# How does the ESPRESSO Echelle spectrograph fold the spectrum so nicely?

The CBC News Technology and Science Q&A This ESPRESSO machine doesn't make coffee but scans the skies for habitable planets includes a nice description of ESPRESSO (Echelle SPectrograph for Rocky Exoplanet and Stable Spectroscopic Observations) and links to the YouTube video Echelle SPectrograph for Rocky Exoplanet and Stable Spectroscopic Observations.

There's something really amazing about the way that the spectrogram has been folded so nicely on to this image. Is this what it looks like on a 2D detector? (sans colour of course)

I see what looks like eight distinct sections, and a really nice spacing between adjacent lines otherwise. How is this accomplished?

Click or open the image in a new tab for full size view (it's worth it; 1.8 MB, 1738x2530)

above: Frame grab from the video at about 00:55. Recorded with Python using this script, then extracting frames using this script in Blender. Click or open in new view for full size - it's worth it.

above: What an Echelle spectrogram "normally" looks like (again with added color). From the really interesting University of Virginia ASTR 3130 (Majewski) Lecture Notes.

• Not sure what you are asking - how to produce an echellogram or why Espresso uses 8 detectors? Dec 8, 2017 at 7:15
• @RobJeffries I'm familliar with how "vanilla" cross-dispersion Echelle spectrographs work. If this really a mosaic of multiple detectors, then that information, along with a description of how the light is actually split among those detectors without significant loss of information would be a basis for an excellent answer. Are the bare silicon die just butted against each other, or are there kinfe-edge mirrors or prisms that cut the beam into sections perhaps? Thanks!
– uhoh
Dec 8, 2017 at 7:49
• @RobJeffries Espresso doesn't use 8 detectors. It uses two 6kx6k CCDs. To answer your question: Espresso uses a dichroic to split the light at about 515nm on two seperate cameras / detectors. For a detailed answer see their optical design paper: researchgate.net/profile/Denis_Megevand/publication/… Dec 8, 2017 at 15:03
• No, the vertical gap (and also the horizontal gaps) is an artifact from how they readout the detector / how the overscan region of the CCD looks like. The real gap between the detectors is horizontally. They just stacked the two images on top of each other. Dec 8, 2017 at 15:11
• @JulianS. Could you construct a short answer? Feb 3, 2018 at 21:45