Transcript CCD Cameras – How they Work

CCD Cameras – How
They Work
And How to Correct for the Errors
They Make Along the Way
CCD vs CMOS – in 25 words or
less
• CCD = “Charge Coupled Device” have a
readout along the edge of the chip, and
produce less noise, usually best in
astronomical setting
• CMOS= “Complementary Metal Oxide
SemiConductor”
• CMOS chips have a voltage converter for
each separate pixel and read-out faster
(but noisier) than CCD’s
A. How Does the Chip Work?
The photoelectric effect is
fundamental to the operation of a CCD.
• Atoms in a silicon crystal have electrons arranged in
discrete energy bands. The lower energy band is called
the Valence Band, the upper band is the Conduction
Band. Most of the electrons occupy the Valence band
but can be excited into the conduction band by heating
or by the absorption of a photon. The energy required for
this transition is 1.26 electron volts for silicon.
• Once in this conduction band the electron is free to move
about in the lattice of the silicon crystal. It leaves behind
a "hole" in the valence band which acts like a positively
charged carrier.
• In the absence of an external electric field the hole and
electron will quickly re-combine and be lost. In a CCD an
electric field is introduced to sweep these charge carriers
apart and prevent recombination.
•
Absorbing a Photon Raises electron to the
Conduction Band, where it can be easily
moved
1.26ev = mid IR Photon – CCD’s sensitive
only to Photons of higher energy than
this.
• Thermally generated electrons are
indistinguishable from photo-generated
electrons. They constitute a noise source known
as "Dark Current" and it is important that CCDs
are kept cold to reduce their number.
• 1.26eV corresponds to the energy of light with a
wavelength of 1 micron. Beyond this wavelength
silicon becomes transparent and CCDs
constructed from silicon become insensitive.
CCD Chip vs. old style Film
• CCD’s record up to 90% of the photons
that fall on them, vs. only a few percent for
film emulsions
• CCD’s are linear over nearly their entire
range (double the light, double the
counts), while film is very non-linear.
Important for photometry
• CCD’s have far higher dynamic range. In
other words, they can record a far higher
range of light intensities than film.
The “Bucket Brigade” Analogy
There are Several Sources of
Noise in Silicon Detectors like
CMOS and CCD’s
• 2. Readout Noise. In the process of
“bucket brigade”ing the pixel’s store of
conduction electrons to the amplifier, they
can gain or lose a few in the counting
process.
• 1. Thermal Noise. Jostling of atoms due to
not being at Absolute Zero temperature
can also knock electrons into the
conduction band.
1. Read-Out Noise
• Bucket-brigading those buckets full of electrons
one column across at a time and then into the
wires, you can gain or lose a few along the way.
• How to deal?
• A. If it’s a major noise source, don’t read out
your chip any more often than necessary.
• B. Spend more money on your detector chip.
(CCD’s are better than CMOS chips in this
regard;
• CCD’s read-out noise typically about a dozen
electrons per pixel. Tiny!
• CMOS chips usually what is used in commercial nonastronomical digital cameras like in your iPhone and point/shoot
pocket cameras, and even in expensive DSLR’s).
• For daylight photography, usually you have PLENTY of photons
and so readout noise doesn’t matter, so CMOS chips are fine
for daylight photography.
2. Dark Frame Correction
• Jiggling of atoms can knock loose
electrons and they are collected just as if
they were knocked lose by a photon of
light from your subject
• But, it’s just noise – thermal noise
A proper dark frame is an exposure
with the aperture closed (like, you
forgot to take off the lens cap).
• Thermal noise is being added to your real
picture at (on average) the same rate as it is
added to the chip with the lens cap on.
• This is true as long as the chip is at the same
temperature in both situations.
• If the chip were at absolute 0 temperature, there
would be no thermal noise.
