metadata.

Pixel size in μm. Empty value uses image metadata.

Target binning. Binning multiplies the pixel size by the binning value.

• Auto uses image metadata XBINNING value when available.
• None does not modify pixel size.
• Values 2 and 4 multiply the pixel size by those values.

Values: Auto, None, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10

Target drizzle. Drizzle divides the pixel size by the drizzle value.

• Auto uses image metadata 'DrizzleIntegration.scale:' value when available.
• None does not modify pixel size.
• Numeric values divide the pixel size by those values.

Values: Auto, None, 2, 4

Force solving images even if it already solved.

Can be useful for example when using old data with a newer SPCC version.

Try solving the image interactively using image solver dialog.

Number of layers used for structure detection. Larger value detects larger stars for signal evaluation. If SPCC fails you can try increasing the value.

Number of layers used for noise reduction. If SPCC fails you can try increasing the value.

Minimum size for a detectable star structure. Can be increased to avoid detecting image artifacts as real stars.

Wavelength and bandwidths for Red, Green and Blue filters with narrowband processing.

Default values are for Astrodon LRGB 2GEN filters using SHO palette.

Wavelength and bandwidths for Red, Green and Blue filters with narrowband processing.

Default values are for Astrodon LRGB 2GEN filters using SHO palette.

Wavelength and bandwidths for Red, Green and Blue filters with narrowband processing.

Default values are for Astrodon LRGB 2GEN filters using SHO palette.

Wavelength and bandwidths for Red, Green and Blue filters with narrowband processing.

Default values are for Astrodon LRGB 2GEN filters using SHO palette.

Wavelength and bandwidths for Red, Green and Blue filters with narrowband processing.

Default values are for Astrodon LRGB 2GEN filters using SHO palette.

Wavelength and bandwidths for Red, Green and Blue filters with narrowband processing.

Default values are for Astrodon LRGB 2GEN filters using SHO palette.

Enable SPCC for narrowband images and use narrowband filter values.

Do background neutralization during SPCC.

Automatically update narrowband mode, white reference and filters. Filters are selected automatically when a single filter is used.

Select white reference for SPCC.

Usually Average Spiral Galaxy is the best choice but for narrowband images Photon Flux should be used.

Values: Average Spiral Galaxy, Photon Flux

Limit magnitude for catalog search. Can be changed from Auto to something like 20 or larger if SPCC fails. Maximum value is 30.

If SPCC fails you can try increasing this to for example 0.90.

If SPCC fails you can try decreasing this value. You can for example try value 0.

Enable ColorCalibration for narrowband images.

When to run ColorCalibration process

• With auto the ColorCalibration process is run in both linear and nonlinear phase (non-stretched and stretched image). Running the ColorCalibration process twice seems to give a better final image in many cases. If SPCC is used for color calibration then with auto option the ColorCalibration process is not run.
• With linear the ColorCalibration process is run only in linear phase (non-stretched image).
• With nonlinear the ColorCalibration process is run only in nonlinear phase (stretched image).
• With both the ColorCalibration process is run in both linear and nonlinear phase (non-stretched and stretched image).When using the SPCC process the color calibration is always in a linear phase (non-stretched image).

Note that when using narrowband data color calibration is not run by default and must be enabled separately.

Values: auto, linear, nonlinear, both

• Auto STF - Use Auto Screen Transfer Function to stretch image to non-linear.

• MultiscaleAdaptiveStretch - Use MultiscaleAdaptiveStretch to stretch image to non-linear.

• Masked Stretch - Use MaskedStretch to stretch image to non-linear.

• VeraLuxHMS - Use VeraLux Hypermetric Stretch to stretch image to non-linear.

• Masked+Histogram Stretch - Use MaskedStretch with a Histogram Stretch prestretch to stretch image to non-linear.

• Histogram stretch - Using a simple histogram transformation to get histogram median or peak to the target value. Works best with images that are processed with the Crop to common area option.

• Arcsinh Stretch - Use ArcsinhStretch to stretch image to non-linear.

Select how image is stretched from linear to non-linear. For galaxies and other small but bright objects you should adjust targetBackground in Postprocessing / AutoSTF settings section to a smaller value, like 0.10 Parameters are set in Postprocessing / AutoSTF settings section. You can adjust settings in Postprocessing / MultiscaleAdaptiveStretch section. Useful when AutoSTF generates too bright images, like on some galaxies. Parameters are set in Postprocessing / Masked stretch settings section VeraLuxHMS can work well on many targets and can create saturated results. Parameters are set in Postprocessing / VeraLux HMS Stretch section Prestretch help with stars that can be too pointlike with Masked Stretch. Parameters are set in Postprocessing / Masked stretch settings and Postprocessing / Histogram stretching settings sections Parameters are set in Postprocessing / Histogram stretching settings section Can be useful when stretching stars to keep good star color. Parameters are set in Postprocessing / Arcsinh stretch settings sectionThere are also some experimental stretches.

• Logarithmic stretch - Parameters are set in Postprocessing / Other stretching settings section
• Asinh+Histogram stretch - Parameters are set in Postprocessing / Arcsinh stretch settings and Postprocessing / Histogram stretching settings sections
• Square root stretch - Parameters are set in Postprocessing / Other stretching settings section
• Shadow stretch - Parameters are set in Postprocessing / Other stretching settings section
• Highlight stretch - Parameters are set in Postprocessing / Other stretching settings section
• None - No stretching, mainly for generating _HT files to be used with AutoContinue.See Postprocessing / Stretching settings section for stretching specific parameters.

