How to Observe with LUCI

LUCI Software Start-Up

Other Software Start-Up

Pointing and Collimating

Running Scripts

Queue Manipulation and Error Recovery

Target Acquisition

Introduction

Longslit Alignment

MOS Mask Alignment

Manual Offsets with the Telescope Service

Calibrations

Twilight Flats

Darks

End of Night

Where is my data?

Basic Troubleshooting

LUCI Software Start-Up

LUCI can be run directly from the mountain obsN workstations as well as from the Tucson remote operations room rm525-N workstations, there is no longer any need to run a remote desktop into the LUCI computer as was done for LUCI1.  The user interface (UI) is still in active development, so keep in mind that some labels may change or move. 

  • Log into an LBTO workstation using your partner account (INAF,LBTB,OSURC,AZ).  In  the remote operations room in Tucson, log in as observer

  • Open an xterm and type open_observer luci2 all.  This will open all of the components (details on these are below) normally used to observe with LUCI2, including (in this order):

    • firefox session, with relevant tabs and bookmarks already configured

    • Two GUIs from the TCS will also be started: the instrument interface IIFGUI and the guiding control subsystem GCSGUI

    • The LUCI Observer GUI.

    • A display client for the readout service (GEIRS).  This takes about 51 seconds total.  When the initialization GUI pops up, change the “Observer” field and click OK.  The GEIRS control and display GUIs will then open.

  • On the Observer panel are buttons for the instrument manager GUI (the LUCI2 button), the readout manager GUI (READ2 button) and the real-time display (RTD2).  Click the RTD2 button on the Observer panel.  The LUCI2 and READ2 GUIs are really only needed if you want to do something manually.  They can be opened/closed as needed.

  • Position the arrangement of windows to your liking.  We suggest something like this.  The small GEIRS control gui is not normally needed, but should be left up.  Put it behind the active display window

  • Most subpanels and buttons on the Observer Panel have tool tips that will pop up if you hover the mouse on that location for a few seconds

  • The open_observer command now fully supports both LUCIs

Other Software Start-up

While you will not normally need the instrument, readout, or telescope GUIs when observing with LUCI, sometimes they are needed.  For example, you might want to enter a small offset on the telescope GUI to adjust the target centering on the slit, or manually take another image to confirm.  The instrument GUI and readout GUI are opened from the “LUCI2” and “READ2” buttons at the bottom of the right status panel on the observer GUI.  The telescope service can be opened from an xterm by typing open_observer luci telsvc.  The manual offsets section below details how to use the telescope service.  Once done with them, these GUIs can be dismissed.

Pointing and Collimation Check

  • Provide the OSA with the coordinates of your first target before the start of the night, preferably at or before sunset

  • Once it is dark enough, they will use the coordinates provided to slew to a nearby bright star and correct pointing and collimation, then turn the observing over to you

Running Scripts

Script execution is controlled from the queue action buttons on the observer GUI.

    • LOAD: navigate to your script in the popup and double-click it.  This will read the script display the individual items in the queue panel
    • GO!: to start the execution of the script or continue a paused script  
    • Reset: When a script is not executing (paused, stopped because of an error, for example), the reset button will reset the pointer to the top of the script and clear out any displayed “ready”, “error” or “done” messages.  
    • ABORT: When you abort a running script, LUCI will, as appropriate, stop waiting for a reply from the telescope and will interrupt any in-progress readout of the detector.  Instrument reconfiguration, in particular MOS mask motions, will not be interrupted.  
    • Locate: The queue panel indicates where in the script each section (telescope, instrument, and readout) is currently executing or done and waiting for other sections to catch up, but it does not auto-scroll.  The active subsection can be found by selecting it in the Locate menu
    • Other: Here you can enter names for the support personnel (metadata) or refresh the Observer gui.  Try refresh if your queue panel seems to be out of sync with the instrument status or where you are in the script.  The edit queue button currently does not work

