At the most basic level, a histogram is a plot of the distribution of tones in an image, from black through all gray tones to white. One of the astrophotographers best friends is the histogram. When it comes to image processing, statistical noise, standard deviations, and signal offsets ultimately matter more than what we might see, as if we understand the statistical nature of noise, we can observe how the SNR changes as we change exposure, integrate more information or apply noise reduction routines. At a mathematical level, noise may be statistically higher in an image that appears less noisy…it depends on how strong the signal is relative the noise in each image (the SNR).
The stronger the signal, the less apparent noise we’ll see. The amount of apparent noise in an image depends on the signal to noise ratio. Noise can be, and usually is, a bit more complex than that…as it doesn’t necessarily only occur at the pixel level, however for the purposes of this article, per-pixel noise is generally what we are concerned with. The more noise we have, the larger the deviations tend to be, and the grainier the image appears. The grainy noise we usually think about when viewing an image, a two dimensional signal, is the result of random deviations from the norm for each pixel. Noise is the deviation from the proper, correct or “normal” value that a pixel should have for the given exposure. Even a single pixel has noise, although it may seem counter-intuitive to think of it that way. Every pixel can be thought of as a one dimensional signal. A digital signal, to be exact…and a two dimensional one.
Signals & NoiseĮvery image we process is a signal. A bit of theory around signals, noise and histograms helps both with the acquisition process as well as with processing. There is an extensive amount of work involved to produce world class images, including both pre-processing (calibration, registration, integration) as well as post-processing. The effort required to process an astrophoto is much more involved than that required for the average daytime photograph. Astrophotography Basics: Signal, Noise and HistogramsĪs astrophotographers, we must be deeply versed not only in image acquisition, but also in image processing.