Reprocessing Data to Create a New Level=2 Event File

Imaging Observations

There are two filtering steps for ACIS imaging observations:

1. Filter for bad grades and for a "clean" status column (i.e. all bits set to 0):

   unix% punlearn dmcopy
   unix% dmcopy "acis_new_evt1.fits[EVENTS][grade=0,2,3,4,6,status=0]" \
         acis_flt_evt1.fits
   

2. The Good Time Intervals (GTIs) supplied by the pipeline now need to be applied; these are stored in the acis_flt1.fits file. Simultaneously, unnecessary columns are eliminated from the output:

   unix% punlearn dmcopy
   unix% dmcopy "acis_flt_evt1.fits[EVENTS][@acis_flt1.fits][cols -phas]" \
         acis_evt2.fits
   

   Be sure to include the @ symbol in the filter expression; the command will not be executed properly if it is omitted.

Finally, there is the option to run destreak, which can improve the quality of your data. For details on how the tool works, see the Destreaking ACIS Data why topic and "ahelp destreak".

The ccd_id parameter should always be set to 8 (the default) to ensure that the destreak algorithm will be applied to the ACIS-S4 chip only:

unix% punlearn destreak
unix% pset destreak infile=acis_evt2.fits
unix% pset destreak outfile=acis_dstrk_evt2.fits
unix% destreak
Input dataset/block specification (acis_evt2.fits): 
Output dataset/block specification (acis_dstrk_evt2.fits): 

The thread is now complete. acis_dstrk_evt2.fits is a level=2 event file with the latest calibration products applied.

Grating Observations

There are three CIAO threads that deal specifically with generating a level=2 event file for an ACIS grating observation. Follow the appropriate one, depending on which transmission grating was used:

Single source:

• HETG/ACIS-S Grating Spectra
• LETG/ACIS-S Grating Spectra

Multiple sources:

• Grating Spectra for Multiple Sources - ACIS (applies to both HETG and LETG)

Once you leave this thread, you do not need to return to complete any further steps.

Common acis_process_events Warnings

It is important to remember that the following messages are warnings, not errors. The tool may print one of these messages, but will still create a valid event file.

Event island contains 1 or more bad pixels.

# acis_process_events (CIAO 4.0): The following error occurred 26941 times:
  dsAFEBADPCNTERR -- WARNING: Event island contains 1 or more bad pixels.

This warning indicates that one or more of the nine pixels in a 3x3 event island is identified as a bad pixel. The 26,941 events for which this is the case (in this example) have a STATUS bit set to one. This message is produced whenever a bad pixel file is used and may be safely ignored.

No files found matching CALDB search.

# acis_process_events (CIAO 4.0): WARNING: No files found matching CALDB search:
        tel=CHANDRA
        inst=ACIS
        det=-
        filt=-
        product=CTI
        start_date=1999-08-20T00:07:44
        start_time=00:07:44
        stop_date=1999-08-20T02:26:38
        stop_time=02:26:38
        query=-
# acis_process_events (CIAO 4.0): WARNING: PROBLEM READING CTIFILE, CTI ADJUSTMENT WILL NOT BE APPLIED.

This warning may also be triggered by product=T_GAIN.

The most common cause for seeing this warning is that the observation does not contain -120C data. Since the CTI and TGAIN is only calibrated for this focal place temperature, it is not possible to use it on other observations.

Since no file is found, acis_process_events continues running as if apply_cti=no or apply_tgain=no and produces a valid output file that has not been CTI-corrected or had the TGAIN applied.

The CTI correction cannot be performed on a graded mode event list.

# acis_process_events (CIAO 4.0): WARNING: The CTI correction cannot be performed on a graded mode event list.
# acis_process_events (CIAO 4.0): WARNING: Although the keyword TGAINCOR is false or missing, the column PHA_RO exists. 
                                 The values in this column are assumed to be the unadjusted pulse height values.

The CTI correction requires that a PHAS column exist in the input file. This column is not created for observations taken with the data mode GRADED, so this warning is produced.