Note that when non-default Target type is selected then this option is disabled.

Values: Auto STF, MultiscaleAdaptiveStretch, Masked Stretch, VeraLuxHMS, Masked+Histogram Stretch, Histogram stretch, Arcsinh Stretch, Histogram direct, Logarithmic stretch, Asinh+Histogram stretch, Square root stretch, Shadow stretch, Highlight stretch, None

STF targetBackground value. If you get too bright image lowering this value can help.

Usually values between 0.1 and 0.250 work best. Possible values are between 0 and 1.

Masked Stretch targetBackground value. Usually values between 0.05 and 0.2 work best. Possible values are between 0 and 1.

Masked Stretch prestretch target value if Masked+Histogram Stretch is used.

Target value is a target median value. Using a prestretch can help with too pointlike stars.

Arcsinh Stretch Factor value. Smaller values are usually better than really big ones.

For some smaller but bright targets like galaxies it may be useful to increase stretch factor and iterations. A good starting point could be 100 and 5.

Useful for stretching stars to keep star colors. Depending on the star combine method you may need to use different values. For less stars you can use a smaller value.

Arcsinh Stretch black point value.

The value is given as percentage of shadow pixels, that is, how many pixels are on the left side of the histogram.

Number of iterations used to get the requested stretch factor.

MAS targetBackground value. Usually values between 0.1 and 0.250 work best. Possible values are between 0 and 0.50.

Controls the black point clipping. Higher values produce Better background but may loose some details.Possible values are between 0 and 1.

MAS dynamic range compression value. Higher values produce softer contrast in highlights. Possible values are between 0 and 1.

Enable or disable MAS contrast recovery.

Contrast recovery can help to recover some contrast lost during stretching.

If checked automatically finds and uses a background reference.

MAS scale separation value.

Smaller value protects smaller elements from local local contrast enhancements.

Values: 8, 16, 32, 64, 128, 192, 256, 384, 512, 768, 1024

MAS contrast recovery intensity.

Enable or disable MAS color saturation.

Color saturation can help to improve color saturation lost during stretching.

MAS color saturation amount.

Controls the intensity of the saturation enhancements. Higher values produce stronger color saturation.

MAS color saturation boost.

Higher values target more on lower saturation levels.

Enable or disable MAS color saturation affecting lightness.

If enabled color saturation affects are masked using a lightness mask.

VeraLux processing mode selection. Ready-to-Use (Aesthetic) Produces an aesthetic, export-ready image with adaptive expansion and soft-clipping. Scientific (Preserve) Produces 100% mathematically consistent output. Ideal for manual tone mapping.

Values: Ready-to-Use, Scientific

VeraLux sensor profile selection. Defines the Luminance coefficients (Weights) used for the stretch.

Values: Default, Rec.709 (Recommended), Sony IMX571 (ASI2600/QHY268), Sony IMX533 (ASI533), Sony IMX455 (ASI6200/QHY600), Sony IMX294 (ASI294), Sony IMX183 (ASI183), Sony IMX178 (ASI178), Sony IMX224 (ASI224), Sony IMX585 (ASI585) - STARVIS 2, Sony IMX662 (ASI662) - STARVIS 2, Sony IMX678/715 - STARVIS 2, Panasonic MN34230 (ASI1600/QHY163), Canon EOS (Modern - 60D/6D/R), Canon EOS (Legacy - 300D/40D), Nikon DSLR (Modern - D5300/D850), ZWO Seestar S50, ZWO Seestar S30, Narrowband HOO, Narrowband SHO, Narrowband Tricolor

Target background median (0.05-0.50). Standard is 0.20.

Analyzes histogram shape to find true signal start. Recommended for images with gradients.

Analyzes image to find optimal Stretch Factor (Log D).

Hyperbolic Intensity (Log D, 0.1-15). Controls the strength of the stretch.

Highlight Protection. Controls the 'knee of the hyperbolic curve.

Controls how quickly saturated colors transition to white (1-10).

Negative: Clean Noise | Center: Balanced | Positive: Soften Highlights. Only for Ready-to-Use mode. Value can be between -100 and 100.

Controls vector color preservation. 1.0 = Pure VeraLux (0-1). Only for Scientific mode.

Damps vector preservation in shadows to prevent color noise (0-3). Only for Scientific mode.

Select if shadows are adjusted before, after or before and after stretch.

Value zero just moves the histogram to the left without clipping any pixels.

Values: none, before, after, both

Percentage of shadows adjustment.

Value zero just moves the histogram to the left without clipping any pixels.

Target type specifies what value calculated from histogram is tried to get close to Target value.

Values: Median, Peak

Target value specifies where we try to get the the value calculated using Target type.

Usually values between 0.1 and 0.250 work best. Possible values are between 0 and 1.

For very bright objects like galaxies you should try with 0.1 while more uniform objects like large nebulas or dust you should try with 0.25.

Target value specifies where we try to get the the histogram median value.

Usually values between 0.1 and 0.250 work best. Possible values are between 0 and 1.

For very bright objects like galaxies you should try with 0.1 while more uniform objects like large nebulas or dust you should try with 0.25.

Use GraXpert for noise reduction.