• Then a dark frame picture would look like this
An Actual Dark Frame from Our SBIG ST2000
camera. 5 min at Temp=-21C
3. Pixel-to-Pixel Sensitivity Variance
• Pixels are super tiny, like 7 microns
typically (7 millionths of a meter) and not
all identical
• They can gain electrons at slightly different
rates. There may even be obvious “hot
pixels” which produce lots more signal
than they should
• The fix for this problem is easy, as we’ll
see in a few minutes
And there’s millions of pixels in
that little gray square!
4. Dust on the Chip Window
• In an astronomical camera, the chip sits in
an airtight little chamber so dew won’t form
on it, and the pixels stay perfectly clean.
• But, the window above it which lets in the
light can get dust on it which may not be
easy to just blow off.
• This dust will show up as “dust donuts” –
out of focus “donuts” of dimness, because
the chip window is not at the focal point of
the optics
5. Optical Illumination of the
Chip Can Be Uneven
• Your square chip is at the back end of a
large, long set of round lenses and
mirrors, and the center, or “optical axis”
will get the most light from the sky, while
the edges and corners will get a little less
• So you see, your pixels will send off the
“wrong” amount of signal for two reasons –
it may receive less proper light, and it may
respond to it with different sensitivity
The Fix for Pixel Sensitivity,
Optical Illumination, and Dust is
All in One - The Flat Field
Correction
• A flat field is an image of a perfectly
uniformly lit field. Example: the twilight sky
(over a small bit of that sky)
• Dividing your image by a flat field
corrects for all these error sources
• Divide?? Yes – every pixel value is
divided by it’s value in the
corresponding flat field image
Color Pictures?
• So, how is color information gotten in this
process? It seems that any photon more
energetic than a 1-micron IR photon will
get counted equally, regardless of color.
• Yes, so you need to color filter what
photons get to your pixel.
• There are Two Ways to do this…
1. Use a Filter Wheel
Advantages of a Filter Wheel
• You use the entire chip for each color, so
you have a more sensitive image
• You can choose a different exposure time
for each filter, if desired.
• Color resolution is single pixel, just like the
image itself. Lower resolution (~2 pixel) for
the Bayer Mask method
2. A Bayer Filter Mask
Advantages of a Bayer Mask
• 1. Need take only one image. “Single Shot
Color”. MUCH faster to accomplish.
• 2. Moving objects such as comets and
asteroids will yield better images, because
a filter wheel tri-color image will have gaps
in time between each color image due to
time needed to save image and re-starting
the next image. You get a dot-dash set of
different color stars, which may not be
appetizing.
Steps in Creating a Final Digital
CCD Image
• Take your Image at the telescope
• Subtract a dark frame taken for the same
exposure length and at the same chip
temperature with the same camera
• Divide by a flat field
• Convert to color. Now you’ve got a fully
corrected, color image!
Color for Cabrillo
• Our Big Dome’s CCD camera uses the
Bayer Mask approach – “single shot color”
• Our Nikon D7000 camera does the same
• No filter wheels needed! They can be a
hassle anyway, which is why I don’t use
them any more at Cabrillo
How to Do All These Steps?
• I’ve already taken dark frames for the various
temperatures we may be able to get down to.
We have a “dark frame library”!
• We also have flat fields already taken
• So for the user, it’s all done in software, and I’ll
help you with the steps as we get our photos!
• It takes much longer to talk about the above
steps than it does to DO them – For you, just a
minute or two to accomplish these steps
• But, there’s more wizardry to come…
After The Corrections…
• There are two more software steps after you have a fully
corrected color picture
• 1. For the CCD images, you’ll take 2 back-to-back 5
minute pictures. We’ll stack these together to make the
equivalent of a 10 minute picture, which will show you
how shot noise is reduced. Free off-the-web software
called Registax will do this for us
• 2. Then we will import into Photoshop and use many of
the options there to polish your picture into final form. It’s
really in Photoshop that you take possession of your
photo and give it your individual touch unlike anyone
else’s!
• So, if we have too much fog and not enough fresh CCD
images, we’ve got plenty already in the bank we can give
to you to work with