SINFONI p2 Tutorial

This tutorial provides a step-by-step example of the preparation of a set of OBs with SINFONI, the "Spectrograph for Integral Field Observations in the Near Infrared" at ESO's Very Large Telescope (VLT). To follow this tutorial you should be familiar with the essentials of the use of p2.  Please refer to the p2 web page and associated sub-pages (see the menu on the left of that page) for an overview of p2 and generic instructions on the preparation of Observing Blocks (OB).

0: Goal of the run

In this tutorial we will prepare OBs for a simple example observing run, with the goal of obtaining Natural Guide Star (NGS) adaptive optics-assisted integral field spectroscopy data for an elliptical galaxy. The aim is to extract spatially resolved dynamics in the central arcseconds, to search for black holes. Additionally, we need observations of a velocity template giant star. The sample OBs will illustrate the use of a variety of features of p2 and the kind of decisions to be taken at the time of preparing an observing run, as well as some aspects that are specific to the preparation of OBs for SINFONI.

1: Getting started

The Phase 2 process begins when you receive an email from the ESO Observing Programmes Office (OPO) communicating to you that the allocation of time for the coming period has been finalized and that the results can be consulted in the corresponding Web page. The receipt of this email shows that you have successfully activated your User Portal account, and the combination of ID and password you are using to log into the portal is also what you require for the use of p2. You follow the instructions given by ESO and find that time was allocated to your run with SINFONI. Therefore, you decide to start preparing your Phase 2 material.

First, you collect all the necessary documentation:

For the sake of this tutorial, we will use the p2 demo facility:

This is a special facility that ESO has set up so that users who do not have their own p2 login data can still use p2 and prepare example OBs (for example, while writing a proposal to get the overheads right!).  Indeed, all users have access to the same facility, and each can see what all others have done.  You cannot use it to prepare actual OBs intended to be executed.  When you prepare such OBs you should use your ESO User Portal credentials for p2 (with

After directing your browser to the p2 demo facility the main p2 GUI will appear as follows:


Runs for a number of instruments appear in the lefthand column, since the p2 demo facility is used for all of them.  Similarly, if you log into p2 (instead of p2demo) with your own ESO User Portal credentials, you will get the list of all the runs for which you are PI, or for which you have been declared as a Phase 2 delegate by one of your colleagues.

Select the folder corresponding to the SINFONI tutorial run, 60.A-9253(H) by clicking on the + icon next to it.  In this tutorial we assume that time was allocated in Service Mode.  This is indicated by the small wrench icon that appears next to the RunID.  "Inside the run" you will see (at least) one folder, called "SINFONI Tutorial".  That folder contains the final product of the tutorial you are now reading.  You can refer to the contents of that folder at any time, perhaps to compare against your own work.

You can now start defining your observations.  In the p2 demo you must first start by creating a folder, within which you will put your own content.  Because the p2 demo is online and accessible to anyone you may see folders belonging to other people here, though after-the-fact cleanup is encouraged and possible.

You can now start defining your OBs.

to the top

2: Creating the first OB

To create a new OB, open your folder by clicking on the + icon next to it (if it's not already open) and then click on the OB button. This will create an "empty" OB on the ESO database, ready for filling in.  It is not only good practice to keep track of things, but also will avoid errors down the road if you provide the OB with a name.  In this case why not name this new OB, say, "Galaxy OB".

The p2 window in your browser will look like this:


to the top

2.1: Setting the target information

To enter information on the target, click the "Target" tab at the top of the OB. 

