Still confused if I should turn on Git LFS (Large File Storage) in my repository

I read a few articles on Git LFS (Large File Storage) and I am more confused about when I should apply LFS instead of having a clear idea.

My situation is following:

  • I work locally, and don’t push to remote servers, although I am thinking to make a local-repository on another drive for additional backup.
  • I work almost only with media files: PNG, JPEG, TARGA etc…
  • I have some large JSON files per project: up to 10MB, but they are references and never change.
  • I also have large files created by Blender, Maya, Unreal Engine, Affinity Photo, Substance Painter etc…

Some of these file grow easily up to 1GB, some have like 100-150MB.

I found out one argument against LFS is to avoid files that are highly compressible. it wasn’t really explained what it means, compressible by Git itself, or ZIP/7z etc…? How do I check this? I don’t know how or even if Git compresses binary files. I know it works well with texts, though.

I checked files created by Blender or Maya, as well as PNGs files, and they are quite well compressible even up to 15% or lower.

So, in this situation, should I apply LFS in my repository or simply use regular one?