After coming across this article at AJATT (like, 3 months ago) I finally got around to implementing it into my regular studies! I have to admit, I’m really happy with the results. I have done things a little differently, though. For example, I use Audacity (free!) to split the audio. Yes, it requires more work, but I’m a bit nitpicky with having the audio match the text on the card. My current card format is (while I generally use Surusu for my text cards, I prefer Anki for audio/pictures):
Front: <lyric from song>
Back: <audio transcription>
On the back, I only add vocab I am really interested in learning. For me, the focus is to just learn to blindly sing along, to catch the nuances in the language, not much the exact content.
An example card learning the theme song for Kaohsiung/高雄‘s MRT system:
I keep all the lyrics below it, and use blue to highlight the current line. It helps keep it organized in my mind where it actually goes in the song. At first I thought splitting it up in this way would actually cause me more issues keeping the song in order in my head, but it actually works really well! The background music helps to keep the rhythm going as you do each lyric. Sounds crazy but it works.
I have also used it to learn a poem that practically every Chinese person will know (like a popular nursery rhyme). It is a poem by Li Bai (李白，唐朝詩人) called “Thoughts On a Still Night”:
I would suggest learning it and reciting it to your Chinese friends, they’ll be impressed! Don’t believe me? Read the Wikipedia page about it: