#!/usr/bin/env python3 """Pair rendered JPEG + H.264 (+ optional AAC audio) as Apple Live Photo on any desktop OS. Modified from LiveCanvas: keeps one AAC audio track (Live Photos support sound) instead of rejecting it. Requires Python >=3.10, Pillow >=10,<12, ffmpeg and ffprobe on PATH. Never imports into Photos. Use verify_live_photo.swift for optional Apple validation. """ import argparse from fractions import Fraction import hashlib import json import math from pathlib import Path import shutil import subprocess import tempfile import uuid from PIL import Image from portable_live_photo.apple import write_live_jpeg, write_live_mov, inspect_live_jpeg, inspect_live_mov def digest(path): with Path(path).open('rb') as stream: return hashlib.file_digest(stream, 'sha256').hexdigest() if hasattr(hashlib, 'file_digest') else hashlib.sha256(stream.read()).hexdigest() def probe(path): result = subprocess.run(['ffprobe', '-v', 'error', '-show_streams', '-show_format', '-of', 'json', str(path)], check=True, capture_output=True, text=True) return json.loads(result.stdout) def check_inputs(photo, movie, seconds, fps): if not all(math.isfinite(n) for n in (seconds, fps)) or not 0 < fps <= 240: raise ValueError('Invalid cover time/fps') info = probe(movie) streams = info['streams'] videos = [s for s in streams if s['codec_type'] == 'video'] if len(videos) != 1 or videos[0]['codec_name'] != 'h264': raise ValueError('Expected one H.264 video stream') audios = [s for s in streams if s['codec_type'] == 'audio'] if len(audios) > 1 or any(a['codec_name'] != 'aac' for a in audios): raise ValueError('At most one AAC audio stream supported') video = videos[0] duration = float(video.get('duration', info['format']['duration'])) if not math.isfinite(duration) or not 0 <= seconds or seconds + 1/fps > duration + 0.0001: raise ValueError('Cover time outside video') if abs(float(Fraction(video['avg_frame_rate'])) - fps) > 0.01 or abs(seconds*fps-round(seconds*fps)) > 0.0001: raise ValueError('FPS mismatch or cover time not aligned to frame') if abs(float(video.get('start_time', 0))) > 0.0001: raise ValueError('Expected zero-based render timeline') if any(float(d.get('rotation', 0)) != 0 for d in video.get('side_data_list', [])): raise ValueError('Render rotation into pixels before pairing') with Image.open(photo) as cover: if cover.format != 'JPEG' or cover.size != (video['width'], video['height']): raise ValueError('Expected JPEG matching video dimensions') if cover.getexif().get(274, 1) != 1: raise ValueError('Render EXIF orientation into pixels first') cover.load() return duration def verify_pair(directory, manifest): directory = Path(directory).resolve() paths = [] for key in ('photo', 'video'): path = (directory / manifest[key]).resolve() if path.parent != directory or not path.is_file(): raise ValueError('Invalid pair resource path') if manifest.get('sha256', {}).get(key) != digest(path): raise ValueError('Pair hash mismatch: ' + key) paths.append(path) photo, movie = paths identifier = inspect_live_jpeg(photo) metadata = inspect_live_mov(movie) if identifier != manifest['identifier'] or metadata.asset_id != identifier or metadata.sample_value != 0: raise ValueError('Live Photo identifier or signed metadata sample mismatch') if abs(metadata.still_time - manifest['coverTime']) > 1/60000: raise ValueError('Timed metadata position mismatch') # Independent FFmpeg demuxer must find the metadata sample at the same timestamp. packets = json.loads(subprocess.run(['ffprobe', '-v', 'error', '-select_streams', 'd', '-show_packets', '-of', 'json', str(movie)], check=True, capture_output=True, text=True).stdout)['packets'] if len(packets) != 1 or abs(float(packets[0]['pts_time']) - manifest['coverTime']) > 1/60000: raise ValueError('FFprobe timed sample readback failed') # FFmpeg 7.x omits duration_time for a sample that follows an empty edit; the writer sets it to one frame. if 'duration_time' in packets[0] and abs(float(packets[0]['duration_time']) - 1/manifest['fps']) > 2/60000: raise ValueError('Timed sample duration mismatch') check_inputs(photo, movie, manifest['coverTime'], manifest['fps']) subprocess.run(['ffmpeg', '-v', 'error', '-xerror', '-i', str(movie), '-map', '0:v:0', '-f', 'null', '-'], check=True, capture_output=True) return metadata def pair(photo, movie, output, seconds, fps): photo, movie, output = Path(photo).resolve(), Path(movie).resolve(), Path(output).absolute() if output.exists(): raise ValueError('Output already exists; choose a new directory') duration = check_inputs(photo, movie, seconds, fps) output.parent.mkdir(parents=True, exist_ok=True) staging = Path(tempfile.mkdtemp(prefix='.livecanvas-pair-', dir=output.parent)) try: identifier = str(uuid.uuid4()).upper() raw = staging / 'remux.mov' subprocess.run(['ffmpeg', '-v', 'error', '-nostdin', '-i', str(movie), '-map', '0:v:0', '-map', '0:a:0?', '-map_metadata', '-1', '-c', 'copy', '-movie_timescale', '60000', str(raw)], check=True, capture_output=True) write_live_jpeg(photo, staging / 'live.jpg', identifier) write_live_mov(raw, staging / 'live.mov', identifier, seconds, duration, fps) raw.unlink() manifest = {'identifier': identifier, 'photo': 'live.jpg', 'video': 'live.mov', 'coverTime': seconds, 'duration': duration, 'fps': fps, 'backend': 'portable-python', 'metadataVerified': False, 'localPHLivePhotoLoad': 'not_available', 'localLoadReason': 'Portable backend does not invoke Apple PhotoKit', 'iPhonePlayback': 'not_tested', 'photosImport': 'not_requested', 'wallpaper': 'not_tested', 'importedIntoPhotos': False, 'sha256': {'photo': digest(staging/'live.jpg'), 'video': digest(staging/'live.mov')}} actual = verify_pair(staging, manifest) manifest.update(metadataVerified=True, readbackStillTime=actual.still_time) (staging/'manifest.json').write_text(json.dumps(manifest, ensure_ascii=False, indent=2), encoding='utf-8') if output.exists(): raise ValueError('Output appeared during generation; refusing to replace it') staging.rename(output) return manifest finally: if staging.exists(): shutil.rmtree(staging) if __name__ == '__main__': parser = argparse.ArgumentParser(description=__doc__) parser.add_argument('photo') parser.add_argument('movie') parser.add_argument('output') parser.add_argument('cover_seconds', type=float) parser.add_argument('fps', type=float) args = parser.parse_args() print(json.dumps(pair(args.photo, args.movie, args.output, args.cover_seconds, args.fps), indent=2))