Evolution Audio Converter

This project is a microservice in Go that processes audio files, converts them to opus or mp3 format, and returns both the duration of the audio and the converted file (as base64 or S3 URL). The service accepts audio files sent as form-data, base64, or URL.

Requirements

Before starting, you'll need to have the following installed:

• Go (version 1.21 or higher)
• Docker (to run the project in a container)
• FFmpeg (for audio processing)

Installation

Clone the Repository

Clone this repository to your local machine:

git clone https://github.com/EvolutionAPI/evolution-audio-converter.git
cd evolution-audio-converter

Install Dependencies

Install the project dependencies:

go mod tidy

Install FFmpeg

The service depends on FFmpeg to convert the audio. Make sure FFmpeg is installed on your system.

• On Ubuntu:

  sudo apt update
  sudo apt install ffmpeg

• On macOS (via Homebrew):

  brew install ffmpeg

• On Windows, download FFmpeg here and add it to your system PATH.

Configuration

Create a .env file in the project's root directory. Here are the available configuration options:

Basic Configuration

PORT=4040
API_KEY=your_secret_api_key_here

Transcription Configuration

ENABLE_TRANSCRIPTION=true
TRANSCRIPTION_PROVIDER=openai  # or groq
OPENAI_API_KEY=your_openai_key_here
GROQ_API_KEY=your_groq_key_here
TRANSCRIPTION_LANGUAGE=en  # Default transcription language (optional)

Storage Configuration

ENABLE_S3_STORAGE=true
S3_ENDPOINT=play.min.io
S3_ACCESS_KEY=your_access_key_here
S3_SECRET_KEY=your_secret_key_here
S3_BUCKET_NAME=audio-files
S3_REGION=us-east-1
S3_USE_SSL=true
S3_URL_EXPIRATION=24h

Storage Options

The service supports two storage modes for the converted audio:

1. Base64 (default): Returns the audio file encoded in base64 format
2. S3 Compatible Storage: Uploads to S3-compatible storage (AWS S3, MinIO, etc.) and returns a presigned URL

When S3 storage is enabled, the response will include a url instead of the audio field:

{
  "duration": 120,
  "format": "ogg",
  "url": "https://your-s3-endpoint/bucket/file.ogg?signature...",
  "transcription": "Transcribed text here..." // if transcription was requested
}

If S3 upload fails, the service automatically falls back to base64 encoding.

Running the Project

Locally

To run the service locally:

go run main.go -dev

The server will be available at http://localhost:4040.

Using Docker

1. Build the Docker image:

   docker build -t audio-service .

2. Run the container:

   docker run -p 4040:4040 --env-file=.env audio-service

API Usage

Authentication

All requests must include the apikey header with your API key.

Endpoints

Process Audio

POST /process-audio

Accepts audio files in these formats:

• Form-data
• Base64
• URL

Optional parameters:

• format: Output format (mp3 or ogg, default: ogg)
• transcribe: Enable transcription (true or false)
• language: Transcription language code (e.g., "en", "es", "pt")

Transcribe Only

POST /transcribe

Transcribes audio without format conversion.

Optional parameters:

• language: Transcription language code

Example Requests

Form-data Upload

curl -X POST -F "[email protected]" \
  -F "format=ogg" \
  -F "transcribe=true" \
  -F "language=en" \
  http://localhost:4040/process-audio \
  -H "apikey: your_secret_api_key_here"

Base64 Upload

curl -X POST \
  -d "base64=$(base64 audio.mp3)" \
  -d "format=ogg" \
  http://localhost:4040/process-audio \
  -H "apikey: your_secret_api_key_here"

URL Upload

curl -X POST \
  -d "url=https://example.com/audio.mp3" \
  -d "format=ogg" \
  http://localhost:4040/process-audio \
  -H "apikey: your_secret_api_key_here"

Response Format

With S3 storage disabled (default):

{
  "duration": 120,
  "audio": "UklGR... (base64 of the file)",
  "format": "ogg",
  "transcription": "Transcribed text here..." // if requested
}

With S3 storage enabled:

{
  "duration": 120,
  "url": "https://your-s3-endpoint/bucket/file.ogg?signature...",
  "format": "ogg",
  "transcription": "Transcribed text here..." // if requested
}

License

This project is licensed under the MIT license.