That's quite exhaustive, but it makes me wonder why isn't anything done to correct for that. Like for example instead of taking one 15h exposure, why not take three 5h exposures and roll the telescope 5 degrees in between, then median filter out the artefacts?
Mainly because it doesn't matter. They're not looking at the stars in the foreground, they're looking at the background which is much further away. The diffraction pattern is actually super dim -- those foreground stars are just very bright due to the exposure.
It took like 5 months to cool web to operational temperatures rolling the telescope would create so much heat all new images would be useless until it cools down again.
That makes no sense, they have to rotate it every time they take a picture otherwise they'd be looking at the same spot all the time. Motors don't emit that much heat and neither do torque wheels.
Though I suppose now that I think of it, it's possible the main mirror assembly actually has no built in roll control but only pitch, since the yaw part could be done by moving the entire telescope while remaining shaded. I've never seen any videos showing the full movement, but the previews for LUVIOR show it having full 3 degree articulation relative to the heatsink segment, so I assumed the Webb also has it given that they're extremely similar designs.
> otherwise they'd be looking at the same spot all the time
It's in an orbit around L2, so it's not statically positioned in space. L2 also moves with the earth around the Sun, so it's not statically limited to any one region of the sky.
LUVIOR is not web. Web doesn't have articulation like LUVIOR its fixed only the mirror segements move. also they don't rorate everytime they take a picture there's limitations beacuse its an infered telescope. https://jwst-docs.stsci.edu/jwst-observatory-characteristics.... Web also has a field of view 15x hubble
