Skip to the navigation links
Last modified: 17 December 2019


CalDB 4.9.0 Public Release Notes

Public Release Date: 17 DEC 2019

NOTE: CalDB 4.9.0 will not be installed in standard data processing until DS 10.8.3 is released at a TBD date!


CalDB 4.9.0 is an upgrade to the Chandra CalDB, which includes the following items:

For the CIAO 4.12 / CalDB 4.9.0 release notes see How CalDB 4.9.0 Affects Your Analysis.


CalDB 4.9.0 is compatible with CIAO 4.12 and later versions ONLY! Please do not attempt to upgrade your CalDB for CIAO 4.11 or older versions with CalDB version 4.9.0. None of the ACIS upgrades in CalDB 4.9.0 will be applied with the older software because of required software changes to do so. Furthermore, the HRC-S chip edge offset values will not be applied in CIAO 4.11 with CalDB 4.9.0, and users will not be notified by the software if this occurs.

As a result of this, no CalDB 4.9.0 "upgrade" patch tar file is being provided in this release. CalDB 4.9.0 is to be considered the new baseline CalDB for all future upgrades. In addition, there will be no further CalDB upgrades provided for CIAO 4.11 or CIAO 4.10.


A. ACIS BADPIX with Framestore Shadow Regions

Location: $CALDB/data/chandra/acis/badpix/
Filenames: Files applicable to TE-mode observations, with framestore shadow regions added:
For early data taken with ACIS FP_TEMP = -90C

For FP_TEMP = -100C

For FP_TEMP = -110C, years 1999-2000 only

For FP_TEMP = -120C

Files applicable to CC-modes only, with NO framestore shadow regions included:

The framestore shadow regions for each chip are being added to the ACIS bad pixel (BADPIX) CalDB library, at this point for all the non-CC-mode observations, including retroactively to the launch date for Chandra. The shadow results from a construction feature at the bottom of the chip, where the first several rows (CHIPY = 1-~5), but varies slightly from chip to chip. Table 1 below gives the actual region included for each separate CCD of ACIS. These regions are being assigned a separate column in the status bits in the BADPIX files (bit 17) which then required that the "status" column in the BADPIX files be changed from 16 to 32 bits. Events occurring in these regions will normally be filtered out in generating the level 2 events files, at least for the non-CC-mode ACIS observations.

Table 1: Framestore shadow region as determined for each ACIS chip. The bold number for CHIPY in the second column also gives the y-extent or width of the region.

CCD_ID Framestore Shadow Region
0 [(1,1) (1024,5)]
1 [(1,1) (1024,7)]
2 [(1,1) (1024,6)]
3 [(1,1) (1024,6)]
4 [(1,1) (1024,8)]
5 [(1,1) (1024,8)]
6 [(1,1) (1024,8)]
7 [(1,1) (1024,8)]
8 [(1,1) (1024,7)]
9 [(1,1) (1024,7)]

Because the application of bad pixels in the CC-mode is different, we have found that a more involved software change will be necessary to implement the framestore shadow event exclusions. Hence, for the time being, we are not setting up to filter framestore shadow region events in the CC-mode. To this end, we are implementing a separate BADPIX file for each of the BADPIX file cases above, to be applied with CC-mode observations. The CC-mode BADPIX files each have a "B" tag at the end of their filenames, and these files do NOT include the framestore shadow regions in them. This situation will remain until further notice.

Because of the "status" column change, necessary for each active file in the CalDB BAPIX library for ACIS, the entire library is being upgraded, and all of the old BADPIX files are being retired (removed) from the CalDB distribution.


The CIAO 4.12 script chandra_repro may be used to apply this BADPIX file upgrade to non-CC-mode observations. It calls the routine acis_build_badpix which implements the framestore shadow region exclusions.


The "ACIS Data" branch of the thread Reprocessing Data to Create a New Level=2 Event File is applicable in this case.

