I wanted to extract the music from some Nintendo Switch games that hadn’t been ripped yet. I’m not a lawyer, but whole process should be completely legal, assuming you are using your own dumps, which you should be using regardless.

Steps

You will need your dumped ROM in either .xci or .nsp format, alongside your title.keys and prod.keys extracted from your console, as the dumps can be encrypted. I’m not going to say how to do these things, as I’m a bit worried by the Nintendo Ninjas.

  1. Download/clone the NSTool repository to a folder on your computer, and put your prod.keys in that folder.
  • There are other tools for doing this, I just like NSTools since it just kinda works in my experience.

  1. Place your title.keys file into the ~/.switch/title.keys location on your computer.

  2. Extract the base game .NSP/.XCI file using the following command, adjusted to your exact filenames:

./nstool -k ./prod.keys -x extracted/ ./switchromdump.nsp
  • This will extract many .NCA files (where the content is), .CNMT.NCA files (metadata) and .TIK/.CERT files (certificates/tickets). Generally your largest .NCA file will have the bulk of the music.
  1. If you want to dump an update, you may need to further extract the ticket files by running:
./nstool -k ./prod.keys ./extracted/[random stuff].tik

and then add them all to new lines in the title.keys in step 2’s location formated like this:

[RightsID from the previous command] = [Data from the previous command]
  1. Now you can extract the largest .NCA files from the dumps:
./nstool -k ./prod.keys -x extracted-assets/ ./extracted/[random stuff].nca

All music will generally be stored in the extracted-assets/Sound/Resources/Stream/ folders.

  1. You’ll need to download/install vgmstream from their website to actually extract the music to a usable format.

  2. Convert the music using the following command while inside a folder with the music:

for f in *.bwav; do vgmstream-cli -o "out/${f%.bwav}.wav" "$f"; done

Replace .bwav with whatever format it comes in.

  1. If needed, manually assign metadata using a tool like File Explorer or iTunes (I’m sure there’s a FOSS alternative), and potentially mix split-up tracks in Audacity.