Observing with MIDI

1
Observing with MIDI

• Christian Hummel (ESO)

2
Mid-Infrared Instrument
3
MIDI schematic drawing
4
Mid-infrared background
5
Observing in the mid-infrared
6
MIDI acquisition
-

7
MIDI chopping
8
Photometry
-

9
Spectrum extraction
PA I1 I2 PB
10
Interferometry
11
Wavelength binning
12
HIGH_SENS (high sensitivity)
13
Background cancellation

• The quality of the initial background
  cancellation depends on the splitting ratios
• A high-pass filter needs to be used to remove
  residual background fluctuations

14
Kappa matrix
and so forth
15
Interferograms and PSDs
16
Channel PSDs
17
Good and bad fringes
18
Correlated flux normalization
Max. and min. field amplitudes
Max. and min. intensities
Visibility amplitude
yields
Interferogram in one MIDI channel
Subtracting the two channels
Normalization factor
19
Beam overlap problems
20
Multiply, then mask

• Only the green overlap area contributes to the
  correlated flux
• Therefore, multiply detector pixels first, then
  use common mask (red) to extract

21
MIDI schematic drawing
22
SCI_PHOT (high precision mode)

• Photometry recorded simultaneously with the
  fringe data (must be chopping). Use kappa matrix
  to convert PA,B into I1,2.
• Changes in beam overlap will simultaneously
  affect all extracted fluxes, thus will divide
  out.
• Kappa matrix can be determined from A and B
  photometry (needs only to be done once per night
  on a bright target).
• Otherwise, same reduction as HIGH_SENS

23
SCI_PHOT
24
Optical distortion
PB
I1
25
Reductions
26
Coherent integration

• Integration by co-adding interferograms
• Requires off-line fringe tracking (post
  processing)
• Maintains visibility phase (second derivative)
• Implemented by EWS package (W. Jaffe)
• Results have been tested to be consistent with MIA

27
EWS processing steps

• Compress (extract) spectra for each frame
• Difference BC outputs and apply high-pass
• Multiply by ei2pd/? and sum over scan
• Fourier transform complex visibility as a
  function of wave number into delay space
• Average several scans and find peak
• Apply both instrumental OPD and group delay to
  align phasors before coherent integration

28
Polychromatic fringes

• Compressed spectra (vertical, with color coding)
  as a function of time (OPD). The pattern repeats
  after each scan.

29
Group delay analysis
D Dins Datm
With
30
Demodulation
ei2pDi / ?

0.4s smoothing
0.8s smoothing
31
EWS product amplitude and phase
ei2pDa / ?
ei2pDi / ?


This complex quantity can now be coherently
integrated
32
Planning MIDI observations

• Target flux
• Is it detectable?
• Target complexity
• MIDI FOV is 2
• Paranal VLTI sky coverage
• The UT shadow
• Calibration and calibrators
• Select them carefully

33
User Support Group at ESO
34
Visibility (as in Can I see it)
35
Interferometric field of view (I)
12 Persei observed on Oct 9, 2001 with the CHARA
Array, K-band, 330m baseline, separation 40
marcsec
Mark 3 (Oct 8, 1992)
36
Interferometric field of view (II)
b CrB
(NPOI)
37
Photometric field of view
Mizar A (V2.3) with B (V4.0) at 14 (Mark III)
38
Calibrator selection

• Close to science target
• Measured immediately before and/or after
• Small known diameter
• Similar brightness as science target

39
SearchCal
40
MIDI calibrators

• During phase-2, users select their calibrators
  for their targets.
• ESO recommends to use the CalVin tool on the web
  to select the calibrators.

41
MIDI data reduction software

• MIAEWS-1.6 SCI_PHOT enabled
• ESO Pipeline some SCI_PHOT support
• MyMidiGui 1.6 compatible with MIAEWS
• Meudon package small user base ? Source code not
  available!

42
ESO Pipeline
43
MIAEWS-1.6

• Written by R. Koehler and W. Jaffe
• Does SCI_PHOT
• Coherent and incoherent analysis
• IDL based
• Fine tuned for UTs, but not yet for ATs

44
xmdv GUI
45
Issue w/ATs sky windows
46
Sky window effect for ATs
47
Recent AT operations on 64 m
48
Calibrator visibility (TF)
49
64 m acquisition
50
FOV edge
51
MyMidiGui and MyMidiPipe
52
Photometry with MyMidiGui
53
MyMidiPipe
54
Global calibration
Calibrator V2 vs seeing
Calibrator V2 vs time
Mark III Mozurkewich et al. 1991
55
Dependence of MIDI V2 on seeing
MIDI