B. ACIS CTI Files With Added Coefficients (set to 0.0)

Location: $CALDB/data/chandra/acis/cti/
Filenames: Files applicable to non-GRADED datamodes:

Files for GRADED datamodes:

Note: The above CTI file libary change in CalDB has zero effect on any ACIS data sets. The following information is provided for completeness and clarity.

With the advent of warmer and warmer ACIS operating temperatures, the ACIS Calibration Team and CXC Science Data Systems Group have developed a modification to the ACIS CTI correction algorithm. Specifically, a higher-order polynomial fit to the fractional change in the serial or parallel trap density distribution per degree Kelvin is being implemented, which can give improved event corrections as the focal plane temperature of ACIS (FP_TEMP in the events lists) increases above -114C. As a result, there is a need to include additional coefficients in the first extension of each of the active CTI correction CalDB files.

Calibrations of these coefficients are in process; however the software and CalDB file modifications are being implemented in advance, so that a simple future CalDB upgrade can implement the new data. For the time being, the above new files are being put out with place holders for the new coefficients included. In this particular release (the version *N0009* files above) have the value 0.0 assigned to the new coefficients. Hence, the included upgrade will have zero effect on users' data sets when applied.

Note: All of the pre-existing CTI correction data have been included in the above files. Because the CIAO 4.12 software requires the presence of the new coefficient columns in the CalDB files, we are replacing the full ACIS CTI calibrations library in CalDB. All of the old CTI files are being retired from the distribution.

See the CTI why page for basic information on the ACIS CTI correction procedure.


No pipelines or tools are affected by this upgrade, because it is a data neutral/zero effect change at this point.


Again, no threads are affected by this change, because it is data neutral at this point.

C. HRC-S GAPLOOKUP Version N0006

Location: $CALDB/data/chandra/hrc/gaplookup/
Filename: hrcsD1999-07-22gaplookupN0006.fits

The above HRC-S GAPLOOKUP file replaces the version N0005 file released with CalDB 4.8.3 on 23 May 2019 to the public. The reason for this upgrade is that the CXC originally release version N0005 of this GAPLOOOKUP file with two keywords added to each of the three U-direction BINTABLE extensions within the file, but the keyword names were too long for interoperability requirements within the FITS file definition. Specifically, the original keywords were named HRC_OFF_X and HRC_OFF_Y, which are nine characters long. The new file above uses the keyword names HRC_OFFX and HRC_OFFY, which are the required eight characters in length, or interoperability. These keywords contain modified chip edge offset values for each of the three HRC-S (1, 2, and 3) plates.

Because the CIAO software had to be changed to accommodate this CalDB file change at the same time, CIAO 4.12 includes the necessary changes to be able the read the offset values. For this reason, CIAO 4.12 and CalDB 4.9.0 are mutually required for these chip edge offsets to be implemented correctly in hrc_process_events. At the same time, CIAO 4.11 still requires the old nine-character keyword names to continue to implement these offsets. Hence CIAO 4.11 cannot accept the CalDB 4.9.0 upgrade, and no upgrade file or procedure will be provided for CIAO 4.11 and any earlier versions of CIAO.

For technical details regarding the HRC-S chip edge offsets, see the CalDB 4.8.3 release notes under the tech details for the HRC-S GAPLOOKUP version N0005 file here.


The CIAO 4.12 script chandra_repro may be used to apply the HRC-S chip edge offsets to recalibrate the LETGS energy scale. That script calls the tool hrc_process_events which reads the GAPLOOKUP file and implements the offsets.

Note also that the HRC-S chip edge offsets have been implemented in Standard Data Processing (SDP) with the release of DS 10.8.1 and CalDB as of October 2, 2019. All HRC-S datasets processed at the CXC in SDP since that date will have this modification included in them.


The HRC data branch of the thread Reprocessing Data to Create a New Level=2 Event File is applicable here, and only to HRC-S data sets.