  • In the Name field type the target name (NGC 1399).
  • For the coordinates you can click on the "resolve" button to the right of the Name field.  This will query the SIMBAD database to look up the coordinates of the target as specified (note that this requires SIMBAD to be available).  Doing so will return the values 03:38:29.080, -35:27:02.60 for the Right Ascension and Declination fields just below the target name.  Of course in any event (resolved button clicked or not) you are free to enter any (other) coordinates you want for your target (if, for example, you have reasons to prefer coordinates from one wavelength over another).
  • Since the coordinates are given for both epoch and equinox J2000, leave these fields with their default values (though note that these values are not retrieved from SIMBAD if the resolve button is clicked.
  • The proper motion of this target is negligible for the purposes of this example, and differential tracking of the telescope is not needed since this is not a moving Solar System target. Therefore, you can leave the last four fields in the Target tab set to their default values of zero.

to the top

2.2: Setting the constraint set

As stated in Section 1, we assume for the purposes of this tutorial that the program has been allocated time in Service Mode. You thus need to specify a constraint set for your OBs. You can do this by clicking on the Constraint Set tab and filling the entries under it:

  • First, give a descriptive name to the constraint set about to be defined. Since you have decided that this constraint set will be applied to all galaxy observations (those for the star will be less stringent on seeing), you type Galaxy constraints in the Name field.
  • You are not interested in accurate photometry for the galaxy, only need to get approximately the flux that you anticipated. Therefore, you should enter Clear conditions in the Sky Transparency entry.
  • Since you need moderately good quality in your images, you specify 0.8 as the value of the Seeing field, consistent with what you proposed for.
  • As this target rises to very low values of airmass you will not be overly restricting the observability of the target, nor compromising the adaptive optics correction by setting the airmass constraint value to 1.2.
  • Since you are doing medium resolution spectroscopy observations in the near-infrared, and the AO reference star (which is sensed in the optical) is quite bright, the lunar illumination has hardly any influence. You can thus leave the default values of 1.0 and 30 degrees for the Lunar Illuminationand Moon Angular Distance fields.
  • The Strehl ratio was predicted by the Exposure Time Calculator (ETC, a portion of which is shown below) for your selected set of conditions, and you need to specify it in the Strehl field (this would not be required if you intended to perform seeing limited observations). Your (imagined) star has a B-R colour of -0.2, an R-magnitude of 11.5 mag, and it is 15.6 arcsec away from the nucleus of NGC 1399. You enter these values into the corresponding fields in the ETC, together with the seeing and the airmass. The Strehl ratio which needs to be inserted into p2 is the on-axis value. For our example, the predicted value we enter is 32.0.
  • Finally, since these observations will use adaptive optics you need to toggle the value for the field Atmospheric Turbulence Model to "default Paranal turbulence model".  This is used in the scheduling of AO observations at the telescope.

Note that in your Phase 1 proposal you already specified some of these constraints (lunar illumination, seeing, transparency). You must make sure that none of the constraints specified in Phase 2 is more stringent than the corresponding one specified at Phase 1 (Phase 2 constraints must agree with Phase 1 request).

to the top

2.3: Setting the time intervals

Observations of the galaxy can be carried out at any time during the semester when the target is above the specified airmass. Therefore you can leave the Time Intervals field unchanged.

However, for this example we will assume now that the observation has to be obtained in Jan 2019, since you will have simultaneous observations at other wavelengths then. You can specify this by clicking the Time Intervals tab. Using the "Add" button and the fields/calendars you can set the time interval:


Here we note two important matters:

  • If your observation could be executed in other, non-contiguous time windows, you could define more intervals in the same way as described.
  • The timeline associated with the Absolute Time Constraints tab will not display useful content for the demo of p2, but will, of course, provide useful content for "real" runs. 

Note that such time critical aspects must also be explained in the section on "Time Critical Aspects" of the Readme. It is not sufficient to enter the constraints only here.

2.4: Back to the Observation Description

After clicking on the Obs. Description tab you see three free-text fields at the top: Observation Description Name, User Comments, and Instrument Comments.  

It may be useful in many cases to have an easy way of identifying an Observing Description, like when having observations in a number of instrument setups performed for the same target. The "Observing Description Name" field allows you to define such names.  In this example OB, the Observing Description will simply be an object/sky sequence using a fixed sky position. We can thus appropriately name it K - 0.1 - FixedSkyOffset. We enter this name in the Observing Description Name field.

