The Ohio State University
College of Mathematical & Physical Sciences
Department of Astronomy
The multi-object capability was commissioned in May 2011. Click here for further details and pictures.
Click here for construction photos.
More information is available via these detailed instrument characteristics and these two papers:
Here is an OSMOS logsheet.
There are numerous useful Prospero scripts in the Scripts/ directory on hiltner. Observers are encouraged to look through these scripts. A brief guide to the most useful scripts may be written in the future.
Photometric magnitude zeropoints for OSMOS were measured in April 2010 with the MDM4K and in September 2011 with the R4K. These magnitude zeropoints correspond to 1 e- per second.
|Vega magnitudes, MDM4K|
|AB magnitudes, R4K|
Paul Martini's page on Useful Astronomical Data has information about Flux Zeropoints.
Basic information about wavelength coverage and resolution is available in the Quick Reference section. More details on performance will be posted shortly.
Region of Interest
One may reconfigure either detector to read out a specific 'Region of Interest' (ROI), rather than the full detector. This mode has the advanges of faster readout and smaller image files and may be particularly valuable for judging when to start twilight flats, focus sequences, and spectroscopic observations. Configuration files for 1k square, 2k square, 1x2k, and 1x4k regions centered on the detector are presently available. To implement an ROI for the MDM4K in Prospero type:
Note that the 2x1k ROI is well-suited to single-object spectroscopy with the triple prism and the 4x1k ROI is well-suited to single-object spectroscopy with the VPH grism. (The old 'init' versions of these ROI scripts were replaced with the present set of 'roi' scripts in October 2011.)
OSMOS uses a Prontor/E100 shutter. This is a leaf shutter that may introduce timing errors in short exposures. To attain better than 1 percent measurement precision, one should employ exposures longer than 10 seconds (including for twilight flats). The figure below demonstrates the shutter timing error for a 1s exposure. In this image the peak at the center of the field is approximately 5 percent higher than at the edges.
Observers interested in multi-object masks presently should contact Paul Martini at least two months before their run.
OSMOS masks may be designed with a software package called OSMOS Mask Simulatior (OMS) OMS is a slightly modified version of MMS, which is the mask design software for MODS (MMS is in turn a slightly modified version of LMS, which is the mask design software for LUCI). In addition to the OMS web page, the web pages for MMS and especially LMS contain additional documentation.
Mask alignment is performed with alignment stars. These stars will have square apertures in the mask at their locations relative to the slits. While in principle only two are necessary to solve for the rotation and translation offsets required for mask alignment, at least four are recommended to guarantee a good solution. To speed the mask alignment process, these stars should be in the central 4x1k region of the detector so that this ROI may be used to align the mask. These stars should also be well distributed across the field to maximize the lever arm for the rotation offset calculation.
The outputs of OMS include a file in gerber format (.gbr extension) that contains instructions to the laser cutting machine on how to cut the mask and mask description file (.oms extension) that describes the mapping between targets and the slit mask. Note that it may take up to a month to arrange for the masks to be cut. Observers will consequently need to send their final mask designs to Paul Martini at least a month in advance of their run to guarantee that the masks will be ready. An image of an OSMOS mask in a cell is shown here
The cost per mask is still TBD, but will likely be $150 to $200. This cost covers the electroformed spherical shell and the labor for the laser machine operation.
Multi-object observations are similar to longslit observations, with the exception that the rotator angle must also be set very precisely (to within 0.01 degrees). This can be achieved with custom mask alignment software available at the telescope and is described in the section on Multi-Object Acquisition.
Guidelines for Calibration
The present MIS calibration system does not uniformly illuminate the entire OSMOS FOV. A reasonable solution is to simply take very long exposures during the afternoon until sufficient signal is obtained in all of the slits for both arc lamps and spectroscopic flats. Twilight spectroscopic flats could be used to remove the non-uniformity of the illumination by the MIS flat lamps. See also the section on Flat Fields for more information.
The zeropoint of the wavelength solution will likely drift between the afternoon calibations and the nighttime observations due to instrument flexure. Sky lines could be used to compute this zeropoint offset. See also the section on Wavelength Calibration for further information on specific dispersers.