Queue Manipulation and Error Recovery

Each line in the queue panel represents an ObsItem (or just item).  Each ObsItem summarizes (from left to right) the LUCI1 readout and instrument configurations (blue), the telescope/target information (yellow), and the LUCI2 instrument and readout configurations (pink).  Blank boxes indicate no change in that configuration from the box above. As the script is executed, the completed parts of the obsItems will turn gray.  Currently running sections are outlined in green or labeled with a green “running”

  • To insert a pause in a script after an ObsItem, right click the item to bring up the context menu and select pause after this item.  You can now insert a pause after an item that is currently executing
  • To continue a paused script, click Go! in the queue actions panel
  • To continue after a normal error, fix the problem and then right click the ObsItem subsection with the red error box to bring up the context menu and select continue.  Note the context menu will change depending on what functionality might be needed
  • Abort currently has a bug in it (Dec 2015), so please avoid using it.  Recovery requires your ISA to restart some lower level software if you do abort.  Under normal circumstances an abort would be used to get out of a script that is currently executing without any problems.  An alternative method is to insert a pause after the current obsItem and wait for it to complete what it is currently doing.  
  • If you are currently paused or stopped because of an error, you may load another script or click Reset and then in the queue panel scroll to the ObsItem where you want to re-start the script execution.  Right click to bring up the context menu and select skip to this item, then click Go! in the queue actions panel


Target Acquisition

Introduction

  • LUCI spectroscopic scripts contain both the acquisition imaging and spectroscopic observations.

  • A normal faint object MOS mask or longslit alignment requires four images: skysourceslit, and source+slit

  • Acquisition scripts prepared with scriptor will automatically take the sky, source, and slit images in this sequence without interaction from the observer and then pause.  You should still preview the acquisition sequence in the queue panel to confirm that this is what it will do.  

    • Sky frames might be omitted from bright source acquisitions.  Crossed filters may be used to minimize the chance of saturating the detector

    • All images in the sequence should be taken with the same readout configuration

  • While paused, the Aladdin display is used to interact with the images and calculate the alignment offset.  This offset is inserted into the queue by the RTD just before the source_slit image is taken.

Note: Please do not leave the mask in turnout for long periods!  Motor current is used to hold the mask in this position and over time it will warm up parts of LUCI.  This should happen less frequently with the new LUCI software than it did with LUCI1.

Longslit Alignment (replace screenshots with labeled “live” acquisition shots)

  • Select the “Long slit” tab in the RTD2 gui
  • Confirm the auto-populated filenames for alignment (source) and background (sky) in the target selection section are correct, then click subtract images or Single image depending on what you need
  • In the Aladdin display, click on your source.  You can zoom and click multiple times to get reasonably close to centered on your target
  • Click set1 then center target position(s) to centroid on your source
  • Confirm the auto-populated filename in the slit selection section is correct, then click use
  • In the Aladdin display, click on your slit.  You can zoom and click multiple times to get reasonably close to centered on your target
  • If you want your source centered somewhere other than Y=1024, uncheck the Use fixed Slit1 Y position box or change the Y coordinate where it will center your source
  • Click set then center slit position to centroid on your slit
  • Click calculate and check that the offsets make sense, then send if they do.  This will insert an offset ObsItem into the queue
  • Click Go! on the observer gui.  This will execute the alignment offset, take the source+slit image and then pause the queue.
  • If you are happy with the source alignment on the slit, click Go! again to start executing the spectroscopic portion of the script.  If not, you may manually execute a small offset on the telescope service gui and take another image to confirm your source position in the slit.