In the Tools / GraXpert section it is possible to set some settings.

NOTE! A path to GraXpert file must be set in the Tools / GraXpert section before it can be used.

NOTE2! You need to manually start GraXpert once to ensure that the correct AI model is loaded into your computer.

Do not use noise reduction. This option disables all noise reduction regardless of what other noise reduction settings are selected.

More fine grained noise reduction settings can be found in the Postprocessing / Noise reduction section.

Do not use star noise reduction. Star noise reduction is used when stars are removed from image.

Amount of noise reduction.

Number of iterations for noise reduction.

Enable separate intensity and color noise reduction.

Amount of high frequency intensity noise to remove.

Amount of color noise to remove.

Enable separate low and high frequency noise reduction.

Amount of low frequency noise to remove.

Amount of high frequency noise to remove.

Pixel scale for transition between HF and LF noise reduction.

Amount of tile overlap for noise reduction.

Amount of noise reduction.

Use NoiseXTerminator for noise reduction.

You can change noise reduction settings in the Postprocessing / Noise reduction section.

Use DeepSNR for noise reduction.

Note that with DeepSNR increasing the noise reduction strength value will decrease the noise reduction.

Select automatically correct time for noise reduction.

If BlurXTerminator is used, then processed linear image noise reduction is used. Otherwise channel noise reduction is used except for OSC/color images where processed linear image is used.

Do noise reduction on each color channels and luminance image separately.

This option does nothing with color/OSC images.

Do noise reduction in non-linear state after stretching on combined and luminance images.

Noise reduction strength for color channel (R,G,B,H,S,O) or color images.

With MultiscaleLinerTransform the strength between 3 and 5 is the number of layers used to reduce noise. Strength values 1 and 2 are very mild three layer noise reductions and strength 6 is very aggressive five layer noise reduction.

Values: 0, 1, 2, 3, 4, 5, 6

Noise reduction strength for luminance image.

With MultiscaleLinerTransform the strength between 3 and 5 is the number of layers used to reduce noise. Strength values 1 and 2 are very mild three layer noise reductions and strength 6 is very aggressive five layer noise reduction.

Values: 0, 1, 2, 3, 4, 5, 6

Do noise reduction on combined image. Image can be from channel combination or from integrated color/OSC image.

On L image noise reduction is done before processing which is the same as channel noise reduction.

Do noise reduction on processed RGB image and possible luminance image in linear stage.

Do color noise reduction using TGVDenoise.

Runs ACDNR noise reduction.

If non-zero, sets StdDev value and runs ACDNR noise reduction.

A mild ACDNR noise reduction with StdDev value between 1.0 and 2.0 can be useful to smooth image and reduce black spots left from previous noise reduction.

Strength for GraXpert denoise.

Batch size for GraXpert denoise.

Values: 1, 2, 4, 8, 16, 32

Use GraXpert deconvolution for stellar and non-stellar sharpening.

In the Tools / GraXpert section it is possible to set some settings.

NOTE! A path to GraXpert file must be set in the Tools / GraXpert section before it can be used.

NOTE2! You need to manually start GraXpert once to ensure that the correct AI model is loaded into your computer.

Do not use sharpening on the image. Sharpening uses a luminance and star mask to target light parts of the image.

Use BlurXTerminator for sharpening and deconvolution.

BlurXTerminator is applied on the linear image just before it is stretched to non-linear. Enhancements option for sharpening can be used to apply BlurXTerminator on non-linear image.

Some options for BlurXTerminator can be adjusted in the Tools / BlurXTerminator section.

When using BlurXTerminator it is recommended to do noise reduction after BluxXTerminator by checking option Combined image noise reduction or Non-linear noise reduction. But it is always good to experiment what is best for your own data.

Sometimes on starless images PSF value can not be calculated. Then a manual value should be given or BlurXTerminator should not be used.

Amount to reduce the diameter of stars. Use a value between 0.00 and 0.50.

Amount to adjust star halos. Use a value between -0.50 and 0.50.

The amount to sharpen non-stellar image features. Use a value between 0.00 and 1.00.

Manual PSF value if a non-zero value is given.

Get PSF value from image using FWHM.

Sometimes on starless images PSF value can not be calculated. Then a manual value should be given or BlurXTerminator should not be used.

Use median FWHM from subframe selector as PSF value. It can be useful when PSF cannot be calculated from the image.

Value is saved to the FITS header and used if available. Value is also printed to the AutoIntegrate.log file with a name AutoIntegrateMEDFWHM.

Sometimes on starless images PSF value can not be calculated. Then a manual value should be given or BlurXTerminator should not be used.

Run BlurXTerminator in correct only mode before color calibration.

Run BlurXTerminator in correct only mode on channel images.

Strength for GraXpert stars deconvolution.

GraXpert deconvolution is used for stellar and non-stellar sharpening.

FWHM in pixels for GraXpert stars deconvolution.

GraXpert deconvolution is used for stellar and non-stellar sharpening.

Strength for GraXpert object deconvolution.

GraXpert deconvolution is used for stellar and non-stellar sharpening.

FWHM in pixels for GraXpert object deconvolution.

GraXpert deconvolution is used for stellar and non-stellar sharpening.

Use median FWHM from subframe selector as FWHM value.

Value is saved to the FITS header and used if available. Value is also printed to the AutoIntegrate.log file with a name AutoIntegrateMEDFWHM.

