Raw Astrophotography Data
Nebulae | Open Clusters | Globular Clusters | Galaxies

The Types of Images

There are four types of images I take: Light Frames, Dark Frames, Flat Frames, and Dark Flat Frames. There is also something known as Bias Frames, but I don’t usually use these. It is best to shoot all images using the camera's raw format. Since I have a Nikon camera, I use the NEF format.
Light Frames

Light frames contain the actual image data of the object you are photographing. There are many imperfections in light frames which need to be removed before they can be stacked. The stacking itself should reduce the random noise in the light frames; however, there are other non-random imperfections which need to be dealt with. Some examples are: thermal noise, vignetting, dust, bias signal, and hot pixels to name a few. Fixing these problems is the job of the dark, flat, and bias frames. After the light frames have been calibrated they can be aligned and stacked to produce the final image.

Light Frame
Raw light frame right out of the camera.
Dark Frames

I like to think that the purpose of taking darks is like to “zeroing” a scale. Each pixel on the camera is slightly different. Some read the signal hotter, and some read the signal colder. Dark frames are images taken with the cap over the aperture of the telescope so no light can get in. This way you have a picture of what black / nothing should look like. However, dark frames are not purely black. They still contain the hot pixels, thermal signal, and whatever other non-random signal is present in your image. By subtracting the dark frames from the light frames you are in a sense “zeroing” them. Dark frames also contain random noise like the light frames. When you subtract a dark frame the random noise also gets subtracted from the light frame which adds noise to the final image. This is why multiple dark frames should be taken so they can be stacked together to reduce the effect of the noise. After the darks frames are stacked, they are subtracted from the light frames to remove the dark signal. It is important that dark frames are taken at the same thermal temperature as the lights since the thermal signal is dependent on temperature. They should also have the same exposure length and ISO.

Note: Dark frames also contain the bias signal which is why I can get away without subtracting bias frames. I have read that subtracting a bias from the darks can help improve DeepSkyStacker’s dark optimization feature.

Dark Frame
This is a dark frame with the histogram stretched out to show what the dark signal looks like.
Flat Frames

Just like with darks I like to think of flats as “zeroing” the image. Only this time flats are correcting optical imperfections, instead of ones with the camera sensor. Most telescopes do not evenly distribute light across the camera sensor. This causes images to be brighter in the center, and darker towards the edges. There is also usually dust on a camera sensor which causes dark blotches to appear. Flats correct for this by taking a picture of what a blank, evenly illuminated surface should look like. The flats are then stacked together and the lights are divided by them. This evens out the illumination throughout the light frame.

Light Frame with no Flat
This light frame has not had a flat applied to it. Notice the uneven distribution of light.

Flats need to be taken with the same focus, camera orientation, and optical setup as the lights. There are many ways to take flats such as using the evening sky, a light box, or pointing the telescope at a white computer screen. I use the last of these methods. To take flats I tape a sheet of paper towels over the aperture to help diffuse the light, and I point my scope at the computer screen which is displaying a blank white image. I then proceed to take 30 – 50 images. For these images I let the camera determine the correct exposure. Next I put the cover over the aperture and I take Dark Flats. These serve the same purpose as the darks but they are applied to the flats and should be the same exposure length and ISO as the flats.

Flat Frame
This flat frame shows the distribution of light over the sensor, and specs of dust.
Bias Frames

Bias frames are used to remove the readout signal from you camera sensor. Even when a pixel has not received any sort of signal there is still variation in how the camera reads data off the sensor. Bias frames can be subtracted from your lights, darks, flats, and dark flats to remove this variation. To capture the bias signal, pictures need to be taken at zero exposure length (or as close as possible) and with the lens cap on. They are not dependent on temperature but should be taken with the same ISO. I do not usually take bias frames since the bias signal is also contained in the dark frames. I have read that removing the bias signal from darks can help programs such as Deep Sky Stacker optimize the dark frames. If bias frames are used they will be subtracted from all other images at the start of the stacking process.

Please email me your questions, comments, and photos.