mp3voladj

• Use ffmpeg to adjust the volume of mp3 files in EBU_R128 standard.
• (I made this roughly, so I'm sure there will be problems...)

Tested Environments

• macOS

Preparation

• Install Docker
• Install jq command
• (Optional) Pull the docker image jrottenberg/ffmpeg (not mandatory as it will be fetched by docker run command)

Script Overview

• This script uses ffmpeg from the jrottenberg/ffmpeg docker image to encode mp3 files according to the EBU_R128 standard.
• It takes two arguments:
  • Argument 1: Full path of the input mp3 file
  • Argument 2: Full path of the output mp3 file

Execution Example

# [Before] Directory Structure
# /your/target/dir
#               ├input
#               │   ├── sample1.mp3
#               │   ├── sample2.mp3
#               │   └── sample3.mp3
#               └output

# Execute script.
$ INPUT_FILE_NAME='/your/target/dir/input/sample1.mp3'
$ OUTPUT_FILE_NAME='/your/target/dir/output/encoded_sample1.mp3'
$ mp3voladj ${INPUT_FILE_NAME} ${OUTPUT_FILE_NAME}


# [After] Directory Structure
# /your/target/dir
#               ├input
#               │   ├── sample1.mp3
#               │   ├── sample2.mp3
#               │   └── sample3.mp3
#               └output
#                    └── encoded_sample1.mp3

Explanation of ffmpeg Options Used

• loudnorm

  • A filter used to normalize the volume
  • Automatically adjusts the volume using an algorithm based on the EBU R128 standard.

• I

  • Specifies the target average volume
  • -23 is a common target volume.

• LRA

  • Specifies the target volume range
  • 7 is a common value.

• TP

  • Specifies the target peak volume
  • -2.0 is a common value.

• measured_I

  • Specifies the measured average volume

• measured_LRA

  • Specifies the measured volume range

• measured_TP

  • Specifies the measured peak volume

• linear

  • Specifies whether to perform volume normalization in linear mode
  • If set to true, it processes in linear mode, which results in lossless volume conversion.

• print_format

  • Specifies the output format of the result
  • summary: Output in a concise format.

• -codec:a libmp3lame

  • Option to specify the audio codec of the output file
  • In this case, libmp3lame, an MP3 encoder, is specified.

• -q:a 2

  • Option to specify the audio quality
  • 2 is an option to specify the quality of the audio.
    • 2 provides relatively high quality while significantly reducing file size.
  • Can be specified in the range of 0 to 9.
    • 0: Highest quality
    • 9: Lowest quality