ING Banner
Home > Astronomy > Telescope Operator Notes > Working with ISIS


ISIS is a high-efficiency, double-armed, medium-resolution (8 - 120 Å/mm) spectrograph, capable of long-slit work up to ~4' slit length mounted at the cassegrain focus of the WHT. Use of dichroic slides permits simultaneous observing in both the blue and red channels, which are optimised for their respective wavelength ranges.

The default detectors are a large, thinned EEV12 4096×2048 device on the blue arm and the RED+ 4096×2048 device on the red arm. Pixel scales are 0".2/pixel in the blue arm and 0".22/pixel in the red arm. Spectral resolution depends on the grating in use and the slit width. Two low noise frame transer devices (QUCAM2 and QUCAM3) are also available for fast spectroscopy mode, both having a pixel scale of 0".19/pixel.

Spectropolarimetry and imaging polarimetry are also available.


ISIS Astronomy page.

CAGB Softwware manual

ISIS Software manual

Acquisition Tool manual


  • USER> FOCUS 97.8-97.85 (2017)
  • ZEROSET (az, alt, rot)
  • CAL (last, or faint on first night)

  • TVSCALE: Slit
  • TVFOCUS: 9250
  • Set the autoguider focus:  TO@taurus> autofocus 2400 (2015)
Telescope focus:  At the start of every night the best telescope focus is measured by taking exposures of a standard star or pointing grid star at different focus positions , using a wide slit (8 arcsec) and measuring the spatial profile. The observer will do this.

To save readout time, you can window the TV so that it displays only the region that is illuminated. Do this using the Options menu and selecting Camera. 
(There are focal reducer optics for the TV with two possible positions: TVSCALE 12 (slit?) or TVSCALE 5 (direct?). The default position is TVSCALE 12 (slit?). In this position, a 18"/mm optical barrel is deployed to permit viewing of the full slit with AG4. With TVSCALE 5 (direct?), about 60" of the centre of the slit is seen.)

Rotator centre:
After an instrument change you should determine the rotator centre and do the 7 star calibrate procedure, which takes about 20 minutes. This would only be done on a service night, unless there are pointing problems, in which case it can be done anytime.
You could use AGCOMP, which is the top surface of the comparison mirror and is believed to be stiffer than the slit viewing mirror, but it's normally done with AGSLIT. 
Measure the star position on the TV screen (x,y), then rotate 180 degrees in Sky PA and measure again. Calculate the midpoint, this is your rotator centre.

7-star Calibrate procedure:
Done with rotator tracking turned off:
USER> cal faint
Centre all the stars on the rotator centre and hit HANDSET after each one is done. 
The sky sigma should be <4 arcsec, but it's OK if it's a little higher.

Determine aperture offset:
There is no need to define an aperture.



NB: if using the R158R grating at very red wavelengths beyond 9000A, do not use mount PA positions between -8 and 64 degrees; rotate by 180 degrees if necessary to get the required angle.

When slewing to a new object watch out for limits (rotator, azimuth) and if necessary use:

TV Acquisition:

The acquisition tool is not used for ISIS. Instead acquisition is done using the handset to make t/s movements until the target is well centred in the slit.

The ISIS slit is tilted 7.5 degrees with respect to the incident light beam, so that the reflected light can be picked up by the TV system to enable slit viewing while observing. There is no filter in the beam (displayed on mimic in red). The DS9 display must be flipped the image in the X direction to agree with the HANDSET X offset direction in Slitview. Do this by selecting X on the Zoom menu. The sky can also be seen in AGCOMP mirror position ("direct view") - with this the X direction is opposite to that in Slit viewing, i.e. X is not inverted. In slit viewing mode the full unvignetted slit width of 4 arcmin can be seen on theTV.

The orientation of the sky with Cass TV slit-viewing depends on the sky position angle (PA). Use the metal dial to work out which way to position the finding charts. 
At Sky PA 270, North is up and East is left and at Sky PA 0 North is right and East is up etc.

Acquisition is done with the slit-viewing mirror in (TO> agslit), taking exposures with TV. The pointing is always very good (<10") and a single object is always found close to the rotator centre. It is not necessary to acquire exactly onto the rotator centre: to save time, you can move only in Y to centre the target in the slit. The position can be changed to avoid dust particles on the slit or defects on the chip. For faint targets, it may be necessary, perform a blind offset from a reference star to the science target.
To accurately recover the same acquisition position on later nights, log the following parameters: RA/Dec of the guide star (to make sure you use the same one), telescope offsets, sky PA, probe position (r and theta) and X- and Y-position of the guide star as seen on the TCS. On the following night make sure the Sky PA and probes are in position on the same guide star, then type 'USER>auto on x y'.

