What is signal-to-noise ratio and why is it important? The simplest definition would be the ratio of captured digital signal vs the captured noise acquired during the imaging process. In general, it is a measure of the quality of the image. Today’s digital cameras generate noise simply from the electronics inside the camera.
Ideally, when you are photographing, you want the captured image to be entirely comprised of the target itself and no noise. For all practical purposes, this is impossible. But with daytime photography, there is usually so much ambient light that very short exposure times don’t allow much time for the capture of significant noise compared to the data captured from the target itself.
Nighttime imaging is an entirely different story because celestial objects (other than the planets and sun) generally are very dim and require longer exposures to capture a recognizable image. Typical astrophotography images are minutes long, which allows the generation of significant noise by the camera electronics. Here are some of the factors contributing to noise generation:
- Exposure time: The longer the exposure time, the more signal you will collect, which will improve the SNR. However, longer exposure times also mean more noise, so it is important to find a balance that works for your image. Longer images are, however, generally better because signal accumulates faster than noise.
- Gain: Gain is a setting on your camera that controls how much amplification is applied to the signal. Higher gain can improve the SNR but also introduce more noise because all of the signal is amplified. Gain can be useful for taking full advantage of the bit-depth of the camera in low-light conditions, such as in astrophotography.
- Read noise: Read noise is a type of noise that is introduced by the camera’s sensor when it reads out the image data. Read noise is typically very low but can become more significant with longer exposure times.
- Sky brightness: The brightness of the sky can also affect the SNR of an astrophotography image. A brighter sky means more noise, so finding a dark location to take your images is important.
There is no doubt that longer imaging times result in a better signal-to-noise ratio. So why not just take really long images to get the best ratio? Well, there are several reasons why this is not always the best strategy.
- The brightness of all objects in a target frame is not uniform. Bright stars and brighter parts of galaxies or nebula can become overexposed with longer exposures. You end up “blowing out” portions of the image in an attempt to get good detail in the dimmer areas.
- The longer the exposure time, the higher chance that something will disturb the exposure. For example, a gust of wind or an unexpected bumping of the tripod will cause unwanted blurring. Or a satellite, plane, or cloud could pass through the image during the exposure time, potentially ruining the entire image. Obviously, you have limited time to take your images during the night. If a 30-minute exposure is ruined by one of the above, you have lost a substantial portion of your imaging time. If you are taking shorter images, like 5-minute ones, losing one or two images is only a small portion of the potential imaging time.
Considering all of the above facts, my general recommendation is to take 3-5 minute exposures. I usually try to take at least 50 images in one session. There are several image processing software to choose from, and they do a very good job of combining all of these images into one high-quality image. The software can even remove background unevenness, such as that caused by artificial lighting. Whether you choose fewer long exposures or many shorter images, keep in mind each will improve the signal-to-noise ratio, but there are advantages and disadvantages to each.