Transcript The Imaging Chain for Optical Astronomy

The Imaging Chain
in Optical Astronomy
Review and Overview
“Imaging Chain” includes these elements:
1.
2.
3.
4.
5.
6.
7.
8.
energy source
object
collector
detector (or sensor)
processor
display
analysis
storage (if any)
Optical Imaging Chain
1: source
5: processing
2: object
3: collector
4: sensor
6: display
7: analysis
Source and/or Object
• In astronomy, the source of energy (1) and the object
(2) are almost always one and the same!
• i.e., The object emits the light
– Examples:
• Galaxies
• Stars
– Exceptions:
• Planets and the moon
• Dust and gas that reflects or absorbs starlight
Optical Imaging Chain in Astronomy
5: processing
1: source
2: object
or
6: display
7: analysis
3: collector
4: sensor
8: storage
Optical Imaging Chain in Radio
Astronomy
1,2
3,4
radio waves
receiver where
waves are collected
waves
converted into
electro signals
5
computer
received as signal
6,7
Specific Requirements for
Astronomical Imaging Systems
• Requirements always conflict
– Always want more than you can have
must “trade off” desirable attributes
 Deciding the relative merits is a difficult task

“general-purpose” instruments (cameras) may not be sufficient
• Want simultaneously to have:
– excellent angular resolution AND wide field of view
– high sensitivity AND wide dynamic range
• Dynamic range is the ability to image “bright” and “faint” sources
– broad wavelength coverage AND ability to measure narrow
spectral lines
Angular Resolution
vs. Field of View
• Angular Resolution: ability to distinguish sources that
are separated by small angles
– Limited by:
• Optical Diffraction
• Sensor Resolution
• Field of View: angular size of the image field
– Limited by:
• Optics
• Sensor Size (area)
Sensitivity vs. Dynamic Range
• Sensitivity
– ability to measure faint brightnesses
• Dynamic Range
– ability to image “bright” and “faint” sources in same system
Wavelength Coverage
vs. Spectral Resolution
• Wavelength Coverage
– Ability to image over a wide range of wavelengths
– Limited by:
• Spectral Transmission of Optics (Glass cuts off UV, far IR)
• Spectral Resolution
– Ability to detect and measure narrow spectral lines
– Limited by:
• “Spectrometer” Resolution (number of lines in diffraction grating)
Optical Collector (Link #3)
Optical Collection (Link #3):
Refracting Telescopes
• Lenses collect light
• BIG disadvantages
– Chromatic Aberrations (due to dispersion of glass)
– Lenses are HEAVY and supported only on periphery
• Limits the Lens Diameter
• Largest is 40" at Yerkes Observatory, Wisconsin
http://astro.uchicago.edu/vtour/40inch/kyle3.jpg
Optical Collection (Link #3):
Reflecting Telescopes
• Mirrors collect light
• Chromatic Aberrations eliminated
• Fabrication techniques continue to improve
• Mirrors may be supported from behind
 Mirrors may be made much larger than
refractive lenses
Optical Reflecting Telescopes
• Concave parabolic
primary mirror to
collect light from
source
– modern mirrors for
large telescopes are
thin, lightweight &
deformable, to
optimize image
quality
3.5 meter
WIYN
telescope
mirror, Kitt
Peak, Arizona
Thin and Light (Weight) Mirrors
• Light weight Easier to point
– “light-duty” mechanical systems  cheaper
• Thin Glass  Less “Thermal Mass”
– Reaches Equilibrium (“cools down” to ambient
temperature) quicker
http://www.cmog.org/page.cfm?page=374
Hale 200" Telescope
Palomar Mountain, CA
http://www.astro.caltech.edu/observatories/palomar/overview.html
200" mirror (5 meters)
for Hale Telescope
•
•
•
•
•
•
•
Monolith (one piece)
Several feet thick
10 months to cool
7.5 years to grind
Mirror weighs 20 tons
Telescope weighs 400 tons
“Equatorial” Mount
– follows sky with one motion
Keck telescopes, Mauna Kea, HI
http://www2.keck.hawaii.edu/geninfo/about.html
400" mirror (10 meters)
for Keck Telescope
• 36 segments
• 3" thick
• Each segment weighs 400 kg (880 pounds)
– Total weight of mirror is 14,400 kg (< 15 tons)
• Telescope weighs 270 tons
• “Alt-azimuth” mount (left-right, up-down
motion)
– follows sky with two motions + rotation
Basic Designs of Optical
Reflecting Telescopes
1. Prime focus: light focused by primary mirror alone
2. Newtonian: use flat, diagonal secondary mirror to
deflect light out side of tube
3. Cassegrain: use convex secondary mirror to reflect
light back through hole in primary
4. Nasmyth (or Coudé) focus (coudé  French for
“bend” or “elbow”): uses a tertiary mirror to
redirect light to external instruments (e.g., a
spectrograph)
Prime Focus
Sensor
f
Mirror diameter must be large to ensure that
obstruction is not significant
Newtonian Reflector
Sensor
Cassegrain Telescope
Sensor
Secondary
Convex Mirror
Feature of Cassegrain Telescope
• Long Focal Length in
Short Tube
f
Location of
Equivalent Thin Lens
Coudé or Nasmyth Telescope
Sensor
Optical Reflecting Telescopes
Schematic
of 10-meter
Keck
telescope
(segmented
mirror)
Large Optical Telescopes
Telescopes with largest diameters
(in use or under construction:
– 10-meter Keck (Mauna Kea, Hawaii)
– 8-meter Subaru (Mauna Kea)
– 8-meter Gemini (twin telescopes:
Mauna Kea & Cerro Pachon, Chile)
– 6.5-meter Mt. Hopkins (Arizona)
– 5-meter Mt. Palomar (California)
– 4-meter NOAO (Kitt Peak, AZ &
Cerro Tololo, Chile)
Keck
telescope
mirror
(note
person)
http://seds.lpl.arizona.edu/billa/bigeyes.html
Summit of Mauna Kea, with Maui in background
Why Build Large Telescopes?
1. Larger Aperture  Gathers MORE Light
–
–
Light-Gathering Power  Area
Area of Circular Aperture = D2 / 4  D2
•
D = diameter of primary collecting element
2. Larger aperture  better angular resolution
–
recall that:


