In Mode 1 (left - SPIRou only), the f8 beam from the telescope enters SPIRou without any additional optics. In Mode 2 (middle - ESPaDOnS only), a pair of mirrors and additional doublets direct the f8 beam to ESPaDOnS. In mode 3 (right - both instruments), dichroics split the light from the f8 beam into separate wavelength domains and feed both instruments. Illustration credit: CFHT/IRAP

VISION is an optomechanical interface that will allow installing SPIRou and ESPaDOnS at the Cassegrain focus of the telescope at the same time. Thanks to a beam splitter that will separate the near IR and optical beams without affecting polarization, and relay optics, VISION can be used to observe with SPIRou only (no intervening optics), or with ESPaDOnS only (using relay optics), or with both SPIRou and ESPaDOnS simultaneously (using the beam splitter).

This project is led by OMP/IRAP in collaboration with CFHT. Funding is provided by CFHT and CNRS in France. The current project timeline has all 3 VISION modes integrated at CFHT (hardware assembled and tested, software integrated, including QSO and Kealahou) at the end of 2024 or early 2025. Kealahou will be modified to allow requesting (Phase 1) and entering observations (Phase 2) for Modes 1 and 2 first, then for Mode 3.


The advantages include: 

  1. Simplified operations: there will no longer be a need for the technical staff to spend time removing one instrument to install the other.
  2. Maximum mechanical stability, which is essential for reaching optimal RV precision with SPIRou
  3. In the single-instrument mode, allow observations with one of the instruments for 1 night during a run for the other instrument (e.g. 1 night of SPIRou observations to catch an exoplanet transit in the middle of a dedicated ESPaDOnS run).  
  4. In the single-instrument mode, extend the sample period of targets that need to be monitored with one instrument or the other, by switching instrument at the appropriate time of the night, over a long VISION run.
  5. In the simultaneous observations mode, acquire data from 370 nm to 2.44 microns, thus expanding the scientific capabilities of both instruments.


VISION guarantees the polarimetric and throughput performances of both instruments over >99% of their spectral domain, without altering the present optical setup or focus.

The first main constraint for VISION is that the polarimetric capabilities of both instruments are preserved.

The geometry of the reflections of either the mirrors (mode 2) or the dichroics (mode 3) is chosen to compensate for any polarization introduced by the oblique reflections at constant incident angles. The beam aperture, aligning issues, or material imperfections will corrupt this ideal polarization-free scheme and degrade the polarimetric capabilities. We thus want to ensure that, in the non-ideal case of a f/8 Cass beam aperture and slight implementation imperfections, the polarimetric properties of the incoming light are not degraded beyond a given threshold, that we take as the ultimate photon noise level in either SPIRou or ESPaDOnS spectra, i.e., 0.01% of the unpolarized continuum. This sets the maximum acceptable level of polarimetric contamination / crosstalk induced by VISION:

  • from Stokes I (unpolarized) to Stokes Q, U and V (linearly and circularly polarized light) spectra: 0.01%;
  • from Stokes Q and U to Stokes V spectra (and vice versa): 1% (as typical Stokes QUV signatures in spectral lines very rarely exceed amplitudes of 1%);
  • from Stokes Q to U (or vice versa), crosstalks are not an issue if within 10%, and can be compensated for by a small rotation of the instrument and / or through the observation of standard stars.

The second main constraint for VISION is that the throughput of both instruments is not degraded by more than ~10% throughout the spectral domain, except in the cutoff region of the dichroic plates in mode 3 (900-950 nm) where throughput will obviously be affected.

The third main constraint specifies the cutoff region of the dichroic plates in mode 3. We set the spectral region between 900 and 950 nm (a region of limited astrophysical interest for SPIRou and ESPaDOnS science) as the wavelength interval where transition from transmission to reflection happens. In this spectral range, throughput progressively ranges from 0 to 1 for SPIRou (transmitted light), and from 1 to 0 for ESPaDOnS (reflected light). Polarimetric properties of the incoming light are also expected to be degraded beyond our main constraint in this region.