M I D I  MID-infrared Instrument for the VLT Interferometer
	HOME INDEX SEARCH HELP NEWS

Note: the aim of this page is to provide general information about the MIDI project and its expected features and performances. To know about the actual state of MIDI for observations, please check the "MIDI page for Astronomers"

	MIDI at Paranal (click for more images)
Instrument Information Scientific Objectives and GTO Instrument Characteristics General design Detector Optics Scanning modes Chopping modes Sensitivity Estimates References The MIDI page for Astronomers

Long-baseline direct interferometry in the mid-infrared (N- and Q-bands, i.e., 8 to 13 µm, and 13 to 26 µm) is a relatively unexplored observational technique. MIDI is the first instrument to allow observations with large apertures and hectometric baselines at such wavelengths. The high-angular resolution and unique sensitivity provided by MIDI  will permit to carry out novel observations in a large number of areas. To give an idea, a few topics extracted from the MIDI GTO program are listed here:

• Dust Tori in nearby AGN
• Inner disks around stars (low-mass YSO, intermediate mass YSO, Vega-type)
• Massive YSO
• Dusty environment of Hot Stars
• Cool Late-type stars
• Extrasolar planets and brown dwarfs

1. General design

MIDI is built around a Mach-Zender type (half-reflecting plate) 2-beam optical recombiner. It uses as light-collectors either two VLT unit-telescopes (UTs), or two VLT auxiliary-telescopes (ATs), and the whole VLTI infrastructure (delay-lines, M16, switchyard, beam-compressors, fringe tracker). Operation with the ATs is foreseen for end-2003. The delay lines are equipped with Variable Curvature Mirrors (VCM), that relay the pupil image correcting for the distance of the delay line carriage. The UTs are normally equipped with MACAO high-order Adaptive Optics units. Although the correction of turbulence does not provide a large again on 8m apertures at the wavelengths of MIDI under good seeing conditions, in practice MACAO is needed for the operation of the fringe tracker FINITO.

*: 4 UT telescopes are being progressively equipped with MACAO units, first two available from end of 2003.
**: 4 AT telescopes are being progressively installed, first two available in the first part of 2004.

2. Detector

MIDI has a mid-infrared detector which has the following characteristics (given performances have been measured by MIDI consortium):

The detector control software of MIDI is based on GEIRS, the generic infrared detector software developed at the MPIA (Max-Planck Institut für Astronomie). Data from the detector are on-line processed by a MIDI-specific piece of software called NRTS (Near Real-Time Software) which runs DQA (Data Quality Analysis) modules. DQA modules exist for signal-to-noise image evaluation, fringe search and tracking, etc.. The NRTS of MIDI is based on the common Instrument VLT Software.

3. Optics

Due to thermal radiation from the environment, most of the optics is enclosed in a cryostat cooled at 40 K (actually, 35 K can be reached thanks to the cold-head of the helium closed-cycle cooler of MIDI). The array detector of MIDI is cooled at 10 K.Thermal radiation from background and stray-light are reduced thanks to pupil reimaging (by the Variable Curvature Mirrors of the VLTI Delay-Lines) and pupil-stops inside the cryostat. The beam-combination is done close to the pupil plane while the signal is detected in an image plane. The light path within MIDI can be summarized as follows (figure below): the collimated beams from the telescopes, reduced to 18 mm by the VLTI Beam-Compressor. To generate interferograms, the difference of optical path length between the beams is time-modulated by dihedral reflectors mounted on piezoelectric transducers (PZTs).

The optics of MIDI is not fully reflective (gold is used as coating for mirrors of MIDI because of its high reflection factor in the infrared), but also refractive. The refractive components of MIDI are made of zinc-selenide (ZnSe), a material which has the property to be transparent in N-band, but has a high refraction index (2.39 at 10 µm and T=50 K), requiring anti-reflection coatings whenever this is possible,

After passing through the pupil-stops, the beams are focused on the field-stops. These stops can be pinholes for spatial filtering, or slits (if spectroscopic mode is used). Full-field imaging is also possible. After re-collimation, part of the beam can be sampled by 30/70 ZnSe beamsplitters to obtain the photometry of the source (if this mode has been selected). The remaining part of the beams are combined thanks to a 50/50 ZnSe plate (acting also as a chromatism-compensation plate). The two interferometric beams (and optionally the two photometric beams) are spectally filtered, dispersed by a grism or a prsim (if spectroscopic mode has been selected), and focused by the "camera" element (consisiting of Ge-coated ZnSe lenses) on the detector plane.

