The latest release of the STIS spectroscopic ETC allows the user to select the dimension of the extraction box perpendicular to the dispersion for point-source calculations. Values of 3, 5, or 7 pixels are supported for this parameter, with a default value of 7 pixels for all spectroscopic modes. For targets with expected countrates much larger than the background, the largest size should be selected so that more flux is included in the calculation. For faint targets such that the sky or detector background will make a substantial contribution to the countrate at the wavelength of interest, the user should check whether a smaller extraction box size will produce the required S/N in less time. After the data are received, the observer can reduce them with various customized CALSTIS extraction boxes to select the one that best meets the scientific objectives.
Note that while the default pipeline extraction uses a 7 pixel high box for MAMA echelle and CCD modes, for 1st order MAMA modes and the NUV PRISM mode this value is 11 pixels. So for these latter modes, the fraction of the source energy included in the standard pipeline reduction will actually be slightly larger (typically by 5-10%, but in a few cases by as much as 15%), than assumed in the current ETC calculation. Future releases of the ETC will include the ability to simulate these reductions as well, but for the current release it was not practical to include different choices of box sizes for different detectors.
The new options for the point source extraction box size allow a more realistic estimate of the S/N for faint, background-limited targets than did the previous ETC version, in which the parameters were a compromise meant to give approximately correct answers for both very bright and very faint targets. If you are concerned that the revised ETC predicts substantially more time will be needed than you had estimated using the Phase 1 ETC, please forward your concerns to your Contact Scientist or Program Coordinator, or send a message to help@stsci.edu. Note that changes of 10% or less are not normally considered significant.
Note that the new version of the COS ETC also includes a modification to the spectral extraction box that will mostly affect faint, background-limited point sources. The Phase I version of the COS ETC may have significantly overestimated the time required to observe faint objects with certain configurations, especially in the NUV, while the Phase I STIS ETC may have underestimated the required times for comparable modes.
The STIS imaging ETC now also allows the user to choose the size of the region used for extended target S/N calculations.
The new version of the STIS ETC also fixes a number of other minor bugs. There is no longer a spurious warning when STIS E230M and E230H observations exceed the global count rate limit of 30,000 counts/s which is appropriate for 1st order MAMA spectroscopic modes. Instead a warning about the global count rate for these echelle modes is issued only when it exceeds 200,000 counts/s.