The Veloce spectrograph


Veloce will be a stabilised, high-resolution (R~80000) echelle spectrograph. Initially Veloce will have only one of its three possible 'arms', each covering a different wavelength region, the first arm 'Rosso' will cover ~580 - 930nm. Veloce will be fed by a 26 hexagonal fibre integral field unit (19 target, 2 calibration, 5 sky/background), with an on-sky target area with a diameter of 2.5" and with spatial scrambling through octagonal fibres. The optional simultaneous calibration will be a Menlo Systems laser comb for extremely precise wavelength calibration and/or a Thorium-Uranium-Xenon arc lamp. Veloce will be suitable for observing single targets brighter than i<14 magnitude and will particularly good for red targets such as the cool M dwarf stars.


An observing guide will be created for Veloce after commissioning. It is likely Veloce will have a set of calibration template configurations and will accept a user input list of stars, these stars and calibration setups can then be queued up for sequential observation.


Veloce will not be configurable. It will be a fixed format spectrograph with a standard calibration routine run at the start and end of the night. The main choices for the user will be the exposure time, and whether to have one, both or no calibration fibres illuminated during the exposure. An observer that does not require high Doppler precision (<10 m/s) may choose not to have the simultaneous calibration fibres on as they may add some small level of contamination to the stellar spectrum, and the instrument should be intrinsically stable enough for non-precision velocity work (more information will be available on intrinsic stability post commissioning). Most precision Doppler velocity users are expected to have only the laser comb fibre active during target observations, as the Thorium-Uranium-Xenon lamp will have strong lines in the longer Veloce wavelengths that will contaminate the spectrum.

Changing exposure times during an observation is not recommended when using simulataneous calibration. Since the flux in the calibration is configured to be spread over the whole exposure, ending an exposure early will mean less flux from the calibration than expected, resulting in a poor calibration. Making the exposure longer than planned will mean the calibration will not sample the exposure evenly resulting in a incorrect estimate for the flux-weighted mean exposure time.

Exposure times and observation planning

The best photometric filters for working with Veloce will be r or i.

The exposure times are unknown until the instrument is commissioned, though it is planned that we will be able to reach a peak S/N > 20 for a i~14 star in a reasonable exposure time e.g. <30min. It would be reasonable to plan for precise Doppler velocities (<5 m/s) for stars with i<12 with exposure times of ~15 min per target (typically two consecutive observations) with a 2 minute overhead for readout and slewing. 15 minute exposures are recommended for precise Doppler velocities to help smooth out the velocity effects of short period stellar oscillations. Multiple exposures are recommended to help mitigate the effects of cosmic rays which are enhanced in deep depletion CCD Cameras such as Veloce will have.

More information is available on the Veloce proposal webpage.

Data reduction

The data reduction and calibration is planned to be automatic, with a spectrum and an estimate of the Doppler velocity available a few minutes after the observation (based on a user selected template cross-correlation).

For more information please contact Duncan Wright (