Conditions:
- code version: preliminary release, 05/06/2004
- input data: H1:LSC-AS_Q channel, S3 run, GPS start time:
751974320 sec
- data length: 4096 sec
- waveform #0: Sine-Gaussian 235 Hz, Q=9 (duration=0.0085 s)
- time stride: 20 or 60 sec
- frequency resolution: 64 Hz
- time resolution: 1/128 sec
- number of simultaneous injections per time stride per waveform:
10
- number of repetitions per stride: 10
- number of different injection amplitudes for efficiency
meausurements per time stride: 5
- total number of injections per waveform per time stride:
10x10x5=500
It is possible to run BurstMon with different
time strides. Because output goes to 1 minute trend the time stride
greater then 60 sec will lead to holes in output trend. The shorter
time stride (20 sec in this test) may allow to get more details, see
blue line on top picture. As one can see the BurstMon gives on everage
a little higher values of
h50 when running with 20 sec stride
compared with stride 60 sec. The variations of the
h50
are noticeably larger for 20 sec time stride. The relative errors of
h50
which depend on fit quality are usually larger for time stride 20 sec,
as shown on second picture. Actually with the time stride less then 60
seconds the 1 minute trend will present averaged output of BurstMon so
that the variations of the
h50 will be smoothed down and details
will be lost.
