elevenlabs-usage

warning

tests

this repo is not fully tested yet (tho mostly). please use with a note in the back of your mind that things may not work, to protect against disappointment.

assumptions/requirements

a secrets.py with elevenlabs_key, some pip installs (pygame, ffmpeg, rich), a dump folder <-- list is probably not exhaustive

index, the juice, the point

crop an audio:
  crop.py --> no function, just 6 lines of code

generate and download audio:
  eleven_labs.py -> use_tts(voice_id, string)

get voice ID for each name:
  eleven_labs.py -> get_name_ID_mapping()

get settings for a given voice ID:
  eleven_labs.py -> get_settings(ID)

get settings for each voice:
  eleven_labs.py -> get_all_settings()

edit a voice's labels and settings:
  eleven_labs.py -> edit_voice(name, new_settings)

play an audio
  play_audio.py -> set example_n to 0 for vlc, 1 for pygame, 2 for pygame+tkinter
    (vlc, pygame, tkinter are python modules)

download gen history:
  download_history.py -> bunch of different functions

things i want

API related

easy, clean access to all data, meaning:
- there is history, voice names, training samples, generated audio and associated text for a voice, settings for a specific audio;
  - i want some representation of data where i can access and search over all of that in an easy way

general audio related

crop audios using text
- get speech tagger, use timestamps to crop
  - OpenAI Whisper is strong, but requires GPU to run
picking and playing audios with some interface that is good (as opposed to normal Windows stuff)

misc

all this in a nice interface, probably with tkinter.
not guaranteeing i will do this. this is just what i want.

AtillaYasar/elevenlabs-usage