Skip to content

Latest commit

 

History

History
56 lines (39 loc) · 1.77 KB

README.md

File metadata and controls

56 lines (39 loc) · 1.77 KB
Raw

lyrical-video-generator

STATUATORY WARNING: this project is currently just a set of scripts written from an experimental purpose

to generate a lyrical video from provided audio file or text file

Usage

  • syntax: ./tasks.sh <help|dep|txt-to-vid|t2v|frame-n-vid|fnv|dataframes|frames|vid|transcribe>

  • hard-coded path as of now (WIP for config)

  • base data dir at './lvg-data'

  • dir for audio files at './lvg-data/transcription'

  • mp3 audio file used as input at './lvg-data/transcription/transcript.mp3'

  • transcripted text file from audio at 'lvg-data/transcript.txt'

  • dataframe file created from text file at './lvg-data/transcript.txt.lvg'

  • image frames created from dataframes at './lvg-data/frames'

  • video name = '/tmp/mygeneratedvideo.avi'

  • Parameters Help

  • dep: to install pip dependencies

  • transcribe: to convert audio file to text

  • dataframes: convert available transcription/text to input data format for create-frame

  • frames: to create frame images for video from dataframes

  • vid: to create video from existing frames

  • tnv (or txt-to-vid): to run all commands from creating dataframes to creating video (if uou have text file)

  • fnv (or frame-n-vid): to create frame images and video from it

Utilities

  • transcribe.py to convert lvg-data/transcription/transcript.mp3 into text at lvg-data/transcript.txt

  • to-frame-data.py to convert transcription lvg-data/transcript.txt into frame data input file lvg-data/transcript.txt.lvg for next step

  • create-frames.py to create frames based on lvg-data/transcript.txt.lvg

  • create-video.py to create video /tmp/mygeneratedvideo.avi from frame files at lvg-data/frames