Actually, contrary to popular belief, most astronomers do NOT stay up all night. In some cases one does (typically if you go to an observatory), but when you are observing with radio telescope arrays or satellites you can essentially e-mail in your observations and then sit back and watch TV while they're being done. (You can't just send an text e-mail with your requests, you have to work with some fairly complex software to generate the observing information.)
Most of the time for astronomers is spent back at their home institution in front of a computer. That is where all the information from telescopes, satellites or whatnot are analyzed using a wide variety of specialized software packages. That data analysis almost always takes a lot longer than the actual observations.
Then you have to write it all up and show how your data is either just darn interesting, or fits with an established theory, or - if you're lucky - doesn't fit with an established theory.
But the process for doing this can take months or years. Rarely is anyone in modern science, not just astonomy, working alone. You need to get input and help from your colleagues (and help them when they are taking the lead in writing up papers).
You need to send the paper in to be printed, realizing that if it is in a hot field it had better be old news by the time the actual print copy appears. You need to send your finished paper out to friends and others in the same sub-field. Post it one electronic document-servers for astronomy and physics. And also go to either specialized or general astronomy conferences. The American Astronomical Society has two big conferences a year. At the confernes you need to meet other people and start to cross connect information as well as promote your own ideas.
Let's see, you also have to write grants. Getting grants for telescope or satellite time is often not that hard. You have to be able to scientifically justify your observations and don't get everything, but it is not terribly difficult to build up more data than you can use in a couple of years. The hard past - - and this is critical - - is that it is much harder to get grant money. If you are a postdoc you may need a lot of grant money just for your salary.
On top of that, many astronomers have to do something else. Often it is to teach a couple of classes at a university. Or, if you are working at some place like the National Radio Astronomy Observatory you need to work on the various software packages and routines that you and others use to process and analyze data.
And on top of that you need a job. My own experience was more or less typical (I was a radio/X-ray astronomer for about a decade after grad school).
After college there were 4 more years of grad school in physics (about half of all astronomers have physics degrees). I actually finished slightly faster than some. Then I had two 1-year teaching positions as a Visiting Professor. After that I got a 2-3 year postdoc at Penn State. It turned out to last for 6 years, except that wasn't obvious at the time and except for one year I spent a lot of time looking for a permanent position. I got one, finally, and went the astronomy group at Fermilab outside of Chicago.
Note that by this time others that I had known in college, even some slow ones, had houses and were well established into careers. Boring ones, but careers none the less. We (myself + wife + kid) were still renting and while my salary was okay, it wasn't particularly impressive for the amount of effort that had gone into the whole process.
And I ended up not liking the work at Fermilab a whole lot (turns out I had a brain tumor - but that's another story).
So after 2 years at Fermi I, very, very reluctantly left science for various computer and technical writing jobs with various business entities. I make more money, do a lot less work, and still at times regret having to leave science sometimes.