Object through slit

Two possible situations:

- target is a relatively bright object and you can see it in imaging mode with exposure times typically shorter than ~10 seconds. Use "centering on the slit" (mode 6 of the GUI) button and press "start exposure". This first removes the slit from the field of view and takes continuously images with the choosen exposure time. You then have to ask to the night assistant to move the target onto the slit "x" position (and in the general case, approximately at "y"=500). In this case, you can finally use again "Search Slit Position" (mode 5 of the GUI) to check if your target is within the slit. It may appear tempting to directly start the science exposure without verifying that object is through the slit. However, just imagine how disappointed you will be after "science" exposure if you realize that you missed the object...

If using the T193 "bonnette" guiding, it is recommended to check about every 15min if your target is still within the slit ("Search Slit Position" mode 5 of the GUI) due to telescope mechanical flexures (see also the Cook Book and the test report). A long spectroscopic exposure is therefore preferably split into several ~15mn exposures at the moment (until the new guiding system directly attached to Mistral will be available).

- The target is too faint to be detected with short exposure times (~10seconds). The previous process will then become very uncomfortable and you will need to use a differential positioning on the slit. For this, you first need to make a deep enough science image (mode 3 of the GUI: "Science Image"+"Star exposure") to see the target. On this image you note the target's x,y coordinates. You then compute the difference [ (xslit-xtarget)*spatial_sampling] in x position between the object and the slit, in order for the night assistant to know by how much he has to move the telescope in RA. You do the same [ (500-ytarget)*spatial_sampling] in Y position to approximate the DEC shift to apply. The value of the spatial sampling is 0.48 arcsec. This process assumes that the field rotator is oriented north-south. This process is not possible for now if other orientations are set. At this step, it is highly recommended to cycle a second time the whole process: deep image without slit, locate the target coordinates, recompute the offset (xslit-xtarget)*spatial_sampling, and apply eventually the RA shift; and same for DEC.

Given the MISTRAL imaging capabilities, trying to observe an object not visible in 10seconds exposures will be, however, pretty rare.