The User Comments field can be used to communicate brief details of the observations (which should be explained in more detail in the README).  Finally, in the Instrument Comments field you should list the IR magnitude of the brightest source in the field-of-view  (something which is also required on the finding chart and/or in the README).

to the top

2.5: Defining the acquisition template

The first template that must be part of any OB is the acquisition template, so let us define it next. In the Template Type list, make sure that the acquisition entry is highlighted. This will list all the acquisition templates available for SINFONI in the Template list next to it.

After reading the description of the templates in the SINFONI User Manual, you have determined that the SINFONI_ifs_acq_NGS template is the most suitable one for this particular observation. You thus click on this template in the Template list, and then on the Add Template button next to it.

You need to decide now on the acquisition parameters. This acquisition template sets a grating and takes an OFF frame at an offset of Alpha offset to sky and Delta offset to sky (both in units of arcseconds) from the target position for sky subtraction purposes. Then it takes exposures in open loop presenting on the Real Time Display at the telescope console the image obtained after DET.NDIT integrations of DET.DIT seconds each, minus the OFF frame. This allows the identification of the object or offset star in the case where the target itself will not be directly visible during acquisition.

Using the ETC, you found out that a DIT of 2s is far from saturating the detector, but the SNR is high enough to allow a centering of the target. For this bright AO reference object, NDIT is best set to 1. Indeed, the acquisition is done on the natural guide star (NGS), not on the target, therefore the DITs in the first and 3rd row can differ. As the label suggests, the "Integration time" is for the target, and the "NGS integration time" is for the reference star. We de-select the option."Target = AO Guide Star".

We have imagined that our selected elliptical galaxy has a foreground star bright enough to be used as an AO reference target near the galaxy nucleus. We enter its coordinates in the fields RA of AO Guide star and Dec of AO Guide star.  As shown above our natural guide star has a B-R colour of -0.2 mag, which we enter in the field: "NGS B-R colour".  It is point-like, so we leave the FWHM of AO Guide Star at 0. We don't care about the Position angle on the sky and leave it at 0.

There is no necessity for us to specify the telescope guide star ourselves, and by leaving the next three fields at the default value, the selection is up to the telescope operator.

We are acquiring quite a bright target, therefore we will be able to identify it even without going to sky first. We can set Alpha/Delta offset to sky to "0", which will save time (not only in "real" exeution time at the telescope, but also in the estimated execution time: you can try this out once you have added the science template and filled it in, see below).

The SkySpider position refers to an observing mode which is not yet offered, and the default value of "ref" should not be modified.

SPAXEL size (arcsec) refers to the size of a "spatial pixel" in your datacube, and you have the choice between three scales. We want to perform AO assisted observations, for which the two smaller size scales are most sensible, and since we want to cover area sonably large field ("reasonably large" being 3 arcseconds on a side), we select the 0.1 arcseconds SPAXEL size. The acquisition is automatically done on the largest scale first (with the exception of the "fast acquisition" template), in order to allow for coordinate or pointing uncertainties, but here you should always specify your intended scale for the science observations.

Instrument setup name gives us the chance to select the wavelength range, we use the pull-down menu to select K-band. Unless there is a very strong reason to do otherwise, acquisition should always be done in the same band as the science observation, because moving the grating takes very long with SINFONI.

The set of parameters that you choose in your acquisition template is thus:

  • Integration time (DIT): 2
  • Number of Integrations (NDIT): 1
  • NGS integration time (DIT): 2
  • Target = AO Guide Star: set to "no"
  • RA of AO Guide Star: 03:38:29.156
  • Dec of AO Guide Star: -35:26:45.118
  • R mag of AO Guide Star: 11.5
  • NGS B-R Color: -0.2
  • FWHM of AO Guide Star (arcsec): 0
  • Pupil tracking mode: set to "no" (these observations are to be done in "classical imaging" mode)
  • Position angle on the sky: 0
  • Telescope guide star selection: CATALOGUE
  • RA of telescope guide star: 0
  • DEC of telescope guide star: 0
  • Alpha offset to sky: 0
  • Delta offset to sky: 0
  • SkySpider position: ref
  • SPAXEL size (arcsec): 0.1
  • Instrument setup name: K

If you specify one of the parameters outside the allowed range, the field shows an error (pink background with red characters). The OB is not considered valid unless all parameters are defined within allowed ranges.