Use the standard uDAS autoguider for CASS. Check that the position of the "Autoguider to TCS Selection" switch is in the correct position. 
This was last run from udasdev2 (whtdas18),  with 'obssys', 'startobssys' and 'startag AG6'. 
The monitor should then display a camera mimic showing temperatures (rather inaccurate) and camera status. DS9 display tool will also appear as well as the TV Guider Control. This GUI allows you to change exposure times, take fields and set the TV and Guider going. Alternatively, all these can be done from AutoGSS. Be careful not to allow the guide box to drift outside the area of the CCD as this causes guiding to fail. The Cass autoguider has a FOV of 1'x1'.

AutoGSS is now used for finding guide stars. See the User Guide

When the AG probes are in position the message PROBE <XXX> <XXX> will appear at the TCS User window.

Autoradial must be in the range 0-40000 microns

Autotheta must be in the range 0-180000 millidegrees  (NB: autotheta of 0-35,000 will vignette the slit viewing camera!)

If the guide star needs to be recentred, you can use the AutoGSS + and - buttons: to move the guide star up, increase autoradial and to move it left, increase autotheta (and vice versa). Size of movement: to move the guide star from the right edge to the left edge, increase autotheta by ~6000, and to move it from the bottom edge to the top, increase autoradial by ~12,000. Start guiding and check the status on the TCS DISPLAY changes from TRACKING to A/GUIDE.


The polarisation optics consist of polariser and analyser in the beam above the ISIS slit: in practice these are a wave-retarding plate and a calcite block, respectively. The latter separates beams with 2 different polarisations. Instead of one spectrum we therefore obtain two representing a polarised and non-polarised signal respectively. For linear polarimetry the polarising element is a half-wave plate (hwp), for circular polarimetry it is a  quarter-wave plate.

The polarisation optics significantly reduce the visibility of the slit reflection, so that during exposures, no object can be seen on the slit. Acquisition is also a little more difficult because a multi-aperture DEKKER (usually 3 slots in the spatial direction) is also in the beam above the slit, which reduces the parts of the slit we can see in the TV view.

Telescope Focus:

The waveplates above the slit also alter the f ratio of the beam. To re-focus the star on the slit when the polsarisation optics are in the beam, use the following values:
  • Decrease telescope focus by about 0.2mm with respect to the standard ISIS focus. This should be checked at the start of the night by the observer, with the polarisaion optics in the beam. Focus with spectropolarimetry using the Half-Wave Plate in was around 97.77 (Feb 2015)
    This can now be automated using a GUI hwp "Dfocus" setting of -0.2mm (the observer will do this if they want to)
  • It may be necessary to decrease the TV focus by 3000, but it's probably not necessary since the focus isn't very different (very unlikely to be needed).
  • A corresponding decrease in the autoguider focus of a few hundred will probably be needed too.

Working in polarimetry mode will mean frequent focus changes as different focus positions may be required for acquisition, linear polarimetry and circular polarimetry. It is reccomended that the observer adds all necessary focus changes to their observing scripts, particularly if they are doing both linear and circular polarimetry. The SA should determine the correct focus offsets on the first night of the run. It is probably NOT necessary to refocus simply for acquisition, therefore if you are doing just one type of polarimetry then you can use this focus value for the science and it will be OK for the acquisition too, so focus changes during the night are not needed.

Acquisition for Spectropolarimetry:

Usually a fixed Sky PA such as 90deg will be used all night. For acquisition, the polarisation optics must be removed so that the target and slit can be viewed on the acquisition TV, otherwise nothing will be seen. e.g. to remove the Half-Wave Plate or Quarter-Wave Plate:

TO@taurus> hwout or qwout

Acquisition can be done using AGSLIT (no use of AGCOMP is needed). The appropriate dekker should be in place throughout, e.g. the dekker called POL18ARCSEC, which has a gap of 18 arcsec between the centre of the central dekker slot and the centre of the left or right dekker slot.

First mark the position of the slit on the TV. You will probably need an exposure time of ~2s to see the slit clearly. Also mark the positions of the dekker slots which are going to be used, usually the central one only. The slit is located towards the top of the 3 dekker slots. With no TV windowing, the slit was found at y=457 (July 2012, similar in 2015). If you are already on a bright target and you can't see the slit or dekker slots clearly, offset the t/s to move the bright target away while you make the marks.

Acquire the target onto the intersection of the slit and the dekker slot you want to use. There might be some strange-looking reflections of the star around the edge of the dekker slot, this is normal. Start guiding when well acquired and wait for a couple more readouts to make sure the acquisition is good. Remember that the polarisation optics need to be put back in before observing (the observer might need reminding) or the TO can do it:

TO> hwin to put the Half-Wave Plate back in
TO> qwin to put the quarter-wave plate back in

Important: The dekker and slit positions will change very significantly due to flexure and you must not rely on the markers for any subsequent acquisitions!


See the ISIS Imaging Polarimetry page, in particular the section on configuring the telescope. This contains full instructions for acquisition and observing in impol mode.

Top | Back

Contact:  (Observing Support Assistant)
Last modified: 13 November 2017