lucatnt/als2cue_web repository overview

This simple web app allows the user to upload an Ableton Live .als project file containing markers and export them as a .cue file. This is helpful in order to import these markers into an app such as Podcast Chapters⁠.

⁠Demo installation

If you're not a developer and just need this tool, you can use my demo install at als2cue.lucatnt.com⁠.

⁠How to use this thing

You can either run this with Python (it requires Flask, so pip3 install flask)

python3 main.py

Or you can use the lucatnt/als2cue_web Docker image

docker run -p 5000:80 lucatnt/als2cue_web

Either way you will end up with a webserver running on port 5000.

⁠Why

I record my podcasts⁠ in Ableton live, and I use Live's markers to save when we switch topics. Each topics then becomes a chapter with its own title, and I embed these titles and timestamps into the mp3 files themselves using the awesome Podcast Chapters⁠ Mac app.

Manually creating chapters in Podcast Chapters, however, is pretty laborious, especially given that I already have the timestamps saved as Ableton Live markers in the .als project file.

So I wrote a little script, als2cue, to extract the markers' timestamps and export them in a .cue files, which Podcast Chapters can read.

⁠How

Thanks to offlinemark⁠'s awesome dawtool⁠, als2cue is able to inspect Live's .als project files and extract all the relevant data.

⁠How it used to work

Before I found dawtool⁠, I parsed Live's .als project files manually, and I wrote what follows, it might still be helpful for somebody.

.als project files are basically zipped XML files.

I examined one of them and determined that markers are saved in the Locators array. Each Locator element has a Time field, which determines the marker position, in "beats" (and this shows that Live is supposed to be a music production program that I'm forcing into a podcast editing app).

To convert beats into minutes and seconds, the projects's tempo has to be taken into account, and it can change throughout the project (thanks to @dokfranco⁠ for the tip).

In Tempo > AutomationTarget > Id we can find the automation that regulates the tempo, we can then analyze the envelope to construct a tempo map throughout the arrangement. Given the tempo map and the position of each Locator, it is relatively trivial to calculate the timestamp of each one of them.

Note: this only worked with Live >= 10, version 9 used a different XML structure that wasn't supported by my DIY parser.