MOS Mask Alignment (replace screenshots with labeled “live” acquisition shots)

  • Select the “MOS” tab in the RTD2 gui
  • Click Open LMS file and navigate to your LMS file and select it
  • Normal MOS mask acquisitions include a rotational correction, so leave this selected
  • The MOS through-slit image is aligned first.  Confirm the auto-populated filename in the slit selection section is correct, then click use
  • Select a SlitID number for one of the reference slits along the top of the mask (1 and 6 are the best to use to get a good solution to the mask rotation), then in the Aladdin display, click on that reference slit in the image.  You can zoom and click multiple times to get reasonably close to centered on your target.  Repeat for at least a second reference slit
  • Click set position and confirm that the shifts look reasonable.  The rotation on LUCI2 is typically a bit under 0.5 degrees.
  • Confirm the auto-populated filenames for alignment (source) and background (sky) in the target alignment section are correct, then click subtract images or Single image depending on what you need
  • Set a reference target ID number, then in the Aladdin display click on that reference source in the image.  You can zoom and click multiple times to get reasonably close to centered on your target
  • Click Set position to centroid on this source
  • Click calculate.  This will auto-find the remaining alignment sources and then calculate the affine transform to align the mask to the sky
  • In the “target offsets” section at the bottom of the MOS tab on the RTD, check if any of the alignment sources have large residuals on the fit.  If so, uncheck its “use” box  and click calculate again.  If all seem off, select a second source and repeat the previous four steps
  • When you are happy with the solution, click send to insert an offset item into the queue
  • Click Go! on the observer gui.  This will execute the alignment offset, take the source+slit image and then pause the queue
  • Inspect the post-alignment image to see if the alignment sources are well-centered in their boxes.  If you are happy with the source alignment on the MOS mask, click Go! again to start executing the spectroscopic portion of the script.  If not, you may manually execute a small offset on the telescope service gui and take another image to confirm your source position in the slit.

Manual Offsets with the Telescope Service

Open an xterm and type open_observer luci2 telsvc. This will open the telescope service gui. The two main functions of the telescope service GUI are to (1) execute manual offsets to tweak spectroscopic mask alignment and (2) abort waiting for the telescope collimation.

(1) To execute an offset manually, enter the offsets you want, set the coordinate system (normally this would be DETXY, but check), click “Right Offset” to enable that offset (the button will turn green) and then click “Commit Offset” to send the offset request to the telescope.  

(2) LUCI will automatically wait for the telescope to reach collimation (an AGw measured wavefront error below 600nm) at each new preset.  If your collimation is good enough, you can save a little time by bypassing this wait for collimation.  Click the Abort wait for coll. button.  LUCI will automatically reset to wait for collimation on the next target.

The remainder of the fields on the telescope service GUI are needed for manually sending telescope presets, binocular operations, or recovering from problems, and will not be described further here.  

Calibrations

Twilight Flats:

  • Inform the OSA/ISA  as soon as possible if you plan to do twilight flats (dusk or dawn).

  • Narrowband flats are taken as early as 45 minutes before sunset (Br_gam) or after sunrise with the telescope parked at zenith, facing away from the sun.

  • Broadband flats are normally taken after sunset or before sunrise.  Ask the OSA to preset in TRACK mode to a Persson standard at relatively high elevation.  Make sure it is off the Galactic Plane.

  • For all filters:

    • Start with short 1×2.75s LIR exposures to check count levels

    • Keep maximum counts below 25,000 (half-well) if possible

    • Aim for a minimum of 5 exposures at each level with Delta_ADU of 5000 or more

  • Broadband filters:

    • K/Ks-band flats usually start just after sunset or complete just before sunrise

      • Monitor brightness in Br_gam before starting at sunset

      • Switch to K/Ks when Br_gam counts drop below 2000 ADU after sunset

    • H, J, or z bands are usually started immediately after the K/Ks flats have completed (at sunset, reverse order at sunrise).

    • We generally recommend small offsets between each flat so that any stars that appear in the frame can be removed when the frames are combined.  You can use a generic dithering script <Need to set these up for LUCI2> to take the data.  

