https://reddit.com/r/space/comments/f4l2jv/_/fhrgh2r/?context=1
An explanation of the whole comment would be great. I tried to look up Fourier transform on Wikipedia but it was a lot of math.
In: Mathematics
Lots of correct answers about Fourier transforms, but no one has directly addressed the second part of the question: why it was needed to render that image. The answer is that when you take an image of an eclipse, the photo has a mixture of fuzzy/cloudy features, sharp edges, and everything in between. Using the Fourier transform you can break down the original image into all the constituent parts (fuzzy stuff, sharp stuff, edges, noise, etc) and then you can selectively pick just the one you want to recreate the image.
More technically, you will generally omit the features that have lower frequencies, and retain certain higher frequencies for an image like this. You can also recreate the original image using multiples of a certain frequency. So, say there’s a sharp feature that shows up well in a certain frequency, you might add multiples of that specific feature into the final image. It takes a surprising amount of experience to get really good at processing images like this, but the outcomes can be amazing.
Latest Answers