In this case, acis_process_events continues to reprocess the data as if apply_cti=no and produces a valid output file that has not been CTI-corrected.

HRC-S Imaging Observations

There are two filtering steps for HRC-S imaging observations:

1. Apply the status filter that is specific to HRC-S observations; a value of 0 demands that the bit be flagged as "good", a value of x indicates that either status (0/1) is acceptable. The pha filter removes about 25% of the background with no X-ray losses; see this FAQ:

   unix% punlearn dmcopy
   unix% dmcopy \
         "hrc_new_evt1.fits[pha=0:254,status=xxxxxx00xxxx0xxx0000x000x00000xx]" \
         hrc_flt_evt1.fits
   

2. The Good Time Intervals (GTIs) supplied by the pipeline now need to be applied; these are stored in the hrc_std_flt1.fits file. Simultaneously, unnecessary columns are eliminated from the output:

   unix% punlearn dmcopy
   unix% dmcopy \
         "hrc_flt_evt1.fits[EVENTS][@hrc_std_flt1.fits][cols -crsu,-crsv,-amp_sf,-av1,-av2,-av3,-au1,-au2,-au3,-raw,-sumamps]" \
         hrc_evt2.fits
   

   Be sure to include the @ symbol in the filter expression; the command will not be executed properly if it is omitted.

The thread is now complete; hrc_evt2.fits is a level=2 event file with the latest calibration products applied.

Next, read the Computing Average HRC Dead Time Corrections thread to see if you need to recompute the deadtime statistics for your dataset.

HRC-I Imaging Observations

There are two filtering steps for HRC-I imaging observations:

1. Apply the status filter that is specific to HRC-I observations; a value of 0 demands that the bit be flagged as "good", a value of x indicates that either status (0/1) is acceptable:

   unix% punlearn dmcopy
   unix% dmcopy \
         "hrc_new_evt1.fits[status=xxxxxx00xxxx0xxx00000000x0000000]"  \
         hrc_flt_evt1.fits
   

2. The Good Time Intervals (GTIs) supplied by the pipeline now need to be applied; these are stored in the hrc_std_flt1.fits file. Simultaneously, unnecessary columns are eliminated from the output:

   unix% punlearn dmcopy
   unix% dmcopy \
         "hrc_flt_evt1.fits[EVENTS][@hrc_std_flt1.fits][cols -crsu,-crsv,-amp_sf,-av1,-av2,-av3,-au1,-au2,-au3,-raw,-sumamps]" \
         hrc_evt2.fits
   

   Be sure to include the @ symbol in the filter expression; the command will not be executed properly if it is omitted.

The thread is now complete; hrc_evt2.fits is a level=2 event file with the latest calibration products applied.

Next, read the Computing Average HRC Dead Time Corrections thread to see if you need to recompute the deadtime statistics for your dataset.

Grating Observations

There are two CIAO threads that deal specifically with generating a level=2 event file for an HRC grating observation. Follow the appropriate one, depending on which detector was used:

Single source:

• LETG/HRC-S Grating Spectra
• LETG/HRC-I Grating Spectra

Multiple sources:

• Grating Spectra for Multiple Sources - HRC (applies to both HRC-S and HRC-I)

Once you leave this thread, you do not need to return to complete any further steps.

Common hrc_process_events Warnings

It is important to remember that the following messages are warnings, not errors. The tool may print one of these messages, but will still create a valid event file.

Out of sequence events discovered in hrc_evt1.fits.

# hrc_process_events (CIAO 4.0): The following error occurred 2482 times:
        dsHPEEVENTSEQERR -- WARNING: Out of sequence events discovered in hrc_evt1.fits.

There are occasional events that appear in the telemetry stream with time tags that make them seem to have occurred out-of-sequence. One special case where this occurs has been documented (see the Out-of-sequence HRC Time Tags memo); other occurrences are most likely to be caused by hiccups in the HRC or by double-bit errors in telemetry (single-bit errors are corrected).

This warning may be ignored in most cases, as long as the number of events flagged as out-of-sequence is a small fraction of the total number of events in the file; see this FAQ for more information.