QPOE Files

The x-ray data are stored in QPOE files (Quick Position-Ordered Events, *.qp) rather than image arrays. These are lists of photons identified by several quantities, including the position on the detector, pulse height, and arrival time. Note that, unlike IRAF images, QPOE files have no associated header file, and are always stored in the current directory, unless explicitly specified otherwise. Non-PROS IRAF tasks can also access QPOE data files in place of image arrays.

A ROSAT QPOE file contains all the basic information for an observation (i.e., WCS Pointing Information; START/END information, standard processing results, live time, and instrument parameters). In addition, the temporal status (i.e., viewing geometry, background rate, aspect error, SAA...) is given at regular intervals.

The size of a QPOE file is determined by the number of events and the number of photon qualifiers retained. The recognized names (and types) of the elements of the event structure are:

name:           type:           description:
x               short           x position in pixels (on the sky)
y               short           y position in pixels (on the sky)
pha             short           pulse height amplitude in channels
pi              short           pulse invariant bin in channels
time            double          photon arrival time in seconds since
                                   the header reference date
detx            short           detector x position (with instrument corrections)
dety            short           detector y position (with instrument corrections)
rawx/dx         short           raw detector x position (RDF/rev0)
rawy/dy         short           raw detector y position (RDF/rev0)
region          short           region number (defined after a qpcopy
                                    with several regions)

The following aliases are recognized by PROS tasks that create QPOE files to obtain files of different sizes (qpcopy, qpsort, etc.):

image or peewee             x,y
energy or small             x,y,pha,pi
time or medium              x,y,pha,pi,time
detector or large           x,y,pha,pi,time,detx,dety
region                      x,y,pha,pi,time,detx,dety,region
full                        x,y,pha,pi,time,rawx,rawy,detx,dety,status

PROS tasks often call for these aliases in the `EVENT DEFINITION' parameter.

As indicated by the aliases, different scientific objectives require different information, and consequently different size QPOE files.

All IRAF tasks which take IRAF image data as input will also accept *.qp files. However, when *.qp and *.imh files are both present with the same root name, IRAF will assume the *.imh file as the default, thus the .qp extension must be given to access the QPOE file.

Tasks which input and output an image file will by default assume the same format as the input file. Since non-PROS IRAF tasks cannot write *.qp files, *.imh extensions must be explicitly added to the output file name when the input is a *.qp file.

If the user does not wish to use all of the photon events in a QPOE file, a filter may be applied to the data. The filter is appended to the *.qp filename in brackets.

Filter syntax is as follows:

The event structure element is named, followed by an ``='' sign, followed by either the specific value or the range limits separated by a colon. For example:

rp110590.qp[pi=2:30]

refers to all the events in rp110590.qp with pi energy values between 2 and 30, inclusive.

The following event structure elements may be filtered: x, y, pha, pi, detx, dety, rawx (dx), rawy (dy), region. More than one event structure elements may be specified in a filter, if each is separated by a comma. For example:

rp110590.qp[x=200:300,y=512:612,pi=7:12]

Note that the order is unimportant.

qpcopy and such tasks that create output qpoe files do not allow blocking with such filters. Such an execution will produce the following error:

  xi> qpcopy "xdata$rp110590.qp[bl=30,pi=5:24]" none /pool/copy large

QPEX warning: bad attribute name `bl'

  xi>

For more information see filters and qpoe files, or type ``help qpoe'' in IRAF.

xdisplay allows both blocking and filtering with the following syntax, for example:

  xi> xdisp "xdata$rp110590.qp[bl=30][227:237,307:317]"

Note this ``filter'' is actually a section, where the coordinates are expressed in `logical' coordinates, those of the image after blocking.

For more information see physical, logical, and world coordinates.

Note, again, that the order doesn't matter.

 xp> xdisp "xdata$rp110590.qp[227:237,307:317][bl=30]"

will also work.

Other filters may be added as well

  xp> xdisp "xdata$rp110590.qp[227:237,307:317][bl=30,pi=7:24]"

Sections are placed in separate square brackets.

A word about blocking:

Blocking factors allow one to compress qpoe files so that application programs ``see'' an image of smaller dimension (blocking is NOT allowed when inputting QPOE files to a single event access task). The compression is performed by pixel summing so that, for a block=16, for example, $16\times16$ pixels values are summed to one pixel. This would serve to compress, in our example, a $4096\times 4096$ image to a $256\times 256$ image.

The ROSAT PSPC qp file images are $15360\times 15360$. ROSAT HRI are $8192\times 8192$. As many display in a standard ( $512\times 512$) SAOimage window, some blocking expedites the process. A blocking factor of 30 would then compress a PSPC image to $512\times 512$, while HRI's are compressed to $512\times 512$ with bl=16.

Pixel size for both HRI and PSPC ROSAT images is 0.5 arcsecs on a side. If you create new images with blocking factors, the CDELT1 and CDELT2 variables of the image header will be updated to contain the present pixel size (in decimal degrees) of the array.

To view your present image pixel size, you might use a command such as

  xr> imhead xdata$rp110590.qp long+ | match CDELT

CDELT1 = -1.3888890000000E-4

CDELT2 = 1.3888890000000E-4

  xr>