Development paused indefinitely, currently working on a personal private project. If anyone would like to create some PRs to improve this go ahead!!!

🚧 UNDER CONSTRUCTION 🚧

Hi! If you want an interface to preview ableton 11 projects, specifically their als (ableton live set) files, this may be for you!

So how does it work?

Preface: Structure of ableton projects

Ableton usually only has other directories (dirs) and als files stored in the root dir as shown below. Palsa creates the projects dir so that users have a place to put their projects for preview. This is not a great way to do it, instead the user should be able to choose the path to their projects dir so that no copying is needed.

projects
-------- project_dir_0
         ----- other_dir
         ----- other_files
         ----- als_0.als
         ----- als_1.als
-------- project_dir_1
         ----- other_dir
         ----- other_files
         ----- als_0.als
         ----- als_1.als

Extracting

The first step of the palsa pipline is extracting. This involves extracting the gzipped contents of the als file, in other words it involves decompressing the contents. The result of this step is the humongous and bloated xml (Extensible Markup Language) content.

Parsing

Not much can be found online about the structure of the content for newer ableton versions, therefore I fed an example als into a large language model (specifically gemeni 1.5 with 1 million context) and asked it to retreive the attributes that are required for previewing files. This was tough as some of the attribute names where quite obscure.

• als_file
  • name
  • group(s)
    • id
    • name
    • color
  • track(s)
    • group_id
    • name
    • color
    • clip(s)
      • name
      • start
      • end
      • loop_data
        • start
        • end

In the future the time signature and tempo will also be parsed so that a grid can be displayed in the preview.

Caching

Finally to speed up loading tens of projects together containing hundreads of als files caching takes place so that the cache can be quickly loaded avoiding extracting and parsing entirely. This speeds up load times drastically.

Final tidbits

Parsing and loading cache of projects as well as their als files is all done in parralel which is just an amazing performance boost. Also, thanks for reading! 😊