Previous Page Contents Next Page


Reproduced here are the first few lines of the output file jlp-0145.1c, the one we used as our first example of a three channel run:

# Reduced chan 1 data from jlp-0145.op1 on pg1159, start  2:33:40
# integtime  5.000  cycltime   5.000 mean    25971 npts   3127
# bjed 2449078.6130668  baryc 489.7908  deltabc -1.3599 dtt 59.184
     2.500     25813
     7.500     26942
    12.500     23651
    17.500     25380
    22.500     23612

The notation is mostly obvious, but note that the integration time on line 2 is the same as the cycle time; these will be different if more than one filter is used. The cycletime is just the time it takes to get back to the same filter again, after cycling through the others. Qed writes separate output files for each filter, with filename extensions like .1c, .1v, .1b, .1u for data from channel 1; the times shown for each data point are the centroid of the integration. The Quilt programs write down the time at the start of the first integration. Since there was no filter motor operation on this run, and the integration time was 5 seconds, the times are all multiples of 2.5 seconds.

All of our ground-based observations are tied to a moving platform -- the earth -- and the distance (and hence the light travel time) to our target objects changes with time. We can correct for earth motions by calculating the amount of the change during a run, and Qed does that. The values shown on the third header line are the Barycentric Julian Ephemeris Date, which represents the calculated beginning time of the run, for later use when we put runs together into a single light curve. The correction (in seconds) to the barycenter of the solar system (baryc) is shown next, followed by "deltabc", the amount of change in the barycentric correction that took place between the beginning of this run and the end of it. The final entry in the line is the difference (in seconds) between the Universal Coordinated Time, by which we set our observatory clocks, and the Terrestrial Dynamic Time, which we must use if we want all of our observations to have the same uniform time base, with leapseconds taken into account. This value is taken directly from qed.tbl for the date of the run, and is repeated here for the convenience of later programs which may not have qed.tbl available. The run start time in standard UTC time on the first line is there for the same reason.


When Qed starts up it looks immediately for the qed.tbl file, and gets quite upset if it can't find a copy. This is just a text file containing essential information about the leap second situation (the first part of the table), the names and locations of different observatories, along with their default extinction coefficients (the next part), the names and positions of the target stars (the third part), and the names and default values for the photometers.

Here's a condensed version of the table, showing examples of entries in each of the four sections: (lines starting with '#' are comments)

# These time increments must be added to universal time (UTC) in order
# to get TDT (Terrestrial Dynamic Time). 
# Leap second increments were added at 0hrs UT on the given date.
# The Astronomical Almanac, pages B4-5 and K9 has additional information.

 1 Jan 1900  32.184
 1 Jan 1972  42.184
 1 Jul 1972  43.184
   --  --  -- -- --
 1 Jul 1993  60.184
 1 Jul 1994  61.184
 1 Jan 1996  62.184
 1 Jul 1997  63.184
 1 Jan 1999  64.184

# observatory longtiudes, latitudes, and extinction values (ch1, ch2)
-104 01 18   30 40 18    0.38  0.28
20 48 42  -32 22 42      0.360  0.250
149 00 30  -35 19 12
149 04 12  -31 16 36     0.350  0.250
-70 48 54  -30 09 54     0.45   0.27

    --    --    --    --    --

# star positions: ra, dec, epoch (if not 2000)
 23 28 48    5 15 30
 12 38 52  -49 49 30
 09 24 17   35 16 48
 16 47 18   32 28 24
 12 01 46   -03 45 36
 18 06 05.3 45 51 02 1950

    --    --    --     --

# Photometer names and reduction parameters
# The first 3 values are dead times in ns
# The next two values are the detector sensitivity ratios ch1/ch3, ch2/ch3
15 15 15 1.0000 0.7000
60 60 60 1.0000 0.7500
60 60 60 1.0000 1.0000
r60 60 60 1.0000 1.0000
60 60 60 1.0000 0.7500
15 15 15 1.0000 1.0000
    --    --    --     --

Observers type into the Quilt9 Set mode window the name of the observatory, photometer and target star they are observing, so it is up to Qed to try to identify them. The algorithm that does this is quite simple, and seems to work most of the time. It takes whatever characters the observer typed in, removes everything that is not a number or a letter, makes all letters lowercase, and then compares the first six letters it finds to the entries in the table, or just the characters found if there are fewer than six. Differing entries for the same observatory name, say, can be separated by the '|' symbol and Qed will examine each of them. If it can find something that matches, it grabs the following line and uses what it finds there to establish the longitude and latitude for the observatory, and keeps the extinction coefficients as defaults. If there is a qed.his table to be found, Qed uses coefficients from there in preference to the ones in qed.tbl.

Target stars are identified the same way, by the first 6 characters of their name, so PG 1159-035 and pg1159 are identified as the same star. Alternative names (e.g. g117 and g117-b15A) can both be included, separated by the '|' symbol. Photometers are similarly identified. Both the observatory and the photometer are identified by their first 3 characters in the qed.his file, and show up as identifiers when you query Qed about them with the query command 'qd'. For this run, 'qd' shows

It's easy to add entries to the qed.tbl file, which is why it is kept separate. Qed treats qed.tbl as a read-only file; it never writes to it, so if an entry gets garbled, Qed is not at fault. The same is true for the qed.ini file. Qed does write to qed.his to keep it current, so it could mess up the file; should this happen, just delete the file and Qed will start a new one and keep it up to date. If you should decide to edit the qed.his file, be aware that each line is 40 characters long no matter how it looks, and Qed counts on that.

It's possible to reduce a run without a working qed.ini file; Qed will use its built-in defaults, and write a new one when you exit the program. Any time you get an updated version of Qed, rename or delete your qed.ini file and let the program make a new one for you.

In theory you can reduce a run without a working qed.tbl file too, but it is very tedious and error-prone. The ';' commands let you enter the longitude and latitude for the observatory, the RA, DEC and Epoch for the target star, and the leapsecond offset (dtt) from the keyboard. It might work in an emergency, but don't use it. Make up a new qed.tbl file (use the entries above as an example) and house it next to qed.exe. If you have a new target star or a different observatory, it is far easier to add that to qed.tbl (be safe: make a copy of the old one first) than it is to enter all that stuff from the keyboard.

Previous Page Contents Next Page