D
Why Build Small Telescopes?
1. Smaller aperture  collects less light
•
 less chance of saturation (“overexposure”)
on bright sources
2. Smaller aperture  larger field of view
(generally)
–
Determined by “F ratio” or “F#”
f
F#
D
f = focal length of collecting element
D = diameter of aperture
F Ratio: F#
• F# describes the ability of the optic to
“deflect” or “focus” light
– Smaller F#  optic “deflects” light more than
system with larger F#
Small F#
Large F#
F# of Large Telescopes
• Hale 200" on Palomar: f/3.3
– focal length of primary mirror is:
3.3  200" = 660" = 55'  16.8 m
– Dome must be large enough to enclose
• Keck 10-m on Mauna Kea: f/1.75
– focal length of primary mirror is:
1.75  10m = 17.5 m  58 m
F Ratio: F#
• Two reflecting telescopes with different F#
and same detector have different “Fields of
View”:
large 
Small F#
small 
Large F#
Sensors (Link #4)
Astronomical Cameras
Usually Include:
1. Spectral Filters
–
–
most experiments require specific wavelength
range(s)
broad-band or narrow-band
2. “Reimaging” Optics
–
enlarge or reduce image formed by primary collecting
element
3. Light-Sensitive Detector: Sensor
Astronomical Sensors
• Most common detectors:
– Human Eye
– Photographic Emulsion
• film
• plates
– Electronic Sensors
• CCDs
Angular Resolution
• Fundamental Limit due to Diffraction in
“Optical Collector” (Link #3)