Do not add contrast on automatically created luminance mask.

Do not use an existing mask but always create a new mask.

Use unscreen method to get stars image as described by Russell Croman.

Unscreen method usually keeps star colors more correct than simple star removal. It is recommended to use Screen method when combining star and starless images back together.

Use StarXTerminator to remove stars from an image.

You can change some StarXTerminator settings in the Tools / StarXTerminator section.

Use StarNet2 to remove stars from an image.

Remove stars from combined RGB or narrowband images just before stretching while it still is in the linear stage. Stars are used only from RGB image, stars from L image are not used. This needs StarXTerminator.

When stars are removed before stretching then a different stretching can be used for the stars and potentially get better star colors.

For OSC data this may not work well. Separating channels might help.

Choose star image stretching and combining settings from Postprocessing / Star stretching and removing section.

Remove stars from light image.

Stars are removed after after star alignment.

If comet alignment is chosen then stars are removed before comet align.

With LRGB or narrowband images remove stars from L, R, G, B, H, S and O channel images separately after image integration. while images are still in the linear stage. Star images are then combined to create a RGB star image. This needs StarXTerminator.

With color images (DSLR/OSC) remove stars after image integration while image is still in linear stage. This needs StarXTerminator.

Choose star image stretching and combining settings from Postprocessing / Star stretching and removing section.

Remove stars after the image has been stretched to a non-linear state. Start from RGB image are saved and they can be later added back to the image. This needs StarXTerminator.

Choose star image stretching and combining settings from Postprocessing / Star stretching and removing section.

When both RGB and narrowband data is available, process stars image from RGB channels and process background image from the narrowband data. In the end of processing starless and stars images are combined to create a narrowband image with RGB stars.

This option can be used for RGB only processing but the end result is basically the same as using separate remove stars options.To use this option RGB channels must be available. If no option to remove stars is selected, stars are removed before streching.

Select AI model for AI model

Use the selected AI model. Default AI model is undefined

AI models are stored in the PixInsight installation directory and have a .pb extension. At least in Windows they are in PixInsight/library directory.

Uses large overlap on tiles to avoid tiling artifacts

• Auto STF - Use Auto Screen Transfer Function to stretch image to non-linear.

• MultiscaleAdaptiveStretch - Use MultiscaleAdaptiveStretch to stretch image to non-linear.

• Masked Stretch - Use MaskedStretch to stretch image to non-linear.

• VeraLuxHMS - Use VeraLux Hypermetric Stretch to stretch image to non-linear.

• Masked+Histogram Stretch - Use MaskedStretch with a Histogram Stretch prestretch to stretch image to non-linear.

• Histogram stretch - Using a simple histogram transformation to get histogram median or peak to the target value. Works best with images that are processed with the Crop to common area option.

• Arcsinh Stretch - Use ArcsinhStretch to stretch image to non-linear.

Select how image is stretched from linear to non-linear. For galaxies and other small but bright objects you should adjust targetBackground in Postprocessing / AutoSTF settings section to a smaller value, like 0.10 Parameters are set in Postprocessing / AutoSTF settings section. You can adjust settings in Postprocessing / MultiscaleAdaptiveStretch section. Useful when AutoSTF generates too bright images, like on some galaxies. Parameters are set in Postprocessing / Masked stretch settings section VeraLuxHMS can work well on many targets and can create saturated results. Parameters are set in Postprocessing / VeraLux HMS Stretch section Prestretch help with stars that can be too pointlike with Masked Stretch. Parameters are set in Postprocessing / Masked stretch settings and Postprocessing / Histogram stretching settings sections Parameters are set in Postprocessing / Histogram stretching settings section Can be useful when stretching stars to keep good star color. Parameters are set in Postprocessing / Arcsinh stretch settings sectionThere are also some experimental stretches.

• Logarithmic stretch - Parameters are set in Postprocessing / Other stretching settings section
• Asinh+Histogram stretch - Parameters are set in Postprocessing / Arcsinh stretch settings and Postprocessing / Histogram stretching settings sections
• Square root stretch - Parameters are set in Postprocessing / Other stretching settings section
• Shadow stretch - Parameters are set in Postprocessing / Other stretching settings section
• Highlight stretch - Parameters are set in Postprocessing / Other stretching settings section
• None - No stretching, mainly for generating _HT files to be used with AutoContinue.See Postprocessing / Stretching settings section for stretching specific parameters.

Note that when non-default Target type is selected then this option is disabled.

Values: Auto STF, MultiscaleAdaptiveStretch, Masked Stretch, VeraLuxHMS, Masked+Histogram Stretch, Histogram stretch, Arcsinh Stretch, Histogram direct, Logarithmic stretch, Asinh+Histogram stretch, Square root stretch, Shadow stretch, Highlight stretch, None

Select how to combine star and starless image.

Possible combine operations are:

• Add - Use stars+starless formula in Pixelmath
• Screen - Similar to screen in Photoshop
• Lighten - Similar to lighten in Photoshop

Values: Add, Screen, Lighten

List of predefined color palettes. You can also edit mapping input boxes to create your own mapping.

Dynamic palettes are the same as Foraxx options in Enhancements / Narrowband enhancements section. With Dynamic palettes the script automatically uses non-linear data>.

Mapping for R channel. Use one of the predefined mappings or edit and create your own mapping.

