Last modified: 2 July 2019

URL: https://cxc.cfa.harvard.edu/csc/data_products/stack/bkgimg3.html

Stacked Observation Detections Background Image File (bkgimg3.fits)

The stacked observation detections background FITS format image files consist of a background counts image for each set of stacked observation detections and science energy bands. ACIS stacks are blocked by 1 and HRC stacks are blocked by 2 in SKY coordinates. The images are recorded in units of counts.

Stacked ACIS Background Map

[Thumbnail image: stacked ACIS background map]

[Version: full-size]

[Print media version: stacked ACIS background map]

Stacked ACIS Background Map

An example of a stack of ACIS background maps for a given set of observations.

The stacked observation detections background image FITS file is named: 〈i〉〈s〉〈stkpos〉_〈stkver〉N〈v〉_〈b〉_bkgimg3.fits

where 〈i〉 is the instrument designation; 〈s〉 is the data source; 〈stkpos〉 is the position component of the stack name, formatted as "Jhhmmsss{p|m}ddmmss"; 〈stkver〉 is the 3-digit version component of the stack name, formatted with leading zeros; 〈v〉 is the data product version number; and 〈b〉 is the energy band designation.

[CAUTION]
Caution

In many L3 image files, "NaN" values are used to represent data that is outside the field of view (e.g., in the full-field exposure-corrected images where [0 counts / 0 exposure]).

Users should be cautioned that performing seemingly simple operations with the affected files may cause programs to run very, very slow, e.g. using DS9's built in 'smoothing' operation or the tool dmimgcalc. The reason is that whenever the NaN values are encountered, a SIGFPE is generated and has to be dealt with; most applications choose to ignore SIGFPEs, but mechanically, they are still processed and use up significant CPU time.

Furthermore, performing any FFT-based operation (e.g., csmooth, wavdetect, aconvolve, apowerspectrum, etc.) will (a) take a very long time, and (b) will yield an output that is all NaNs. Since every pixel contributes to every frequency in Fourier space, and any operation with a NaN yields another NaN, all the output pixels are NaN.

CIAO users can replace NaN values with zeros (or any other value) accordingly:

ciao% dmimgthresh in.fits out.fits cut=INDEF value=0

Related Science Documents