FOCAS contains 8 grisms providing low and moderate spectral resolution. Each grism can be used with slits of 0.2, 0.4, 0.6, 0.8, 1.0, or 2.0 arcseconds in width, and any of the four order sorting filters. The following table lists the 5 sigma per pixel sensitivities for 1 hour of integration, split into 3 exposures of 20 minutes each.
Note for VPH grisms. Observers should remember the image shift problem as explained at the end of this page. Two VPH grisms (850nm, 900nm) are available from S10A.
Note for the Echelle grism. As for the Echelle grism, only a single-order mode with an order-sorting filter (i.e., without a cross disperser) is available, although this grism was originally designed for the Echelle spectroscopy.
Important notes about this table: These numbers are the sensitivities (in magnitudes relative to Vega) for a point source in 0.5'' seeing, observed with a 0.4'' slit and extracted over a 1.0'' (10 pixel) aperture. A wider slit will transmit up to 60% more flux, with a gain of up to half a magnitude at the expense of spectral resolution.
Dispersion direction: Please note that the dispersion direction is vertical on the CCD image. A negative value of dispersion in the following table means that wavelength decreses from bottom to top (CDELT2<0.0 in the original CCD image). Only the Echelle grism has a positive dispersion.
For extended sources, the limiting surface brightness sensitivities per pixel (in magnitudes per square arcsecond) are approximately 1.7 magnitudes brighter than the numbers in the table.
Most observations of faint sources with FOCAS are not clearly in one of the background-limited, detector-limited, and object-limited regimes. Simple scaling from the numbers in these tables (especially to brighter sources, when one moves into the object-limited regime) does not always work. Typically, the magnitudes listed produce ~100e per pixel per hour. If your source is much brighter than this, the noise will be dominated by the object; e.g., an object with R=19.4 observed with the 75l/mm grating (100 times brighter than the 5 sigma limit) will produce 10,000e in an hour, and the S/N will be ~100, not 500.
The signal-to-noise ratio can also be improved via on-chip binning; for example, 2-pixel binning in the wavelength direction is optimum for the 0.4'' slit.
Please check the grism selection guide for more details about observational parameters.
(*) This (4000A) is the wavelength at the maximum efficiency. It is noted that the center wavelength is 4500A for 600_450nm grism.
(**) Important note for image shift in VPH grisms: The VPH grisms produce image shifts along the slit direction (i.e., orthogonal to the dispersion direction). The values listed above are in pixels (~ 0.1 arcseconds) where positive values denote that the shift direction is from CCD1 to CCD2.
Choosing the correct order sorting filter: The L550 order sorting filter blocks the long wavelength side better than L600; for observations of 2nd order light, L550 is recommended to avoid 1st order contamination.
The negative value of the dispersion for the Echelle grism denotes that the dispersion direction of the Echelle grism is opposite to those of other grisms.
The SY47 order sorting filter admits some 2nd order light longward of 9100A; use the SO58 filter if you wish to observe to the CCD red cutoff.
In 1st order, 300R is more sensitive than 300B longward of 7300A.