Color palette used to map SII, Ha and OIII to R, G and B

There is a list of predefined mapping that can be used, some examples are below. For more details see the tooltip for palette combo box.

SHO - SII=R, Ha=G, OIII=B (Hubble) HOS - Ha=R, OIII=G, SII=B (CFHT) HOO - Ha=R, OIII=G, OIII=B (if there is SII it is ignored)

Mapping formulas are editable and other palettes can use any combination of channel images.

Special keywords H, S, O, R, G and B are recognized and replaced with corresponding channel image names. Otherwise these formulas are passed directly to the PixelMath process.

Option All runs all narrowband palettes in a batch mode and creates images with names Auto_+palette-name. You can use enhancements options, then also images with name Auto_+palette-name+_enh are created. Images are saved as .xisf files. Use Save batch result files buttons to save them all in a different format. To use All option all HSO filters must be available.

Values: Auto, 0.75*H + 0.25*S

Mapping for G channel. Use one of the predefined mappings or edit and create your own mapping.

Color palette used to map SII, Ha and OIII to R, G and B

There is a list of predefined mapping that can be used, some examples are below. For more details see the tooltip for palette combo box.

SHO - SII=R, Ha=G, OIII=B (Hubble) HOS - Ha=R, OIII=G, SII=B (CFHT) HOO - Ha=R, OIII=G, OIII=B (if there is SII it is ignored)

Mapping formulas are editable and other palettes can use any combination of channel images.

Special keywords H, S, O, R, G and B are recognized and replaced with corresponding channel image names. Otherwise these formulas are passed directly to the PixelMath process.

Option All runs all narrowband palettes in a batch mode and creates images with names Auto_+palette-name. You can use enhancements options, then also images with name Auto_+palette-name+_enh are created. Images are saved as .xisf files. Use Save batch result files buttons to save them all in a different format. To use All option all HSO filters must be available.

Values: Auto, 0.50*S + 0.50*O

Mapping for B channel. Use one of the predefined mappings or edit and create your own mapping.

Color palette used to map SII, Ha and OIII to R, G and B

There is a list of predefined mapping that can be used, some examples are below. For more details see the tooltip for palette combo box.

SHO - SII=R, Ha=G, OIII=B (Hubble) HOS - Ha=R, OIII=G, SII=B (CFHT) HOO - Ha=R, OIII=G, OIII=B (if there is SII it is ignored)

Mapping formulas are editable and other palettes can use any combination of channel images.

Special keywords H, S, O, R, G and B are recognized and replaced with corresponding channel image names. Otherwise these formulas are passed directly to the PixelMath process.

Option All runs all narrowband palettes in a batch mode and creates images with names Auto_+palette-name. You can use enhancements options, then also images with name Auto_+palette-name+_enh are created. Images are saved as .xisf files. Use Save batch result files buttons to save them all in a different format. To use All option all HSO filters must be available.

Values: Auto, 0.30*H + 0.70*O

Mapping of Luminance channel with narrowband data if both RGB and narrowband data are available.

With empty text no mapping is done.

Values: Auto, max(L, H)

Linear fit setting before running PixelMath to combine narrowband channel.

Auto does not use linear if Auto STF or Histogram stretch is used for stretching. Otherwise it uses Min. Min does linear fit using the channel with minimum median value as the reference image.

Max does linear fit using the channel with maximum median value as the reference image.

When a channel is selected it uses that channel as the the reference image.Other selections use linear fit with that channel image as the reference image.

Values: Auto, Min, Max, H, S, O, None

Do narrowband mapping using non-linear data. Before running PixelMath, images are stretched to non-linear state.

Force narrowband mapping using formulas given in Settings / Narrowband processing section.

Use multiple narrowband mappings.

List of narrowband palettes that are processed when the multiple mappings option is used. The list is filled from the palette selection dialog.

A special processing is used to add Ha to the RGB image.

If Ha to RGB mapping is used then a narrowband color palette is not used.

Some good combinations are:

• Continuum Subtract - Bright structure add
• Continuum Subtract - Med subtract add
• Continuum Subtract - ScreenInformation for processing and formulas are collected from multiple sources, including:
• Night Photons website
• VisibleDark YouTube channel

Some useful combinations to try.

A special processing is used to add Ha to the RGB image.

If Ha to RGB mapping is used then a narrowband color palette is not used.

Some good combinations are:

• Continuum Subtract - Bright structure add
• Continuum Subtract - Med subtract add
• Continuum Subtract - ScreenInformation for processing and formulas are collected from multiple sources, including:
• Night Photons website
• VisibleDark YouTube channel

Values: Combine Continuum Subtract, SPCC Continuum Subtract

A special processing is used to add Ha to the RGB image.

If Ha to RGB mapping is used then a narrowband color palette is not used.

Some good combinations are:

• Continuum Subtract - Bright structure add
• Continuum Subtract - Med subtract add
• Continuum Subtract - ScreenInformation for processing and formulas are collected from multiple sources, including:
• Night Photons website
• VisibleDark YouTube channel

Values: Continuum Subtract, Basic

A special processing is used to add Ha to the RGB image.

If Ha to RGB mapping is used then a narrowband color palette is not used.

Some good combinations are:

• Continuum Subtract - Bright structure add
• Continuum Subtract - Med subtract add
• Continuum Subtract - ScreenInformation for processing and formulas are collected from multiple sources, including:
• Night Photons website
• VisibleDark YouTube channel

Values: Stretched, SPCC linear

A special processing is used to add Ha to the RGB image.

