ACIS Energy-Dependent Subpixel Event Repositioning (EDSER)

About the EDSER Algorithm

For sources near the optical axis of the telescope, the size of the point spread function is smaller than the size of the ACIS pixels. Li et al. (2004, ApJ, 610, 1204) describe various sub-pixel event-repositioning algorithms that can be used to improve the image quality of ACIS data for such sources.

Their algorithm EDSER (Energy-Dependent Subpixel Event Repositioning) was incorporated into the tool acis_process_events in CIAO 4.3, making it possible to reprocess any dataset to apply the subpixel algorithm. EDSER can be applied to all Chandra observing modes - except for continuous-clocking mode (CC33_FAINT or CC33_GRADED) - and to data on both front-illuminated and back-illuminated CCDs. The EDSER algorithm will also become the default in standard data processing (SDP) in version DS 8.4 (28 June 2011).

Before the use of EDSER, the coordinates of ACIS events were randomized by +/- 0.5 pixel to avoid possible aliasing affects associated with the CCD pixel grids. Information on that correction is available in the pixel randomization why topic.

Note that most users will not notice a difference in the data with the EDSER subpixel resolution applied. The exception is users working with high-resolution (< 1 arcsec) data on-axis. Figures 1 and 2 show examples of optimized image resolution by subpixel repositioning of individual X-ray events.

Fig. 1. Three examples of optimized image resolution by subpixel repositioning of individual X-ray events. X-ray images of planetary nebulae BD +30°3639 (panel a), NGC 7027 (panel b) and NGC 6543 (panel c). The left panels X-ray images are obtained by binning events before removing position randomization and applying subpixel event position corrections ("original" image). The center panels are images obtained by binning events after removing event position randomization ("unrandomized" image). The right panels are images obtained by binning events after removing randomization and applying subpixel event position corrections ("event relocated" image). The comparisons between "original," "unrandomized," and "event relocated" images illustrate the superior spatial resolution afforded by subpixel event repositioning. From Kastner et al. 2002, ApJ, 581, 1225.

Fig. 2. A comparison between ACIS images of the inner 3"-radius circum-nuclear region in the Seyfert 1 galaxy NGC 4151 before and after subpixel repositioning. (a) Raw 0.3-1 keV ACIS image; (b) Same image after EDSER algorithm (Li et al. 2004) and subpixel binning (1/8 native pixel), demonstrating the improved resolution. Note the 1 arcsec scale. From Wang, et al. 2011, ApJ, 728, 1.

Caveat: PSF Calibration

The CXC believes that the sub-pixel event-resolution algorithm does improve the ability to separate small scale structure, but we have not yet calibrated ACIS on sub-pixel scales, so obviously the potential for artifacts exists. As reported last year to the Chandra Users' Committee (Report, PDF), we have a plan to enhance our simulation tools to represent the EDSER-enhanced PSF. This is a significant calibration, modelling, and development effort since we need to model the grade distribution and have a much more detailed internal model of the instrument. Refer to the Chandra Ray Trace caveats for some other limitations on the PSF modelling tools.

In principle, sub-pixel analysis should not affect the HRMA ray tracing. However, at the sub-arcsecond scale, HRMA calibration is still on-going (e.g. Probing higher resolution: an asymmetry in the Chandra PSF). Therefore, there may be uncalibrated mirror-related effects at the subarcsecond scale in addition to the unmodelled ACIS detector pixel effects.

Applying & Removing the Adjustment

To use the algorithm EDSER, run acis_process_events with the parameter pix_adj=EDSER (the default value). This may also be done by running the chandra_repro script with pix_adj=EDSER.

The size of the CHIPX and/or CHIPY coordinate adjustment depends on the flight grade and energy of an event. Although the chip coordinates CHIPX and CHIPY remain unchanged in the output event data file, the detector coordinates DETX and DETY and the sky coordinates X and Y are adjusted.

The ACIS Sub-Pixel Event Resolution analysis guide provides guidance on how to use the data in analysis.

To remove sub-pixel coordinate adjustments, run acis_process_events with the parameter pix_adj=NONE or RANDOMIZE.

• If pix_adj=NONE, then no pixel adjustment is applied. The detector and sky coordinates are computed using the unadjusted chip coordinates.

• If pix_adj=RANDOMIZE, then a pixel randomization is applied to the chip coordinates before the detector and sky coordinates are computed.

Using a level=2 event file as input

Since application of a sub-pixel event-repositioning algorithm does not affect any of the status, grade, or time filters, it may be run on level=2 imaging data. If you are analyzing grating data, you must run this thread on the level=1 event file, then complete the appropriate "Obtain Grating Spectra" thread to create an new level=2 event file.

In this case, several additional parameters are set so that acis_process_events does not recompute the values of FLTGRADE, GRADE, and PHA (doevtgrade=no) or the values of ENERGY and PI (calculate_pi=no). A custom "level 2" eventdef is specified. The reason for this is explained in the "Why is my level 2 event file larger than the one from the pipeline?" FAQ.

unix% punlearn acis_process_events
unix% acis_process_events infile=acis_evt2.fits outfile=acis_new_evt2.fits \
      acaofffile=pcad_asol1.fits doevtgrade=no calculate_pi=no \
      pix_adj=EDSER \
      eventdef='{d:time,s:ccd_id,s:node_id,i:expno,s:chip,s:tdet,f:det,f:sky,l:pha,l:pha_ro,f:energy,l:pi,s:fltgrade,s:grade,x:status}'

# acis_process_events (CIAO 4.4): WARNING: Cannot load MTL file 'NONE', will not apply FPTEMP dependent cti corrections
# acis_process_events (CIAO 4.4): WARNING: The CTI adjustment to PHA will not be (re)computed or removed because doevtgrade=no.
# acis_process_events (CIAO 4.4): WARNING: The time-dependent gain adjustment to PHA will not be (re)computed or removed because doevtgrade=no.
# acis_process_events (CIAO 4.4): WARNING: The values of PHA, ENERGY, PI, FLTGRADE and GRADE may be inaccurate because doevtgrade=no.
# acis_process_events (CIAO 4.4): WARNING: The values of ENERGY and PI may be inaccurate because calculate_pi=no.

The warnings are expected; they are triggered because mtlfile=NONE (by default), doevtgrade=no, and calculate_pi=no.

Additional Resources

• ACIS Sub-Pixel Event Resolution analysis guide: guidance on how to use the data in analysis.
• Chandra ACIS Subpixel Resolution: AAS poster by Kim, et al. (2011)
• The specification for implementing subpixel event-positioning in acis_process_events.