MRS to LJ Speech Processor

MRS2LJSpeech.py

# This script generates the folder structure for ljspeech-1.1 processing from mimic-recording-studio database
# This is a modified version of what was written by Thorsten Müller

import glob
import sqlite3
import ffmpeg
import os

from shutil import copyfile
from shutil import rmtree

# Setup Directory Data
cwd = os.path.dirname(os.path.abspath(__file__))
mrs_dir = os.path.join(cwd, os.pardir, "mimic-recording-studio")
output_dir = os.path.join(cwd, "dataset")
output_dir_audio = ""
output_dir_audio_temp=""
output_dir_speech = ""

# Create folders needed for ljspeech
def create_folders():
  global output_dir
  global output_dir_audio
  global output_dir_audio_temp
  global output_dir_speech

  print('→ Creating Dataset Folders')

  output_dir_speech = os.path.join(output_dir, "LJSpeech-1.1")

  # Delete existing folder if exists for clean run
  if os.path.exists(output_dir_speech):
    rmtree(output_dir_speech)

  output_dir_audio = os.path.join(output_dir_speech, "wavs")
  output_dir_audio_temp = os.path.join(output_dir_speech, "temp")

  # Create Clean Folders
  os.makedirs(output_dir_speech)
  os.makedirs(output_dir_audio)
  os.makedirs(output_dir_audio_temp)

def convert_audio():
  global output_dir_audio
  global output_dir_audio_temp

  recordings = len([name for name in os.listdir(output_dir_audio_temp) if os.path.isfile(name)])
  print('→ Converting %s Audio Files to 22050 Hz, 16 Bit, Mono\n' % "{:,}".format(recordings))

  for idx, wav in enumerate(glob.glob(os.path.join(output_dir_audio_temp, "*.wav"))):
    percent = (idx + 1) / recordings
    print('› \033[96m%s\033[0m \033[2m%s / %s (%s)\033[0m ' % (os.path.basename(wav), "{:,}".format((idx + 1)), "{:,}".format(recordings), "{:.0%}".format(percent)))

    # Convert WAV file to required format
    (ffmpeg
      .input(wav)
      .output(os.path.join(output_dir_audio, os.path.basename(wav)), acodec='pcm_s16le', ac=1, ar=22050, loglevel='error')
      .overwrite_output()
      .run(capture_stdout=True)
    )

    # Delete Temp File
    os.remove(wav)

  # Remove Temp Folder
  rmtree(output_dir_audio_temp)

def create_meta_data():
  print('→ Creating META Data')

  conn = sqlite3.connect(os.path.join(mrs_dir, "backend", "db", "mimicstudio.db"))
  c = conn.cursor()

  # Create metadata.csv for ljspeech
  metadata = open(os.path.join(output_dir_speech, "metadata.csv"), mode="w", encoding="utf8")

  for row in c.execute('SELECT audio_id, prompt, lower(prompt) FROM audiomodel ORDER BY length(prompt)'):
    metadata.write(row[0] + "|" + row[1] + "|" + row[2] + "\n")
    copyfile(os.path.join(mrs_dir, "backend", "audio_files", "default_user", row[0] + ".wav"), os.path.join(output_dir_audio_temp, row[0] + ".wav"))

  metadata.close()
  conn.close()

def main():
  print('\n\033[48;5;22m  MRS to LJ Speech Processor  \033[0m\n')

  create_folders()
  create_meta_data()
  convert_audio()

  print('\n\033[38;5;86;1m✔\033[0m COMPLETE【ツ】\n')

if __name__ == '__main__':
  main()

This was based on the work of Thorsten Müller as seen in this video: https://www.youtube.com/watch?app=desktop&v=4YT8WZT_x48

What does this do?

• Creates Folder Structure needed for LJ Speech
• Creates MetaData for LJ Speech
• Copies Mimic Recording Studio Recordings (MRS) into Temp Folder
• Processes MRS recordings and converts them to 22050 Hz, 16 Bit, Mono
• Cleans up temp files

Generates some pretty command line output like this:

Setup

I am using this script with the following folder structure ( my projects name is mimic-my-voice ):

mimic-my-voice
- mimic-recording-studio
- TTS
  - MRS2LJSpeech.py

• mimic-recording-studio - This is a clone of Mimic Recording Studio
• TTS - This is a clone of Coqui AI TTS
  • MRS2LJSpeech.py is then placed in the root of TTS folder

How to use

1. You may need to install ffmpeg if you have not already done so.
2. You may also need to install the ffmpeg-python package

   pip install ffmpeg-python

3. Then you can just run the following in terminal ( in the folder that has both mimic-recording-studio & TTS )

   python3 ./TTS/MRS2LJSpeech.py

Hi @manifestinteractive.
That's really a great improvement of my original version, so thank you Peter for this 👏 👏 👏.

I've added some logic to choose which recording session you want to export when you use MRS for more recording sessions. I hope it's okay for you that i took your optimized version, added multi recording session stuff and released it here - of course giving the credits to you 👍 .

https://github.com/thorstenMueller/deep-learning-german-tts/blob/master/helperScripts/MRS2LJSpeech.py