If Ha to RGB mapping is used then a narrowband color palette is not used.

Some good combinations are:

• Continuum Subtract - Bright structure add
• Continuum Subtract - Med subtract add
• Continuum Subtract - ScreenInformation for processing and formulas are collected from multiple sources, including:
• Night Photons website
• VisibleDark YouTube channel

Values: Bright structure add, Screen, Med subtract add, Max, Add, None

Do noise reduction on Ha image after mapping.

Boost value for continuum subtracted image using ExponentialTransformation.

Do gradient correction on Ha image before mapping.

Smoothen background which may help with gradient correction. It may sometimes help with extreme cases for example when using ABE.

Select a percentage value below which smoothing is done. Usually values below 50 work best. Possible values are between 0 and 100.

Smoothen background which may help with gradient correction. It may sometimes help with extreme cases for example when using ABE.

Select a percentage value below which smoothing is done. Usually values below 50 work best. Possible values are between 0 and 100.

Remove stars before combining Ha to RGB.

Boost, or multiplication factor, for combining R and Ha.

A bigger value will make the mapping more visible by increasing the amount of Ha.

This value is used with all other methods except Add.

Boost, or multiplication factor, for Ha channel with Add method.

For R channel a value 1-Boost will be used.

Note that Add method really works well only with SPCC color calibration.

A special processing is used for narrowband to (L)RGB image mapping. It is used to enhance (L)RGB channels with narrowband data.

This mapping cannot be used without RGB filters.

This mapping is similar to NBRGBCombination script in PixInsight or as described in Light Vortex Astronomy tutorial Combining LRGB with Narrowband. You can find more details on parameters from those sources.

If narrowband RGB mapping is used then narrowband Color palette is not used.

With narrowband RGB mapping you can choose:

• Which narrowband channels mapped (L)RGB channels to enhance those.
• Boost for (L)RGB channels.
• Bandwidth for each filter.
• Test the mapping with a test button

If there is no Luminance channel available then selections for L channel are ignored.

The following steps are done:

• Enhance narrowband channel using bandwidth values.
• Add narrowband to RGB channel using formula: RGBch + (NBch - med(NBch)) * Boost
• Optionally run linear fit on narrowband channel using RGB channel as a reference.
• For the final mapped channel tame max(narrowband channel, RGB channel)Information for processing and formulas are collected from multiple sources, including:
• Light Vortex Astronomy website
• NBRGBCombination script

Use RGB image for bandwidth mapping instead of separate R, G and B channel images. R channel bandwidth is then used for the RGB image.

Do gradient correction on narrowband image before mapping.

Do linear fit on narrowband image before mapping. Not used with Add option.

Mapping of narrowband channel to L channel. If there is no L channel available then this setting is ignored.

Values: H, S, O, (empty)

Mapping of narrowband channel to R channel. If no mapping is selected then channel is left unchanged.

Values: H, S, O, (empty)

Mapping of narrowband channel to G channel. If no mapping is selected then channel is left unchanged.

Values: H, S, O, (empty)

Mapping of narrowband channel to G channel. If no mapping is selected then channel is left unchanged.

Values: H, S, O, (empty)

Boost, or multiplication factor, for the L channel.

A bigger value will make the mapping more visible.

Boost, or multiplication factor, for the R channel.

A bigger value will make the mapping more visible.

Boost, or multiplication factor, for the G channel.

A bigger value will make the mapping more visible.

Boost, or multiplication factor, for the B channel.

A bigger value will make the mapping more visible.

Bandwidth (nm) for the L filter.

To make changes more visible you can lower the RGB bandwidths to something like 40 or 60.

Bandwidth (nm) for the R filter.

To make changes more visible you can lower the RGB bandwidths to something like 40 or 60.

Bandwidth (nm) for the G filter.

To make changes more visible you can lower the RGB bandwidths to something like 40 or 60.

Bandwidth (nm) for the B filter.

To make changes more visible you can lower the RGB bandwidths to something like 40 or 60.

Bandwidth (nm) for the H filter. Typical values could be 7 nm or 3 nm.

To make changes more visible you can lower the RGB bandwidths to something like 40 or 60.

Bandwidth (nm) for the S filter. Typical values could be 8.5 nm or 3 nm.

To make changes more visible you can lower the RGB bandwidths to something like 40 or 60.

Bandwidth (nm) for the O filter. Typical values could be 8.5 nm or 3 nm.

To make changes more visible you can lower the RGB bandwidths to something like 40 or 60.

Do not use SCNR to remove green cast.

SCNR is automatically skipped when processing narrowband images.

Skipping SCNR can be useful when processing for example comet images.

Choose how to do linear fit of RGB images. Linear fit is done only on RGB channels.

For narrowband images linear fit settings are in the Settings / Narrowband processing section.

• Auto does linear fit using Min RGB.
• Min RGB does linear fit using RGB channels with minimum median value as the reference image.
• Max RGB does linear fit using RGB channels with maximum median value as the reference image.
• Min LRGB does linear fit using LRGB channels with minimum median value as the reference image.
• Max LRGB does linear fit using LRGB channels with maximum median value as the reference image.When a channel is selected it uses that channel as the the reference image for RGB images.When Luminance channel is selected also the luminance channel is used for linear fit. If the luminance channel is not present then the red channel is used as the reference image.

