The first thing you have to understand is that an NSF is NOT a sound file. It is a genuine program that is designed to run on a real NES/Famicom console. An NSF is basically an NES ROM with the code for game and graphics removed.
With that being said, there is always the option of using a tool for exporting NSFs to WAVs, which you can then open in Audacity (or indeed any audio editing software). NSFPlay is one, but there exist others (foobar2000 + foo_gep + LAME.exe will also do the trick, and I think Winamp + any NSF plugin does too but I'm not sure).
Thanks man, I was just confused on how to separate individual channels in Foobar. I figured out, that I needed to select the GEP controls and isolate a channel, then convert it to a wav format within Foobar. Once I did that, I then imported the channels to Audacity. Thanks so much for your input.
Foogep is kinda iffy for channel soloing, as the noise/DPCM channels bleed into each other at the very beginning of the NSF. It's rather annoying.
I'm not certain this is the reason, but check under Preferences->Output and adjust the fade lengths to see if that fixes it. If not, no idea. fb2k doesn't support the tri+noise+dpcm volume attenuation anyway.
Oh yeah, I ran into the same problem in making towave (since I use the same library). I had to run the emulation for a bit before starting the track over and recording then. Seems to be a problem with GME. And I'm not sure about the attenuation; I hadn't paid that close attention, and I hadn't run it on too many NSFs.