The acquisition template is now complete and should look like this:


to the top


2.6: Defining the Observation Description

Once the acquisition is completed, the science observation begins. The science observation is defined in a set of one or more science templates.

So, the science templates need to be added. When reading the manual and considering the scientific requirements of your program, you noticed that you will need a large sky throw in order to really end up on a sky field which is outside your galaxy, and you wish that a particularly "good" piece of sky shall serve as your "off" position. We are using the template SINFONI_ifs_obs_FixedSkyOffset.

You select Template Type science, then highlight SINFONI_ifs_obs_FixedSkyOffset and click on Add Template. The template will be attached to the OB next to the acquisition template selected and filled in previously.

The template has to be filled in with the specifications for integration on your science target, i.e. the galaxy. The ETC tells you that within reasonable limits for DIT (up to 900s), there is no risk of saturation. So your only constraints are to get a reasonable number of positions on target and on sky in order to efficiently remove bad pixels, to get the highest number of counts within one frame, and to stay within the 1h exposure time limit imposed by the scheduling requirements of Service mode observations. Weighing these constraints against each other, you decide to go for DIT=600s, with two AB cycles.

The Jitter Box width specifies the full width of the box around the nominal position in which the pointings (both for Object and for Sky) are shifted around. A size of 0.5 arcsec is sufficient for our purpose to get rid of bad pixels.

You spend equal amounts of time on sky and on object, and therefore set both, NDIT for the OBJECT positions and NDIT for the SKY positions to 1. To set Number of exposures per offset position to something other than 1 makes sense only for bright targets or time series, so wecleave it at the default value. Alpha/Delta offset to sky give us the chance to specify a location where the sky is to be taken, and an offset of 100 arcsec north (100 in Delta) moves the telescope to a clear field.

The Spectral Dithering flag can only be used in Visitor mode, so we leave it untouched.

Then, you select again the SPAXEL size and the Instrument Mode, like in the acquisition.

Your science template now has the following parameters:

  • Integration time (DIT): 600
  • Jitter box width: 0.5
  • Number of AB or BA cycles: 2
  • NDIT for the OBJECT positions: 1
  • NDIT for the SKY positions: 1
  • Number of exposures per offset: 1
  • Alpha offset to sky: 0
  • Delta offset to sky: 100
  • Spectral dithering: remains off (set to "no")
  • SPAXEL size (arcsec): 0.1
  • Instrument setup name: K

After filling in these details you click on the "Exec. Time" button (at the top of the OB) to compute the execution time of the OB, and to verify that that execution time is compliant with the Service Mode one-hour length rule.  The reported execution time is 53 minutes, 22 seconds.  Thus, your OB should look like the figure below:


to the top

3: Defining the template star OB

In order to determine the velocity dispersion of the stars in the galaxy, a velocity template star spectrum is required. You selected a giant star (spectral type M0III) for this purpose. The science template to use (not only for telluric standards, but also for velocity standards) is SINFONI_ifs_cal_StandardStar.

Create a New OB like you have done before, and name it M0III template star.  Add the SINFONI_ifs_acq_noAO acquisition  template (the star is so bright, and we care only about spectral resolution not about image quality, therefore we use the no AO mode), fill it in in much the same way like for the science OB.  A few parameters less need to be specified, but some are new.  For example you should request to flatten the Deformable Mirror to get at least a somewhat reasonable image quality, and you could in principle add a blind offset to the "real" target, but we leave it at zero, since we want to observe the star itself.

