My system, possibly easier:
1) Record a core track. I usually video this, so ideally uke and vocals together. I would run the camera, and also Audacity (or Sonar in my case). The tricksy part is that you need to get the balance between uke and vocals right, because if they're both on one track, there's no messing with the balance later
2) Record any other tracks. I might or might not video these, depending how long I want to spend building up my video, but I might add more ukes, guitar, bass, backing vocals - go crazy, experiment, see what works. I would use my headphones at this stage so that I can hear the original track without getting it leaking into the mic and polluting my other tracks.
3) Mix down. I'll play around with the levels of the tracks, add some reverb, maybe some other effects depending on what I'm looking for. The more you do this, the better you get at spotting what will work, but you never get perfet.
4) Export to wav (not mp3 - you lose some quality going to that format, so why do it? The wav file is just going to get imported into a video anyway, so the size is not important)
5) Do the video. I generally use Windows Movie Maker, because it's relatively simple. In the simplest cases, I'll import the video and I'll uncheck the "break into clips" option, so I just get one long chunk. I'll split it at the point I see myself click on the computer to start the recording. I'll then drop the audio in, jiggle things around til it's in synch (you can have the sound the camera recorded and the sound from the audio running together, so you can tell when it's right). Bish bash bosh.
If I'm feeling clever, I might cut in some bits of me playing other instruments, or very rarely I might seperately film an MTV style video, but normally it's the above. Movie Maker makes nice compact files, so they're quick to upload. If I'm doing a collaboration, I'll use Multiquence to run the videos side by side, but it's much more of a faff, and the files it outputs are WAY bigger.