4. Scanning modes

There are several possibilities for MIDI to acquire and process "interferograms" (fringe pattern images), in order to compute visibility. All modes (see below) can be used in wide-band (escept group-delay tracking) or spectroscopic mode (dispersion of the recombined and photometric beams). In this case, a visibility modulus can be measured for each spectral channel (number and width of spectral channels are determined from dispersive element resolution and from binning used). Note that these scanning modes are being progressively tested and commissioned, and not all might be available initially.

• "Fourier" mode:  the most common mode, in which OPD (optical path difference) is scanned over a large range (about 100 µm) by one of the piezoelectric translation stages (PZTs). The interferometric signals are sampled with a few (3 to 5) points per fringe. The duration of a scan can be longer than the atmospheric coherence time at 10 µm. The zero-OPD position can in theory be measured for each scan and sent to the VLTI  OPD-controller to correct the delay-line position (internal coherencing fringe-tracking). A raw visibility is computed from each scan. Some scans can be rejected from computation if no fringes are detected within.
• "ABCD" mode: 4 samples of the fringe of largest amplitude (corresponding to OPD=0) are taken for each scan. The duration of a scan must be very short compared to the atmospheric coherence time at 10 µm. Each sample can be taken with the PZTs either moving or fixed. This mode normally requires accurate external fringe-tracking to "freeze" the fringe, though the OPD can be computed from each scan.
• "Speckle" mode: similar to "Fourier" mode, except that the signal is too low to detect fringe and to measure OPD in a scan. Scans are acquired in "blind" mode. Power spectra of the scans are integrated to yield the information that will be used to compute raw visibility.
• "Coherent-integration" mode: similar to "Fourier: mode, except that the OPD is retrieved from the external fringe-tracker with an accuracy allowing to integrate scans after piston-correction. The raw visibility is computed from this integrated "like frozen" scan.
• "Group-delay tracking" mode: Used with dispersion only. No PZT are used. The OPD is computed from the integrated power spectrum of a series of detector frames acquired.

5. Chopping modes

MIDI has different techniques to measure thermal background, in order to later remove the background contribution for fringe visibility calculation.
Note that these chopping modes are being progressively tested and commissioned, and not all might be available initially.

• Virtual chopping: the signal from the star is sampled by a pinhole, while the signal from the background is sampled by another pinhole within the field-of-view of VLTI. Optionally, the location of the star in the FOV can be moved after a few minutes to swap the target of each pinhole (star or background). No fringes are acquired.
• Staring interferometry: fringes are acquired for a few minutes, then chopped-and-nodded photometry from the star is acquired for a few minutes. Then, fringes are searched and acquired, etc...
• Chopping interferometry: The following pattern is repeated: fringe finding, then fringe acquisition for one coherence time, then background measurement by chopping for one coherence time. This mode is normally used with spatial filtering.

The sensitivity achieved by MIDI is under measurement (commissioning task). The following table lists the sensitivity estimates given at Final Design Review of MIDI. The sensitivity depends on the telescope diameter and on the fringe-tracking mode used. The actual sensitivity achieved will depend on several factors, primarily the background level and its fluctuations.