One other difference to science OBs is that you are restricted in the selection of DIT, NDIT and Sky Throw (for the acquisition only, for the real SINFONI_ifs_cal_StandardStar template you can choose any values): Since standard stars are typically very bright, DIT has to be lower or equal 1s, NDIT=1 and both Sky Throw values are supposed to be zero (if you happen to have a standard star which is so faint that for the acquisition you need other parameters, you need to submit a waiver request, but for our example, that is not the case).

Display the calib list of templates, and add the SINFONI_ifs_cal_StandardStar template.

Some of the parameters are the same as in the SINFONI_ifs_obs_FixedSkyOffset template, but the following paramters are slightly different:

List of number of integrations allows to specify different values for the number of DITs for the different positions, for example to have different NDITs for the Object and the Sky positions. We want again equal amounts of time spent on Object and on Sky, therefore it is sufficient to fill in only one number, we set it to 10 (multiplied by a DIT of 3s, this gives the required time per position of 30s). The Number of offset positions parameter indicates through how many of the positions in the List of offsets in RA or X and List of offsets in DEC or Y the template is supposed to cycle. We set it to 2, for one Object and one Sky.

The List of observation types pops up an interactive window, where you have to specify for each position in your offset list, for as many frames as you have Number of offset positions, whether the position is on Object or on Sky. In the popup window, you populate the righthand list (Selected) by clicking on the arrows next to have to the O and S labels in the lefthand (Available) list.  When you have completed the list, you click on done, which will update the field in p2.  This information is used by the pipeline, for example. We set this to the sequence O S.

The Offset coordinate type allows the selection between Detector and Sky. Note that this selects only the frame of reference. The offsets are always given in arcseconds, (never in pixels, not even for "Detector"), and they are relative! For our List of offsets we select 0 0 in RA, and 0 30 in DEC. This means that the first image will be taken on source, the second on a sky field 30 arcseconds to the north.

Your standard star template now has the following parameters:

  • Integration time (DIT): 3
  • List of number of integrations 10
  • Number of offset positions: 2
  • List of observation types: O S
  • Offset coordinate type: Sky
  • List of offsets in RA or X: 0 0
  • List of offsets in Dec or Y: 0 30
  • Spectral dithering:
  • SPAXEL size (arcsec): 0.1
  • Instrument setup name: K

After doing this, and calculating the execution time of the OB as above, the contents of your OB should look like this:


and we will assume that this completes your run.

to the top

4: Finishing the preparation and other final steps

With the completion of the standard star OB, we consider the examples developed in this tutorial to be finished. 

Though, for a real run, you would also make sure to:

  1. click the "Check" button (to the right of the Exec. Time button) for each OB.  Check performs consistency checks on the selected OB(s).
  2. click on the Certify" button (to the right of the Check button).  Certify does the same as Check, but additionally (if the check was successful) marks the OB(s) as fully compliant and reaxdy for ESO's review.  It also marks the OB(s) as readonly.
  3. (optionally) select one or more certified OBs and click on the "Revise" button (to the right of the Certify button) to revoke the certified status and allow re-editing.  If this is done it steps 1 and 2 above should be re-done for the OB(s) in question.
  4. complete the information in the README file
  5. click on the "Notify ESO" button (to the right of the Revise button).  This will lock all certified OBs as well as the README, in preparation for ESO's Phase 2 review process.

As mentioned above, all users have access to the p2 demo facility.  And, unlike the situation that existed with p2's predessesor, P2PP3, all OBs are created directly on the ESO database (there is no longer a "check-in" requirement).  Hence, all users can see what you leave behind.  Indeed, after a while, if not properly maintained, the p2 demo runs could and would get cluttered with all manner of OBs, containers, etc.  Thus, as a courtesy to the next user who follows this tutorial, we would like to ask you to finish these exercises by deleting the OBs, containers, etc. that you make.

Finally, please note that when looking at the tutorial run in the p2 demo facility you will see that both OBs are marked as (+) Accepted (see the small screenshot below).  This is a state that these OBs have been set in (by ESO) solely to avoid them being inadvertently deleted from p2.


Instrument selector