Automated Scripting For 15,000 Bounces Of One Song

[ricschnupp1081]

We need to output 15,000 different mixes of a single 1-minute song, each version with a slightly different lyric.

Recording the performances isn't the difficult part. We are only required to replace one word for each version. The awful part will be in the bouncing process.

PROBLEM ::
How can the task of outputting mixes (bouncing) be scripted so that it happens automatically, without requiring any keystrokes or mouse work for each mix? A solution will require a script that:
1) automates the selection of the next vocal take (or playlist),
2) selects the area to be bounced,
3) gives the bounce a unique title,
4) bounces
5) starts the script over again.

Is Applescript a solution?

Thanks for your help,
Ric

[BGMusic]

This is some sort of a joke, no?

[soundthinker]

ProTools isn't directly scriptable. You may be able to do this with Quickeys.
I don't think Logic is scriptable. Reaper has ReaScript, but I'm not familiar with it's capabilities.

[ronwasserman]

You don't need a script, you need an intern.

[FajitaTone]

can you even have 15,000 playlists?

[ekuehnl]

Bounce the song once without the lyric in question.

Record the 15000 individual lyrics and export them to a folder.

Then use the batch function in a different app like Peak/SoundForge/etc for a faster than real-time mix of the bounced song with the each of the 15000 instances of the lyric. If you set it up correctly, you can just point the app at the folder with the 15000 files and click the button. Then come back a day later and it will be done.

Easy. And well worth the cost of one of those other apps in time saved. You might even be able to do it with a demo copy.

-EK

[jeremyroberts]

I like EK's idea of using a different app, and only bounce the lyric change then merge.

15,000 60-second files = 250 hours without any pre-or post-processing time.

Ric needs to use an app that will mix/bounce faster than real time, else this job will take months. At realtime, 15,000 60-second files will take a minimum of 11 days assuming everything is fully scripted. And that is without any file operation overhead. SO figure 30 seconds per file for overhead. SO add 6 days.

Ric MUST use a faster than real time app to do this.

And data management/file naming conventions must be well thought out.

Is this a "your kid's name embedded in the toy" project? Why not just use a multi--channel playback engine, so instead of custom recording a zillion songs -- add a channel of playback???