OSMOS will typically be ready for operation when observers arrive for their first night. The MDM staff will perform the following tasks before a run, but they will be necessary for observers to follow in the event of a power failure or lightning shutdown:
Then type startup in Prospero to load the current instrument configuration. The last step is to place the collimator and camera lens barrels at their nominal position. This is described in the next section.
Finally, make sure that the OSMOS instrument hatch is open. OSMOS has its own instrument hatch that is separate from the MIS hatch. The lever to open and close this hatch is on the opposite side of the instrument from the access doors for the slit and filter wheels.
There is a separate web page with Detailed Startup and Usage Information.
The instrument configuration is controlled with six basic commands:
|pr> go||take an exposure|
|pr> exp T||set the exposure time to T seconds|
|pr> object name||set the object name to name|
|pr> filename name||set the filename to name|
|pr> print wheel||print the current population of wheel [slit, disp, or filter]|
|pr> call script||run the Prospero script named script.pro|
|pr> snap||take an image but do not save it to disk|
|pr> newext num||change the file number to num|
The tedit command may be used to update the tables that map wheel positions to their contents (e.g. that position 6 is open). Please do not edit these tables without good reason! And please also double check these tables (ideally with the help of the observatory staff) if you suspect the tables are incorrect.
Prospero has a powerful scripting capability that can be used to automate many common tasks. Examples include mask alignment (via oalign.pro), telescope focus (via focus4k.pro) automated, guided dither sequences for deep imaging (via four.pro and nine.pro, and changing the ROI (e.g. roi1k.pro, roi4k.pro). These and many other scripts are available in the Scripts/ subdirectory on hiltner. Further details on how to write Prospero scripts are available in the Prospero online documentation. Should the OSMOS-related scripts discussed here become compromised, here is a backup .tgz archive of some scripts.
A quick way to get close to the best focus is to find a reasonably bright star, switch to the 1k ROI, turn on movie mode (type the movie command), and focus by eye until the star appears reasonably in focus. (Type stopmovie to exit movie mode.)
A simple Prospero script called focus4k.pro is available to help refine the telescope focus once it is close. This script takes a sequence of 5 focus frames in steps of 10 units starting with an input, minimum value. The syntax is:
Instructions on how to use the guider are available in these Detailed Startup and Usage Instructions. More details about how to use the guider are available from the guide to Autoguiding and Acquisition at MDM by John Thorstensen.
JSkyCalc24mGS is also used for spectroscopic acquisition, which is described further in the next sections on Longslit Acquisition and Multi-Object Acquisition.
Note that the OSMOS FOV includes nearly the entire unvignetted footprint of the MIS, which means that guide stars need to be near (or ideally in) the partially vignetted beam for imaging (or multi-slit spectroscopy) that employs the full FOV. The box labeled "Probe can block detector inside this region" in the figure at the top of the JSkyCalc24mGS manual is much smaller than the equivalent, circular region for OSMOS.
OSMOS does not have a slit viewer. Instead, targets are placed in a slit with the use of telescope offsets to the known position of the slit projected onto the sky. The object can then be imaged through the slit after acquisition to confirm that it has been acquired.
Guider offsets are much more precise than telescope offsets at the 2.4m and consequently it is much more efficient to offset the guide box the desired amount and then move the telescope to put the guide star back in the guide box. There are two algorithms for longslit acquisition. One uses a pyraf code by John Thorstensen and this is described in the Detailed Startup and Usage Instructions.
Alternatively, one may use the -l option in oalign.py, which is also used for Multi-Object Acquisition. (There is also a Prospero script called olsalign.pro that performs a similar function to oalign.pro.) Usage of these scripts is described in the next section on Multi-Object Acquisition.
Finally, the old long slit acquisition procedure describes how to perform these steps more manually and less efficiently.
Acquisition of the targets in a multi-object slit mask requires very precise translation (on order 0.1" in x and y) and rotation (on order 0.01 degrees) offsets. This requires an image of the slit mask (which should be obtained immediately prior to alignment to minimize flexure) and an image of the field. The relative positions of the alignment boxes and the alignment stars may then be used to calculate the translation and rotation offsets.
