This page gives an overview of the image quality as derived from observations with VLT instruments. This information helps to assess how good the delivered image quality in the data is. Ideally it is limited by the seeing conditions during the time of observation, or, for instruments with AO, by the performance of the AO.

Most of the IQ monitoring was done by processing science frames. With the rescoping of the QC group processes, science processing was terminated on Sep. 30, 2011, and so was the IQ trending. Still, the historical information could be useful.

For a discussion of differences between the seeing derived from VLT instruments and the seeing measured by the DIMM monitor see Sarazin et al., 2008, Messenger 132, p. 11. More information about the seeing on Paranal can be found here.

top GENERAL INFORMATION

Most of the VLT instruments, which can observe imaging data, also measure the size and shape of the sources observed. They do so, however, in slightly different ways and also collect different ambient parameters. Therefore anyone who wants to compare data observed with different instruments should first read the descriptions of the measurements. The links are provided in the table above, which also provides some guidelines about the wavelength coverage of the instruments. A comparison of the image quality information collected by FORS1/2, ISAAC, and VIMOS can be found here. Also when comparing data obtained in different wavelength bands keep in mind that image quality data from different wavelengths relate roughly like

IQ₁ ⋅ λ₁ = IQ₂ ⋅ λ₂

All instruments provide in addition to the image quality derived from the instrument's data also that from the DIMM (IQ_DIMM) and in some cases also the image quality from the Shack-Hartmann sensor of the Active Optics (IQ_Tel). If present this should be used rather than the DIMM seeing to compare to values obtained from the VLT observations. One should be careful to check the airmass corrections done to any of these data. Both the image quality from the Shack-Hartmann sensor and the DIMM seeing are corrected to zenith before storing them in the frame headers. Some instruments correct these values then to the airmass and wavelength of the corresponding instruments observations (FORS1/2, VIMOS, CRIRES), while other don't (HAWK-I). The standard equation to convert between DIMM values (which are corrected to zenith and observed at λ_D ≅ 0.5μm) and instrument values (which are assumed not to be corrected to zenith) is the one from Marc Sarazin:

FWHM_ins ⋅ λ_inst^1/5 ≅ FWHM_D ⋅ λ_D^1/5 ⋅ airmass^3/5 ⋅ k
with
k = √(1-78.08 ⋅ λ_D^2/5 ⋅ airmass^-3/5 ⋅ FWHM_D^-1/3)

• The field-of-view of VIMOS (16' by 18') is about 4 times larger than that of the FORSes (6.6' by 6.8' in standard resolution, 3.4' by 3.4' in high resolution). Thus the degradation of the optical quality with increasing distance from the optical axis of the VLT plays a larger role for VIMOS than for FORS1/2. In addition each quadrant of VIMOS has its own optics, which introduces additional distortions, this time centered on the center of the quadrant.
• VIMOS does not have an Atmospheric Dispersion Corrector. While VIMOS data are usually observed at low airmass, there is a correlation between ellipticity and image quality on one hand and airmass on the other hand (with image quality decreasing with increasing airmass).
   

When trying to compare data taken with ISAAC and HAWK-I one should keep in mind that these two instruments have very different field-of-views. ISAAC has 2.5' by 2.5 with 0.148"/pixel, which may be reduced to 1.2' by 1.2' above 3μm to have a better sampling. HAWK-I on the other hand covers a total area of 7.5' by 7.5 with 0.106"/pixel, i.e. it has both a better sampling and a larger area, but observes only up to 2.5 μm. Another major difference, which is important for long-term image quality studies, is their time of operations: ISAAC started operations on April 1, 1999, while HAWK-I started operations on April 29, 2008.

In both cases the image quality is determined by the pipeline. ISAAC applies a filtering for extended objects, which would bias the image quality and ellipticiy to larger values. HAWK-I applies a filter for circularity.

VIRCAM is the imaging instrument at VISTA, the infrared survey telescope, and has started operations on April 1, 2010. It has the largest field-of-view among the near-infrared imagers (16 x 11.6'x11.6', with significant gaps between the individual detectors), in combination with the coarsest sampling (0.34"/pixel). Here the FWHM averaged over all sources on a given detector is used as image quality.

More details can be found in the description links listed above.

If one wants to compare the image quality from AO instruments to that of non-AO instruments one should select data taken with open-loop AO.

Both SINFONI and CRIRES derive the image quality from just one object, so no statistical filtering is applied. As CRIRES is a spectroscopic instrument the image quality is derived indirectly by measuring the spatial FWHM of the reduced spectrum and multiplying this value with the image scale of 0.086"/pixel. At its lowest resolution SINFONI has a field-of-view of 8"x8" and the image quality is derived from observations of telluric standard stars or PSF stars. For both instruments observations with open loop AO and with closed loop AO are analysed and monitored.

NACO monitors only the quality of the AO correction (the Strehl ratio) of the standard stars used for zeropoints and not the actual image quality. As this instrument never observes without AO correction, its image quality cannot be compared to non-AO instruments.