STScI Logo

Hubble Space Telescope
DrizzlePac: Introduction

AstroDrizzle, a new task that replaces MultiDrizzle, is the flagship software for a new STSDAS package called DrizzlePac. The name "AstroDrizzle" is derived from Astrometric Drizzle to accentuate its new design, built from the ground-up, for improved handling of astrometry and geometric distortion.

Images processed by AstroDrizzle are essentially identical to those from MultiDrizzle. In some cases, AstroDrizzle-combined images are an improvement due to fixed MultiDrizzle bugs. Its most notable enhancements lie in the handling of astrometric information that is contained in images in the form of new header keywords, and for some instruments, new FITS extensions.

Distortion information is provided as linear distortion corrections and higher order polynomial distortion corrections, incorporated directly into the WCS of calibrated image headers using the Simple Image Polynomial (SIP) convention. Non- polynomial distortion corrections, required only for some instruments, are stored in look-up tables as FITS extensions. Therefore, references files are no longer needed for drizzling images, as was the case for MultiDrizzle. Multiple astrometric solutions, each uniquely-named, can now be stored in images. An astrometric solution can be extracted to a small FITS file called a "headerlet" for easy dispatch to collaborators working on the same image--they will be able to apply the headerlet to their copy of the image to use the same astrometric solution.

Other enhancements include the capability to save task parameters to "configuration files" that can be loaded back into AstroDrizzle for later use with the same or similar datasets. Users also have the capability to use several processors in parallel to speed up processing time when running AstroDrizzle.

drizzlepac is a suite of supporting tasks for astrodrizzle which include tweakreg and tweakback for aligning images in different visits; pixtopix for transforming a X,Y pixel position to its pixel position after distortion corrections; skytopix and pixtosky for transformations between X,Y pixel positions and sky coordinates.

TweakReg is a powerful new DrizzlePac task for aligning dithered images using point sources. It replaces TweakShifts in the Dither package. While HST images in a single visit are generally well-aligned, offsets as large as 0.5" can occur for the same pointing taken over different visits. TweakReg aligns calibrated (but still geometrically distorted) images by matching sources between an image and reference image or catalog. Source files for each image can be created in TweakReg using a routine similar to DAOFIND, or an existing catalog can be supplied by the user. For images heavy with cosmic rays, cosmic ray- cleaned images can often be generated by AstroDrizzle for use in TweakReg for improved source identification. Sources in an image are matched to those in the reference image or catalog using methods similar to the IRAF task xyxymatch. A fit is computed for the offset between each image and its reference. TweakReg may need to be run several times using different parameter values for source identification and statistical rejections until enough good sources can be measured to produce good offset fits. When the user is satisfied with the results, TweakReg is run one last time with the selected task parameter values to update the WCS information in the image headers.

TEAL (Task Editor and Launcher) is a new graphical user interface for displaying DrizzlePac task parameters. It is a particularly useful tool for new DrizzlePac users because dependencies between parameters are built into the code to help prevent user errors. From the TEAL interface, task parameters can be saved to uniquely-named "configuration files." These ASCII files can be loaded into the task for later processing, edited and renamed, and shared with collaborators.

Additional information is available in the DrizzlePac Handbook. The DrizzlePac website provides latest information about the software, supplemental materials not included in the handbook, and examples of how to use DrizzlePac tasks. Questions and comments should be sent to