settled-reach/tooling/db/audio_post.py

#!/usr/bin/env python3
"""Audio post-processing wrapper around ffmpeg.

Subcommands:
  convert   — WAV to OGG (libvorbis, quality 6)
  normalize — LUFS normalize to -16 LUFS (broadcast standard)
  trim      — Remove leading/trailing silence
  pipeline  — trim + normalize + convert (full post-processing chain)

All operations write to a new file (never overwrites input).
"""

import argparse
import os
import subprocess
import sys
import shutil
import json


def check_ffmpeg():
    if not shutil.which("ffmpeg"):
        print(json.dumps({"ok": False, "error": "ffmpeg not found in PATH"}))
        sys.exit(1)


def run_ffmpeg(args, description):
    """Run ffmpeg, capture output, return success."""
    cmd = ["ffmpeg", "-y", "-hide_banner", "-loglevel", "error"] + args
    print(f"  {description}", file=sys.stderr)
    result = subprocess.run(cmd, capture_output=True, text=True)
    if result.returncode != 0:
        print(json.dumps({
            "ok": False,
            "error": f"ffmpeg failed: {result.stderr.strip()}",
            "command": " ".join(cmd)
        }))
        sys.exit(1)
    return True


def cmd_convert(args):
    """Convert WAV to OGG (libvorbis)."""
    output = args.output or args.input.rsplit(".", 1)[0] + ".ogg"
    run_ffmpeg(
        ["-i", args.input, "-c:a", "libvorbis", "-q:a", str(args.quality), output],
        f"converting {os.path.basename(args.input)} → {os.path.basename(output)}"
    )
    size = os.path.getsize(output)
    print(json.dumps({"ok": True, "output": output, "size_bytes": size}))


def cmd_normalize(args):
    """LUFS normalize audio file."""
    output = args.output or _suffixed(args.input, "_norm")
    run_ffmpeg(
        ["-i", args.input, "-af",
         f"loudnorm=I={args.lufs}:LRA=11:TP=-1",
         output],
        f"normalizing to {args.lufs} LUFS"
    )
    print(json.dumps({"ok": True, "output": output}))


def cmd_trim(args):
    """Trim leading/trailing silence."""
    output = args.output or _suffixed(args.input, "_trimmed")
    af = (
        "silenceremove=start_periods=1:start_silence=0.05"
        f":start_threshold={args.threshold}dB,"
        "areverse,"
        "silenceremove=start_periods=1:start_silence=0.05"
        f":start_threshold={args.threshold}dB,"
        "areverse"
    )
    run_ffmpeg(
        ["-i", args.input, "-af", af, output],
        f"trimming silence (threshold: {args.threshold}dB)"
    )
    print(json.dumps({"ok": True, "output": output}))


def cmd_pipeline(args):
    """Full post-processing: trim → normalize → convert to OGG."""
    base = args.input.rsplit(".", 1)[0]
    trimmed = base + "_trimmed.wav"
    normalized = base + "_norm.wav"
    output = args.output or base + ".ogg"

    # Step 1: trim
    af_trim = (
        "silenceremove=start_periods=1:start_silence=0.05"
        f":start_threshold={args.threshold}dB,"
        "areverse,"
        "silenceremove=start_periods=1:start_silence=0.05"
        f":start_threshold={args.threshold}dB,"
        "areverse"
    )
    run_ffmpeg(
        ["-i", args.input, "-af", af_trim, trimmed],
        "step 1/3: trimming silence"
    )

    # Step 2: normalize
    run_ffmpeg(
        ["-i", trimmed, "-af",
         f"loudnorm=I={args.lufs}:LRA=11:TP=-1",
         normalized],
        f"step 2/3: normalizing to {args.lufs} LUFS"
    )

    # Step 3: convert
    run_ffmpeg(
        ["-i", normalized, "-c:a", "libvorbis", "-q:a", str(args.quality), output],
        f"step 3/3: converting to OGG"
    )

    # Clean up intermediates
    os.remove(trimmed)
    os.remove(normalized)

    size = os.path.getsize(output)
    print(json.dumps({"ok": True, "output": output, "size_bytes": size}))


def _suffixed(path, suffix):
    base, ext = os.path.splitext(path)
    return base + suffix + ext


def main():
    parser = argparse.ArgumentParser(description="Audio post-processing (ffmpeg wrapper)")
    sub = parser.add_subparsers(dest="command", required=True)

    # convert
    p = sub.add_parser("convert", help="WAV → OGG")
    p.add_argument("input", help="Input WAV file")
    p.add_argument("--output", "-o", help="Output file (default: same name .ogg)")
    p.add_argument("--quality", "-q", type=int, default=6, help="Vorbis quality 0-10 (default: 6)")
    p.set_defaults(func=cmd_convert)

    # normalize
    p = sub.add_parser("normalize", help="LUFS normalize")
    p.add_argument("input", help="Input audio file")
    p.add_argument("--output", "-o", help="Output file")
    p.add_argument("--lufs", type=float, default=-16, help="Target LUFS (default: -16)")
    p.set_defaults(func=cmd_normalize)

    # trim
    p = sub.add_parser("trim", help="Trim silence")
    p.add_argument("input", help="Input audio file")
    p.add_argument("--output", "-o", help="Output file")
    p.add_argument("--threshold", type=int, default=-50, help="Silence threshold in dB (default: -50)")
    p.set_defaults(func=cmd_trim)

    # pipeline
    p = sub.add_parser("pipeline", help="Full post-processing: trim + normalize + convert")
    p.add_argument("input", help="Input WAV file")
    p.add_argument("--output", "-o", help="Output OGG file")
    p.add_argument("--quality", "-q", type=int, default=6, help="Vorbis quality 0-10 (default: 6)")
    p.add_argument("--lufs", type=float, default=-16, help="Target LUFS (default: -16)")
    p.add_argument("--threshold", type=int, default=-50, help="Silence threshold in dB (default: -50)")
    p.set_defaults(func=cmd_pipeline)

    args = parser.parse_args()
    check_ffmpeg()
    args.func(args)


if __name__ == "__main__":
    main()