SPHEREx Data Explorer at IRSA Known Bugs and Issues

Software version released: December 2025.

This list last updated: December 2025

Reporting Bugs
Suggested Browsers
Known Major Bugs
Known Minor Bugs
Idiosyncrasies

Reporting Bugs

The known bugs and issues in this version of the SPHEREx Data Explorer at IRSA are listed here. If you think you have found a bug, before reporting it, please check this list, and read the online help. It may be a "feature" we already know about.

If you have found a real bug then please report it to the Help Desk. Please include your operating system version and your browser software and version.

Suggested Browsers and Platforms

Safari: at least version 15
Firefox: at least version 120
Google Chrome: at least version 120
Edge: at least version 120
Tablets: iPad Safari : Seems to work ok, though not on the iPhone. Cursory testing of a Droid tablet suggests it probably works ok.

Older browser versions than these may work, but we may not be able to support bug fixes for them.

Known Major Bugs

Jobs may be lost. Spectrophotometry jobs take a long time to run. The jobs should be retained in the same browser session for up to 2 weeks.
1. If there is a situation in which servers go down (such as regular Tuesday morning blocktimes every other week), the jobs may be lost, even those that have already finished running. (We are working on ways to ameliorate this.)
  Workaround: Download results from spectrophotometry jobs as soon as they complete to be sure that you have them. Watch for blocktime banners and don't start big jobs right before a posted blocktime.
2. In testing, we have found some transient and difficult-to-trace gremlins that may wipe the record of your jobs at your end overnight (e.g., much faster than two weeks), even though we can still see them at our end on the server.
Workaround: (a) Logging in (see the SPHEREx Data Explorer online help for more information) before submitting your jobs seems to keep the gremlins at bay for longer. (b) You should download the data as soon as you can.
Duplicate source names causes chaos. When you upload a list of targets, it's not supposed to require that you identify a column for names from you. However, if you did specify names in at least one other time in the session, it may require that you do so in subsequent requests. Moreover, if your uploaded file and specified 'name' column has duplicate source names, it causes chaos and uninformative errors.
Workaround: Restart the tool, re-upload your file, and don't specify a name. OR, load your file into IRSA Viewer, remove ("hide") the offending name column, save it as a tbl file without the name column, and then upload that modified file into the SPHEREx Data Explorer.

Known Minor Bugs

Search panel goes transluscent When you move your mouse focus away from the search panel, it goes transluscent to make it easier to see the background image. Sometimes, when you move your mouse back into the search panel, it doesn't become opaque right away.

Workaround: Click again in the search panel.

Posted: 5 December 2025
Unphysical values for parameters can be passed
For single source spectrophotometry, there is some parameter checking, but for multi-object spectrophotometry, there is no parameter checking currently. Unphysical parameters will be accepted and the code will attempt to work with them.

Workaround: Make sure you are passing it valid parameters. Remember that the tool is only fitting flux, so you need to give it good positions and all the necessary Sérsic parameters.
Targets not overlaid.
If you upload a list of targets, if your target list has positions that you specify that do not have column headings that are lower-case "ra" and "dec", there's a bug -- the tool may be unable to overlay the target position on the images.
Workaround: Change your column headings to be "ra" and "dec", and the visualization should work better.
VO catalog results not displayed
For some VO catalog searches, the tool retrieves the catalog but is unable to display it.

Workaround: Download the catalog yourself from another VO access point (see here) and upload it separately to the tool.

Known Idiosyncracies

'Lost' Downloads
Depending on how your browser is configured, when you download data, your browser may save it in a counter-intuitive place. Under Safari (on the Mac), for small downloads, it might not appear to have done anything at all. It actually has; it's trying to be helpful by doing it quickly. To find where your browser has put your data, (a) search on your local disk for recently modified files; (b) look in the preferences for the browser configuration option of where to put downloaded files; or (c) look for a directory called "Downloads" or something similar.

Regions files saved as overlays from large catalogs do not include all the sources.
If the file has >1,000 sources, it will show catalogs hierarchically -- that is, the tool will bin up the catalogs based on HEALPix pixels, and it will show symbols with a number indicating the number of sources in that region. Then, when you zoom in, it will dynamically adapt to show you smaller and smaller cells until it shows you all the individual sources. Saving the regions file at will not save all the sources.

Workaround: Save the catalog from the table as a regions file; that will save the entire catalog, but not any of the other overlays from the image.

Strange behavior when requesting catalogs
Our catalog search options depend on other services, often other IRSA services, to search and display the catalogs. Sometimes these other services are down for maintenance or may be experiencing unusually heavy loads. On those occasions, the user may get an "Unresponsive script" error or a "Call failed on the server" error. Please do let us know if you continue to get these errors; we may not be aware of an overloaded or down server.

Slow behavior when requesting large tables
When rendering long tables, the table can take several seconds to finish loading. For example, if you conduct a search returning 1000 entries, and change the page size from 50 to 2000, the browser will struggle for a few seconds trying to load the table. A message appears to let you know that that is what is going on.