There are two scripts that aid in efficient mask alignment: The Prospero script oalign.pro and the python script oalign.py. These scripts are meant to be used in conjunction with one another. oalign.pro is a Prospero script (note the .pro extension) that facilitates the acquisition of the images necessary for mask alignment and the proper configuration of the instrument and oalign.py is a python script that is used to calculate the offsets needed for mask alignment. The rotation offset is sent via the Rotator GUI and the translation offsets are best sent by moving the guider stage. (Please read the Guide to the new MDM Hiltner Rotator for usage instructions for the rotator.) oalign.py must be run in a separate terminal from Prospero.
A step-by-step guide to a mask alignment is provided in this OSMOS Mask Alignment Cookbook.
Filter and slit exchanges
Filter and slit exchanges should only be performed by, or in consultation with, Observatory staff. You should contact them with any special configuration information in advance of your run.
Twilight flats are the most reliable flats to calibrate imaging data. A good strategy for evening twilights is to use movie mode and 1k Region of Interest until the sky is no longer saturated. For example, if you are using the MDM4K type:
Note that exposure times longer than 10s are recommended for flats to avoid significant Shutter Timing errors. Note also that for the standard BVRI set (Harris set), the order of the filters from most sensitive to least sensitive is IRVB.
The MIS flat lamp is useful for the removal of small-scale spatial variations along the slit, although they are not useful for large-scale variations because they do not uniformly illuminate the slit. Spectroscopic twilight flats should provide an adequate measure of large-scale variations along the slit.
The best wavelength calibration source for the triple prism is a bright, compact planetary nebula. None of the MIS lamps offer a good selection of unblended lines over the entire wavelength range accessible with the prism. The best wavelength solution obtained during commissioning employed a 4th-order spline and yielded an rms of about 1nm. Here is a screenshot of the identify task in IRAF with planetary nebula lines labeled and here is a list of the emission lines. Many of these lines are actually blends, particularly in the red, and therefore there is room for improvement.
Medium Blue VPH Grism
In addition to a longslit through the center of the field, two other slits are available that are +/- 30mm (about +/- 5.7 arcmin) from the center slit (but parallel to it). These offset slits allow one to take advantage of one of the unique properties of VPH gratings, namely that the wavelength of peak diffraction efficiency varies as a function of the angle of incidence inside the grating with respect to the fringe plane. These offset slits are referred to as the inner slit and outer slit to differentiate them from the center slit and these names refer to their position in the slit wheel, i.e. the slit closest to the axis of rotation is the inner slit. Here are the wavelength ranges and approximate wavelengths of peak efficiency for these three positions:
Matthias Dietrich has created a series of Wavelength Calibration Figures and Tables for the MIS lamps. Please see this information for further details.
The calibration lamps in the MIS do not fully or uniformly illuminate the entire OSMOS field. This will result in lower illumination at the ends of the slits. There is approximately a factor of two decrease in a long slit lamp at 5 arcminutes from the center of the field. Some points that are farther off axis still appear completely vignetted. Complete vignetting does not occur for the long slit, but may impact multi-slit observations. Below is an image of the Ne lamp through the U filter that shows the illumination pattern.
There is no comprehensive data analysis package for OSMOS; however, Jason Eastman has written an IDL package that should be suitable for processing imaging data. His package is described at the bottom of his manual for the 4k detector.
John Thorstensen has also adopted his qccds.html script for OSMOS. OSMOS-specific documentation should be available in the near future.
The IRAF script bias4k.cl may be useful for quick-look bias subtraction at the telescope.
The connection to the IC is lost
Instrument Status Window in Prospero does not appear
No connection to the guider camera
Shutter does not open fully in cold weather
The shutter has not opened fully in very cold weather. This problem was reported in early January 2011 and manifests as partial to complete vignetting of the detector. The temperature at the time was approximately -7 C and, while these conditions are rare at the site, observers are strongly encouraged to be alert to this problem under comparable conditions.
The connection to the IC is lost
Once the connection to the IC was lost during a call to roi4k. The error messages were:
Updated: 2012 April 10 [pm]