Scan Audio Dna

Instruction

You are the Metadata Scanner. The user has provided a finished audio master. Your job is to "listen" to the file and extract the data needed for the DDEX publishing manifest.

Procedure

1. Validate: Ensure the file exists and is a valid audio format (WAV/MP3).
2. Execute Scan: Run the analysis tool to get the sonic data.
3. Output: Return a JSON object with bpm, duration_ms, key, and loudness.
   • CRITICAL: Do not modify the input file. Read-only access.

Tool Code (Python)

import subprocess
import json
import os

def scan_master(file_path):
    # 1. Get Duration via FFprobe
    probe_cmd = [
        "ffprobe", "-v", "error", "-show_entries",
        "format=duration", "-of", "default=noprint_wrappers=1:nokey=1",
        file_path
    ]
    duration = float(subprocess.check_output(probe_cmd).strip())

    # 2. (Optional) Simulate BPM/Key scan if Essentia is missing (for the prototype)
    # In production, this would call the Essentia C++ extractor
    audio_dna = {
        "filename": os.path.basename(file_path),
        "duration_seconds": duration,
        "duration_ms": int(duration * 1000),
        "technical_metadata": {
            "sample_rate": 44100, # In real implementation, parse from ffprobe
            "channels": 2
        },
        "mediator_status": "SCANNED_READY_FOR_PUBLISHING"
    }

    print(json.dumps(audio_dna))

# Agent: Call scan_audio_dna(input_file)