Values: Auto, Min RGB, Max RGB, Min LRGB, Max LRGB, Red, Green, Blue, Luminance, No linear fit

RGB channel linking in image stretching.

Auto option uses the following defaults:

• RGB images using mono camera with separate filters for RGB channels use linked stretching.
• Color OSC/DSLR images use unlinked stretching.
• Narrowband images use unlinked stretching. But if linear fit or color calibration is done with narrowband images, then linked stretching is used.Note that some stretching methods do not support unlinked channels.

Values: Auto, Linked, Unlinked

LRGBCombination lightness setting. A smaller value gives a brighter image. Usually should be left to the default value.

LRGBCombination saturation setting. A smaller value gives a more saturated image. Usually should be left to the default value.

Do linear fit on luminance using RGB as a reference before LRGBCombination process.

Saturation increase in linear state using a mask.

Saturation increase in non-linear state using a mask.

Use chrominance instead of saturation to boost colors.

Automatically save the final image also as a 16-bit TIFF image.

By default the final image is automatically saved only in XISF format.

Automatically save the final image also as a JPG image.

By default the final image is automatically saved only in XISF format.

Quality of the JPG image.

If selected save buttons will save all processed and iconized files and not just final image files.

For mono RGB images, channel images are saved before channel combination.

For mono narrowband images, channel images are saved before RGB mapping.

Save stretched starless channel images. Images are saved as .xisf and .tif files.

These can be used for example for narrowband color mapping in Photoshop. For a great example see Utah Desert Remote Observatories YouTube channel.

For mono RGB images, channel images are saved before channel combination.

For mono narrowband images, channel images are saved before RGB mapping.

For OSC/color images, channel images are extracted and saved before stretching.

When using this options is is often useful to select Gradient correction on channel images in Settings tab, Noise reduction on Channel image in Postprocessing tab and BlurXTerminator option Correct only on channel images in Postprocessing tab.

Run auto contrast on stretched channel images.

If selected output files are not written into subdirectories

Keep integrated images when closing all windows

Keep temporary images created while processing and do not close them. They will have tmp_ prefix.

Keep processed images after every step a PixInsight process is applied to them. They will have process name prefix.

Only images after image integration are kept. Images are not saved to disk.

Use unique file names by adding a timestamp when saving to disk.

Add window prefix to AutoIntegrate.log and AutoContinue.log files.

Generate .xdrz files even if Drizzle integration is not used. It is useful if you want to try Drizzle integration later with Start from ImageIntegration option.

Save cropped image files with _crop postfix.

Stop after image calibration step.

Stopping after calibration could be useful if you for example want to check the quality of calibrated light files and maybe set some filtering rules.

Generate master calibration files (bias, dark, flat dark, flat) from calibration files without requiring light files.

This is useful for remote telescopes where calibration files do not change often. Generated master files can later be reused with the Master files option to speed up light file processing.

Checking this box will enable early PSF signal test. Then light files are filtered for PSF signal values below weight limit before any processing.

Weight limit is set in Preprocessing / Filtering section.

Stop after Debayering step. Later it is possible to continue by selecting Debayered files and choosing None for Debayer.

Run only binning to create smaller files.

Stop after Extract channels step. Later it is possible to continue by selecting extracted files and run a normal mono camera (LRGB/HSO) workflow.

Run only image integration to create integrated channel images or RGB image.

Stopping after integration could be useful is you for example want to set some exclusions areas for DBE or for finding the true background.

Run only channel combination to linear RGB file. No auto stretch or color calibration.

Run only image integration on *_r.xisf files to create automatic cropping info.

Light file list should include all registered *_r.xisf files. The result will be LowRejectionMap_ALL.xisf file that can be used to crop files to common area during AutoContinue.

RRGB image using R as Luminance.

Run in batch mode, continue until all selected files are processed.

In batch mode, just click the Run button and the script will ask for files to process before starting.

Batch mode is intended for processing mosaic panels. When batch mode completes only final image windows are left visible.

Final images are renamed using the subdirectory name. It is recommended that each part of the batch is stored in a separate directory (like for example P1, P2, etc.).

Batch mode works only with calibrated light images.

Run in fast mode where images are downsampled to a smaller size and only 10% or minimum of 3 images are used.

With full processing, all light and possibly calibration images are downsampled.

With AutoContinue, all integrated images (Integration_[LRGBHSO]) are downsampled. For other AutoContinue starting points the fast mode is ignored.

Fast mode can be useful to do a quick overview of processing before committing to full processing.

S uses smaller images, M uses bigger images.

Values: S, M

Divide light files into Number of substacks substacks and stack them separately. Stacked files are named as Stack_num_Integration_RGB.

When this option is enabled only image integration is done and final images are not generated. The idea is to use substacks to generate the final image.

Note that this works only with color (OSC) files but no checks for that are done.

Number of substacks.

Force creation of a monochrome image. All images are treated as Luminance files and stacked together. Quite a few processing steps are skipped with this option.

Create synthetic L image from all light images.

Create synthetic image for any missing image.

Start processing from ImageIntegration. File list should include star aligned files (*_r.xisf).

This option can be useful for testing different processing like Local Normalization or Drizzle (if Generate .xdrz files is selected). This is also useful if there is a need to manually remove bad files after alignment.

This option is also useful when doing comet alignment. Then input files should be comet aligned *_ca.xisf files.

