Hacker News Reader

Comment by ericb | Hacker News Reader

I'm not an audio person, but if the component tracks (sublayers, like audio, vocals, guitar) of a song were separate, it seems like a reasonable use. People could fork songs with different vocals, combine different parts of songs for mashups, etc.

edit: As an aside, as an artistic statement on why autotune is terrible, I was thinking it would be interesting if someone would show the world what classic songs, ruined by autotune, would sound like.

View on HN

It would be cool if there was some standardized "make" utility for multitrack audio, where you could distribute component tracks, and people could fork, come up with their own mixes, effects, mastering, etc. and generate a text file that contained all the metadata required to bounce the tracks down. Of course, effects would be a problem, since most of them are commercial and everyone uses a different toolbox of effects.

I'm sure if everyone agreed on a DAW suite, something like this would be rather easy to implement.

> I'm not an audio person, but if the component tracks (sublayers, like audio, vocals, guitar) of a song were separate, it seems like a reasonable use.

Yeah, but not for Git. If you tried that with Git you'd basically have a nice wrapper around tarballing your work directory every morning.

Not so sure about that... Git would let you do something like this, and if the tracks are separately stored, you could pick and choose commits, and fork, branch, merge, etc.

git branch newleadsinger

git commit -m "added new main vocal track"

git commit -m "trying new drum track"

git commit -m "guitar track moved up a key"

git commit -m "more cowbell"

git commit -m "less cowbell"

git commit -m "deleted cowbell track"

git commit -m "added echo to vocal track"

Reply on news.ycombinator.com