Die erste Simulation eines Schwarzen Lochs im Jahr 1979 im Vergleich zum ersten Bild eines Schwarzen Lochs im Jahr 2019.

Isn’t there a bunch of valid skepticism about the actual accuracy of the 2019 image?

I watched the documentary on it and it seems that a lot of.. interpretation.. went into creating the image.

‘real’ and ‘simulation’ are doing a whole lot of heavy lifting here.

The ‘real’ image has a _lot_ of filling in by interpolation – there are far fewer ‘pixels’ than there appear to be. From memory, something like fifteen (no, not 15*15) over the entire image.

This is then fed into a model of what the telescope, atmosphere, and source were thought to be doing, and how they all behave, and the inputs messed with until they got a picture that looked reasonable, based on similar simulations run using virtual images of simulated black holes and how they behave when put through the whole processing chain.

If you run the procesing chain on a simulated ‘bar’ or ‘circle’ or other black-hole like shapes, and do not get out something that looks sorta like what went in, you know the processing chain is probably broken.

https://www.youtube.com/watch?v=UGL_OL3OrCE I strongly recommend this lecture by Katie Bouman who was one of the developers of this method of imaging.

The goal is not to have ‘it looked like this’ images – quite – but to be able to state with certainty some properties of the black hole – and the image is a more-or-less accurate byproduct.