Darks

  • Darks are better taken in the afternoon, before removing the blind mask, to minimize persistence effects from the previous night’s data

  • With the blind mask in the FPU they can be taken even with the lights on in the enclosure.

  •  Darks are stable over several nights; they do not need to be taken daily!  For most IR observations they are actually not needed at all.  The normal use for darks is to make a nighttime skyfkat out of science data.

  • Procedure:

    • Need to set up standard dark scripts for LUCI2

End of Night

  • If you plan to take darks the next afternoon, move the blind mask into the FPU, otherwise just leave whatever mask is in the FPU or have all masks in storage to minimize MOS motions
  • In all cases, configure LUCI with both blind filters and the mirror to block most light and stop grating tilt stabilization, plus ensure that the flexure compensation is turned off
  • Watch the status panel to make sure this completes
  • It is best to close at least the main Observer Panel and GEIRS windows when you are done and start with a fresh login at the beginning of your next night.  All windows can be closed by clicking the “X” at top right or selecting close from the File menu

Where is my data?

At the telescope, LUCI, like all facility instruments at LBTO, writes all new images directly to the /newdata partition.  This is mounted (read-only) on all of the observer workstations. The archive software detects these new FITS images and copies the data to /Repository/ as well as to Tucson for further distribution. The /Repository directory is also mounted (read-only) on all observer computers on the mountain (obsN) as well as in RemoteOps (rm525-N) in Tucson. The /newdata directory gets cleaned of previous nights’ data at noon each day, while /Repository directory preserves access to uncompressed FITS images from the recent past (up to ~6 months).  When off-site, or for access to older data, please use the usual archive web interface at archive.lbto.org.

If you are running an RTD, it also downloads a local copy of the FITS data and stores it in ~/.luci/readoutN.  You can take the data from here, but it is not an official copy and may be deleted randomly to save space.  We recommend you use /newdata for the current night and /Repository for earlier nights in your observing run.

Basic Troubleshooting

Everything listed under this Basic Troubleshooting section can be completed by the visiting observers and thus the directions are listed here. If these solutions do not work or the problem is more serious, please inform the OSA or ISA on duty.

 

Restarting Software Components

GUI Components Freeze or are Otherwise Unresponsive

The Grating Unit times out while trying to Stabilize

Calibration Unit (CU) fails to move in/out

 


Restarting software components

The open_observer script can also re-start individual components in case you run into trouble and must close something.  In an xterm, type open_observer luci2 and you will get a summary of the options.  As an example, if you close the observer GUI, it can be re-opened with open_observer luci2 observer

 


GUI components seem to freeze or are unresponsive

A new script has been installed called luci_snapshot that records various information useful to the SW developers in tracing down the bugs currently causing trouble.  Please open an xterm and run this script before killing any GUIs.  

The observing GUIs (observer, instrument, readout, and real-time display) are controlled by a single java wrapper.  If one thread is busy, this blocks interaction with the other GUI components sharing that thread.  However, the GUIs can also hang.  First verify that the UI is just not busy.  The two main things to check are that there is no popup waiting for acknowledgement, or if the RTD is downloading the most recent file.  If it is waiting for a file, just wait until the operation completes and try again 

If there is still a problem, close the guis and reopen them with the open_observer luci2 command.  To force a shutdown of one of the GUIs, you can run xkill on a command line and click on the problem GUI to close it.

Need to detail when the following are necessary:

  • restart scheduler

 

 

The Grating Unit times out while trying to Stabilize

  • Manually re-select the same CWL on the LUCI2 GUI and commit

  • Change CWL +/- 0.1um or so and commit

  • Re-submit science script with the correct CWL

  • Avoid high EL observations (>80 degrees) if possible

  • Ask OSA to note date, time, EL, and ROT in IT#2772


Calibration Unit (CU) fails to move in/out

  • Try to move it from the instrument GUI 
  • If it still fails, notify the OSA/ISA.  Sometimes it disconnects from the controller