Lot of good suggestions here, so in the same "based on my experience" vein I'll add a couple thoughts
Start small. Don't try to build that 1/1 scale gundam just yet. There are several reasons in my mind for doing this, but mostly because your first effort will suck. If you make it three to five minutes, thats a lot of sucking, if you even finish it.
Acomalas's Intentional Accident Plan (AIAP)
1) Make a plan
2) Inevitably mess up executing the plan.
3) Instead of deleting mistake, see what it looks like. Hmmmm
Don't underestimate the gift of chance. Many people marvel at the fact that the best parts of their video are all accidental. Let yourself make these accidents, because they usually look pretty damn cool and serve as encouragement. Leave the monotanous meticulous editing for when you have developed a bit more amv masochism.
Barring all that useless advice, I'd have to say I actually do my most concentrated music video work when I have no free time to spare. If I have a relaxed schedule, I get bored. I don't make AMVs out of boredom, I need fuel to make them. Sometimes they feed me back if I am lucky, otherwise I might give up on them.
So best environment to do a video? For me: 10 page paper due in two days, speech tommorrow, friends are going out but I can't cuz I have dem paper to write, well two days is a lot of time and I'm pretty stressed <adobe premiere insidiously loads itself in the background> No adobe! I have work to do! Yes, your filters are very pretty..but you see--hey, that is a good idea, I better try it before I forget it. <five hours later> Well, the idea didn't work, but I somehow finished the second half of the video instead. And now for a night of lovely--oH HELL the bus!!
On a related note, I'd like to say that if you live off campus, get a damn car. Trust me