If filter type is not included in the file keywords it cannot be detected from the file name. In that case filter files must be added manually to the file list.

When possible use already processed files. This option can be useful when adding files to an already processed set of files. Only files generated before image integration are reused.

Option works best with a Json setup file that is saved after processing or with Autosave generated AutosaveSetup.json file because then star alignment reference image and possible defect info is saved.

With image calibration it is possible to use previously generated master files by adding already processed master files into calibration file lists. If only one calibration file is present then the script automatically uses it as a master file.

If selected default file select pattern is all files (*.*) and not image files.

If selected process icons will be created for all processing steps.

Select which windows are left on the desktop at the end of processing.

Default, info windows and intermediate images are iconized and left on the desktop.

Close info windows, close info windows that are only helper images and not part of the actual processing result:

• LowRejectionMap_ALL, contains the cropping information
• AutoBackgroundModel, contains the automatically detected background area as a previewKeep only final images, close info windows and all intermediate images and leave only the final images on the desktop. Intermediate images include for example integrated channel images, processed channel images and masks.

Note that info windows and intermediate images are not always generated.

Note also that closed images cannot be reused in AutoContinue. If images are saved on disk they can be opened again before running AutoContinue.

Values: Default, Close info windows, Keep only final images

Reset parameters to default values before loading a setup.

This ensures that only parameters from the setup file are set and user saved default parameters are not set.

Uncheck this option if you want just add parameters from the setup file to the current parameters.

Get the full flowchart data before processing when a normal processing is done using the Run button. This makes it possible to follow progress in the complete flowchart if Show Flowchart is checked.

If this option is not checked or AutoContinue or batch processing is done, flowchart data is generated during processing.

Note that this parameter is not reset or saved to Json file.

If checked then the current processed image is shown in the flowchart background.

Note that this parameter is not reset or saved to Json file.

If checked then the operation processing time is shown in the flowchart.

Note that this parameter is not reset or saved to Json file.

If checked then the flowchart image is saved into AutoProcessed directory after processing is complete.

Note that this parameter is not reset or saved to Json file.

Save setup after successful processing into AutosaveSetup.json file. Autosave is done only after the Run command, it is not done after the AutoContinue command.

File is saved to the lights file directory, or to the user given output directory.

Setup can be later loaded into AutoIntegrate to see the settings or run the setup again possibly with different options.

Open directory dialog instead of files dialog.

All files that match the file pattern on the right will be added as image files. Files are searched recursively from the selected directory and all subdirectories.

Selected directory is used as the default output directory.

File pattern can have multiple file types separated by space.

File pattern for files that will be added as image files.

File pattern can have multiple file types separated by space.

Disable blinking of files.

Switch between flowchart and image view if flowchart is available.

Can be checked during processing. In that case live updates to the flowchart are shown.

If Flowchart setting Get flowchart data before processing is checked then the live flowchart view uses full processing flowchart.

Note that this parameter is not reset or saved to Json file.

When checked, a preview image during the processing is always shown in a stretched (non-linear) format. Image name on top of the preview window has text AutoSTF when image is stretched for preview.

When unchecked preview image is shown in original format.

Stretched format can be useful for visualizing the current processed image.

Use Resample process to reduce the size of preview images. The target image size can be changed in the Interface tab.

Resample will make it faster to blink through images. Only the preview image is resample, original image is not modified.

Note that resample may alter how preview image and histogram are shown during the preview.

Target size for preview image resample.

Note that resample may alter how preview image and histogram are shown during the preview.

Start the script with empty window prefix

Enable manual control of icon columns. Useful for example when using multiple Workspaces.

When this option is enabled the control for icon column is in the Interface / Interface settings section.

This setting is effective only after restart of the script.

Run only SubframeSelector to output image weight information and outlier filtering into AutoIntegrate.log AutoWeights.json. Json file can be loaded as input file list.

With this option no output image files are written.

Print some additional debug information to the log output files.

Do not perform any actual processing. This is useful for testing and debugging.

Print some additional debug information when generating flowchart.

Print PixInsight process values to the console and to the AutoIntegrate log file.

Create a JavaScript file with executed processes. It is a runnable script that can be executed in PixInsight.

Test Ha RGB mapping. This requires that you have opened:

• RGB files Integration_[channel], channel = R, G and B.
• Integration_H that is used in the mapping.

To get required Integration_[R|G|B] and Integration_H files you can run a full workflow first.

Values: Mapping, Continuum, All mappings

Astrobin filter configuration for L filter.

Give Astrobin filter number to be used in AstrobinInfo.csv file.

Astrobin filter configuration for R filter.

Give Astrobin filter number to be used in AstrobinInfo.csv file.

Astrobin filter configuration for G filter.

Give Astrobin filter number to be used in AstrobinInfo.csv file.

Astrobin filter configuration for B filter.

Give Astrobin filter number to be used in AstrobinInfo.csv file.

Astrobin filter configuration for H filter.

Give Astrobin filter number to be used in AstrobinInfo.csv file.

Astrobin filter configuration for S filter.

Give Astrobin filter number to be used in AstrobinInfo.csv file.

Astrobin filter configuration for O filter.

Give Astrobin filter number to be used in AstrobinInfo.csv file.

Astrobin filter configuration for C filter.

Give Astrobin filter number to be used in AstrobinInfo.csv file.