On the Web:

• MIDI web-site at MPIA
• MIDI web-site at NFRA/ASTRON
• ESO Press-Release on MIDI first fringes (18 December 2002)
• ESO Very Large Telescope Interferometer

• C. Leinert, U. Graser, L.B.F.M. Waters, G. Perrin, W. Jaffe, B. Lopez, F. Przygodda, O. Chesneau P. Schuller, A. Glazenborg, B. Grimm, W. Laun, S. Ligori, J. Meisner, K. Wagner, E. Bakker, B. Cotton, J. de Jong, R. Mathar, U. Neumann, C. Storz, "The 10 micron instrument MIDI - getting ready for observations on the VLTI", Proceedings of SPIE, in press (2002)
• M. Nafati, G. Perrin, V. Coudé Du Foresto, G. Chagnon, "Principles of the data reduction software of the VLTI mid-infrared instrument MIDI", Proceedings of ASP Conference, 266, p 124 (2002).
• S. Wolf, T. Henning, G. D'Angelo, "Detecting Gaps in Protoplanetary Disks with MIDI at the VLTI",  Proceedings of ESO workshop "The Origins of Stars and Planets: The VLT View", p 325 (2002)
• L.B.F.M. Waters, C. Leinert, U. Graser, G. Perrin, B. Lopez, W. Jaffe, "The Scientific Potential of MIDI in the 20 Micron Window", Proceedings of ESO workshop "Scientific drivers for ESO Future VLT/VLTI Instrumentation", p 314 (2002)
• I. Pascucci, T. Henning, B. Stecklum, M. Feldt, C. Leinert, "VLTI observations of ultra-compact H II regions - simulations of MIDI visibilities", Proceedings of 36th Liège International Astrophysics Colloquium, pp 145-149 (2001)
• J. Meisner, "Fringe tracking and group delay tracking methods for MIDI", Proceedings of 36th Liège In ternational Astrophysics Colloquium, pp 225-231 (2001)
• C. Leinert, U. Graser, L.B.F.M. Waters, G. Perrin, B. Lopez, A. Glazenborg-Kluttig, J.C. de Haas, T.M. Herbst, W. Jaffe, P.J. Lena, R. Lenzen, R.S. le Poole, S. Ligori, O. von der Lühe, R. Mundt, J.W. Pel, I. L. Porro, "10-µm interferometry on the VLTI with the MIDI instrument: a preview", Proceedings of SPIE, 4006-05, pp 43-53 (2000)
• B. Lopez, C. Leinert, U. Graser, L.B.F.M.Waters, G. Perrin, T. Herbst, H. Rottgering, D. Rouan, B. Stecklum, R. Mundt, H. Zinnecker, P. de Laverny, M. Feldt, J. Meisner, A. Dutrey, T. Henning, F. Vakili, "The Astrophysical Potentials of the MIDI VLTI Instrument", Proceedings of SPIE, 4006-06, pp 54-67 (2000)
• S. Hippler, W. Jaffe, R. Mathar, C. Storz, K. Wagner, W.D. Cotton, G. Perrin, M. Feldt,, "MIDI: controlling a two 8-m telescope Michelson interferometer for the thermal infrared", Proceedings of SPIE, 4006-06, pp 92-98 (2000)
• S. Ligori, U. Graser, B. Grimm, R. Klein, "Design and tests of the MIDI detector subsystem", Proceedings of SPIE, 4006-06, pp 136-146 (2000)
• J.W. Pel, A.W. Glazenborg-Kluttig, J.C. de Haas, H. Hanenburg, R. Lenzen, "Cold optics of MIDI: the mid-infrared interferometric instrument for the VLTI", Proceedings of SPIE, 4006-06, pp 164-173 (2000)
• R.R. Rohloff, W. Laun, U. Graser, N. Ortlieb, M. Lebong, "Cryo design for the VLTI MIDI instrument", Proceedings of SPIE, 4006-06, pp 277-288 (2000)
• I.L. Porro, U. Graser, U. C. Leinert, B. Lopez, O. von der Lühe, "Estimated Performance for 10-micron Interferometry at the VLTI with the MIDI Instrument", Proceedings of ASP Conference, 194, p 325 (1999)
• C. Leinert, U. Graser, "MIDI - the Mid-infrared interferometric instrument for the VLTI", Proceedings of SPIE,  3350, pp 389-393 (1998)

Send comments to <smorel@eso.org> Last modified: Mon Mar 31 21:47:00 CET 2003