keepr

organize your audio samples.. but don't touch them.

problem

• too many audio samples
  • 250 gigs scattered about in different subdirectories
  • moving them would immediately cause chaos in past project files

solution

• create folder filled with subfolders that we populate with symlinks.
  • use file names, wav data, and parent directory names for hints
  • allows for easy browsing of audio samples from any standard DAW browser by:
    • key
    • tempo
    • percussion type
    • whatever we think of next

keepr is fast. really fast. on my system the bottleneck becomes I/O. When reading and writing to a single NVMe drive keepr averages around 700MBp/s disk read and spikes up to nearly 2GBp/s disk read.

will you ever finish it

do I ever finish anything? idk maybe. it works right now better than the old version (which was a shitty bash script that ran fdfind), so it's lookin good so far.

• guess tempo by filename
• separate wave files and midi files
• validate wave files
• guess key by filename
• guess drum type by parent directory
• create symlinks for all of the above\
• be stupid dumb fast
• verify various theories with wave/midi data
• sort MIDI files
• more taxonomy
• unit tests
• in-app documentation
• more to-do items

recognition

• kr/walk
• go-audio/wav
• go-music-theory/music-theory
• gomidi/midi