D
• But Also Limited by Resolution of Sensor!
Charge-Coupled Devices (CCDs)
• Standard light detection medium for BOTH professional
and amateur astronomical imaging systems
– Significant decrease in price
• numerous advantages over film:
– high quantum efficiency (QE)
• meaning most of the photons incident on CCD are “counted”
– linear response
• measured signal is proportional to number of photons collected
– fast processing turnaround (CCD readout speeds ~1 sec)
• NO development of emulsion!
– regular grid of sensor elements (pixels)
• as opposed to random distribution of AgX grains
– image delivered in computer-ready form
CCD Basics
• Light-sensitive electronic element based on
crystalline silicon
– crystal = “lattice” of atoms at regular spacings
– acts as though electrons have two states:
• “bound” to atom
• “free” to roam through lattice
CCD Basics
• Incident photon adds energy to electron to
“kick” it up into the “free” states
– energy of photon must be sufficiently large for
electron to “reach” the free states
– to be absorbed, the photon wavelength  must be less
than maximum max  1100 nm (near infrared)
Energy
Electrons in “Free” States
(“conduction band”)
Electrons in “Bound” States
(“valance band”)
photon
CCD Basics
• Silicon structure is divided into pixels
– e- transferred and “counted” one pixel at a time
http://www.byte.com/art/9510/img/505099d2.htm
Sensor Resolution
• Obvious for Electronic Sensors (e.g., CCDs)
• Elements have finite size
• Light is summed over area
of sensor element (“integrated”)
• Light from two stars that falls on
same element is added together
• stars cannot be distinguished
in image!
x
Same Effect in Photographic
Emulsions
• More difficult to quantify
• Light-sensitive “grains” of silver
halide in the emulsion
• Placed “randomly” in emulsion
• “Random” sizes
• “large” grains are more sensitive
• (respond to few photons)
• “small” grains produce better
resolution
Photographic techniques:
silver halide
• Film
– Emulsion on “flexible” substrate
– Still used by amateurs using sensitive film
• B&W and color
• Special treatment to increase sensitivity
• Photographic Plates
– Emulsion on glass plates
– Most common detector from earliest development of
AgX techniques until CCDs in late 70’s
Eye as Astronomical Detector
• Eye includes its own lens
– focuses light on retina ( “sensor”)
• When used with a telescope, must add yet another
lens
– redirect rays from primary optic
– make them parallel (“collimated”)
• rays appear to come from “infinity” (infinite distance away)
– reimaging is performed by “eyepiece”
Eye with Telescope
Without Eyepiece
With Eyepiece
Light entering eye
is “collimated”
Eye as Astronomical Detector
• Point sources (stars) appear brighter to eye through
telescope
2
D
• Factor is
2
P
– D is telescope diameter
– P is diameter of eye pupil
– Magnification should make light fill the eye pupil (“exit pupil”)
• Extended sources (for example, nebulae) do not appear
brighter through a telescope
– Gain in light gathering power exactly compensated by image
magnification, spreads light out over larger angle.
Atmospheric Effects on Image
• Large role in ground-based optical astronomy
– scintillation modifies source angular size
• twinkling of stars = “smearing” of point sources
– extinction reduces light intensity
• atmosphere scatters a small amount of light, especially at short
(bluer) wavelengths
• water vapor blocks specific wavelengths, especially near-IR
– scattered light produces interfering “background”
• astronomical images are never limited to light from source
alone; always include “source” + “background sky”
• “light pollution” worsens sky background
Scattering
• “Wavelength Dependent”
– Depends on color of light
– Long wavelengths are scattered “less”
Scattering by Molecules
"Rayleigh Scattering" 
1

4
• Molecules are SMALL
• “Blue” light is scattered MUCH more than
red light
– Reason for BOTH
• blue sky (blue light scattered from sun in all
directions)
• red sunset (blue light is scattered out of the sun’s
direct rays)
Scattering by Dust
"Mie Scattering" 
1

• Dust particles are MUCH larger than
molecules
– e.g., from volcanos, dust storms
• Blue light is scattered by dust “somewhat
more” than red light
Link #5: Image Processing
Link #5: Image Processing
• Formerly: performed in darkroom
– e.g., David Malin’s “Unsharp Masking”
• Subtract a blurred copy from a “sharp” positive
• (or, add a blurred negative to a “sharp” positive)
• Now performed in computers, e.g.,
–
–
–
–
contrast enhancement
“sharpening”
“normalization” (background division)
…
Example of Unsharp Masking
http://www.hawastsoc.org/messier/fslide53.html
Unprocessed
http://www.seds.org/messier/m/m042.html
After Unsharp Masking
n.b., Increased visibility of fine structure in bright and dark regions
of “cloud” after unsharp masking
Blurring vs. Sharpening
• Blurring:
– Local “Averaging” of Pixels in Scene
– “Averages out” fine detail in image more than
large-scale structure
• Sharpening:
– “Inverse” of Blurring  Local “Differencing”
of Pixels
Image Processing to Correct for:
• Atmosphere (to extent possible)
– e.g., images obtained of object at different “heights” in
sky exhibit different atmospheric “extinction”
– images usually can be corrected to compare brightnesses
• CCD defects and artifacts
– “dark current”
• Pixel gives output response even when not exposed to light
– Bad pixels
• Due to manufacturing flaws
• “Dead”, “Hot”, “Flickering” (time-variable response)
– Variations in pixel-to-pixel sensitivity
• every pixel has its own Quantum Efficiency (QE)
• Characterized by measuring response to uniform “flat field” and
subsequently “divided out”
Links #6 and #7
Image Display and Analysis
Image Display and Analysis
• This step often is where astronomy really begins.
• Type and extent of display and analysis depends
on purpose of imaging experiment
• Common examples:
– evaluating whether an object has been detected or not
– determining total CCD signal (counts) for an object,
such as a star
– determining relative intensities of an object from
images at two different wavelengths
– determining relative sizes of an extended object from
images at two different wavelengths
Link #8: Storage
Storage
• Glass plates
–
–
–
–
Requires MUCH climate-controlled storage space
Expensive to store and retrieve
available to one user at a time
now being “digitized” (scanned), as in the archive
you use with DS9
• Digital Images
– Lots of disk space
– cheaper all the time
– available to many users