Back to Astrophotography
STACKING
What is stacking? And how is it useful in digital astrophotography?
Imagine that you have four images (called subs) of the same object. Each
sub you captured one after the other. The four subs are too dark. All four of them are
ONE-QUARTER as bright as you need for a good bright image.
Now, if you use a computer to SUM the light from all four subs your compound image is as
bright as you needed it to be. The light from all subs has been combined to give you a
single bright image.
This is called stacking.
The purpose of stacking is to take multiple images and using the computer's stacking
software create either i) the SUM or ii) an AVERAGE of them.
By summing the short exposures we have in effect created a single long exposure.
Sadly folks we live in an imperfect world. Digital cameras add a layer of noise (a random
image) over the top of every picture they take. For daytime photos the noise is of no
consequence. The noise layer is so dim that it is easily swamped by bright daytime images.
For dim dark astrophotography shots however the noise is a major problem and you can see
it in your shots as a grainy flecked appearance. Noise looks like someone has thrown
salt and pepper across your picture with patches in the image dotted with stray
odd-colored pixels.
By summing the image data we also sum the noise that exists within it. Our four short
images when added together make the image four times brighter but it still looks noisy.
The noise has also been summed.
Also, when we sum our four images the brighter parts of the image will be summed to the
maximum level possible, and those parts of the image turn white, their detail is lost
(overexposed) and cannot be recovered.
Here's where Averaging comes to the rescue. The brighter parts of our combined image
aren't overexposed.
Whether summing or averaging when we combine our four images the signal to
noise ratio is increased. This is because the signal (the collected light from the
object) always contributes to the signal on a pixel, but noise is random and so doesn't
always contribute. The more data you collect the less noisy your picture appears.
The final stacked image is clearer, the data (object) is clearer and stands out higher
above the reduced layer of noise. The more subs you stack the better your signal to noise
ratio.
Once you have a good signal to noise ratio you can turn on the "amplifier" by
cranking up the curves in your Photoshop to amplify the detail that would otherwise have
been hidden down among the noise.
To help with stacking a huge boost to your signal-to-noise ratio will be
achieved if you subtract a dark frame.
DARK FRAMES
Another way to reduce noise significantly is to create a dark frame. A dark frame is a
picture that you take when the lens of the camera has the cap on. No light gets in but the
noise is still there. In effect you take a picture of the camera's noise. You
then use computer software to subtract the noise layer from your photos.
This works because certain aspects of noise are statistically predictable. There is
a certain predictable "average" level of noise for each pixel (sort of like how
temperature is an average measure of the random movement of molecules). Also some
pixels in the camera's CMOS/CCD chip are always much noisier than others and are near
their brightest level even in the dark. These are called hot pixels.
Armed with your dark frame you then use the computer software to subtract it from your
photos. You'll be amazed at how well this works. Fuzzy noisy photos clear up almost
like magic.
Back to Astrophotography
|
