Binospec and BinoMask special issues

This page covers some special or black-belt issues for the Binospec spectrograph at MMT, and the BinoMask slitmask design software.

Large target catalogs

If you have a lot of targets, for example if you have a primary sample and a large catalog of 100s-1000s of lower-priority filler targets, slitmask design becomes more difficult. Two problems are: (1) it’s hard to see where your primary targets are in the js9 sky view; (2) with many targets, optimizing the mask placement over RA/Dec and PA can take a long time and time out, giving you a “Problem connecting to backend server” error and failure to place slits.

A method for dealing with this is to make two catalogs: the first contains your primary targets only, and the second contains your primary and filler targets. (Thanks to Christopher Willmer for suggesting this method.) Begin designing the slitmask with only the primary targets, and in Generate Slits, vary the PA and mask center to optimize the mask placement. Then record the mask RA, Dec, and PA, but don’t submit the mask. Go back to the Target List tab and select the full catalog with filler targets. (It should retain the optimized mask RA, Dec, and PA, but it is wise to record them to be sure.) Now go back to Generate Slits and keep the RA, Dec, and PA fixed at the previous values. This will place slits, including the optimized primary targets, and add filler targets as available.

Variable slit lengths

Currently, BinoMask only allows you to select a global minimum slit length (6 arcsec is recommended). It will extend slits that don’t collide with other objects, but you cannot specify that object 1 needs a 60 arcsec slit while objects 2 through 1000 can use 6 arcsec slits, (eg a large central galaxy and small satellites), or that two classes of objects need different lengths.

It is possible to make different slit lengths by designing a mask with the usual short slit lengths, and then hand editing the JSON file that is produced by Binomask to lengthen some of them. Design the mask, and at the end, before submitting, save the mask to a file – this should download a file called mask.json. This is a text file with data structures, and is easiest to read in a JSON viewer/editor – there are many online, for example

You will need to find the structure of the slit of the object that you want to lengthen. The slits are numbered within the “sidea” and “sideb” structures, and each slit structure contains RA, Dec, and object id, along with many other numbers. The field “polygon” in the slit structure has 8 numbers (4 X,Y pairs) that specify the slit corners in mm, not in arcsec. X is the dispersion direction and Y is the direction along the slit; you want to edit the Y-numbers, which are indexes 1,3,5,7 (0-based). The plate scale of the 6.5-m f/5 MMT is 0.16 mm/arcsec. After you lengthen a slit, you must remove the slits that would overlap with it in the Y-direction, by deleting their data structures from the JSON file. The standard spacing between slits in the Y-direction is 1.5 arcsec (0.24 mm), so leave that margin between your new slit and the remaining slits.

Clearly, this is a black-belt technique that risks creating a bad mask specification or overlapping slits, so leave yourself plenty of time to work on it. You may be able to load the modified json file back into Binomask to visualize it. You may not be able to submit the modified file in the normal way; if so, email the JSON file to Sean Moran at SAO and Ben Weiner at MMTO.

Querying GAIA for guide stars

Normally, Binomask will simply query a server at the CDS for guide stars from GAIA and you won’t have to worry about it. However, we have had an issue where the CDS server was down, and no guide stars were retrieved. If this happens, you can query the GAIA catalog from a mirror of Vizier, such as the CADC mirror. You want stars that are roughly between Gmag 11 and 18 – bright stars for wavefront, and guide stars of mag 16-18. You need to make a CSV file that has name, ra, dec, magnitude, pm-ra, pm-dec, epoch. Rough instructions follow.

Visit the GAIA DR2 Vizier.

This should download a .tsv file with perhaps several hundred stars. It has a long header, a footer, a header line of “RA_ICRS;DE_ICRS;Source;Epoch;pmRA;pmDE;Gmag”, a line of dashes, and the data. Make a .csv file with a single comma-separated header line by editing the file:

This should make a file that you can upload to Binomask in the guide stars tab. I don’t think the column order matters as long as you get the header names in the correct order to match the data.

Benjamin Weiner, bjw @