aboutsummaryrefslogtreecommitdiff
path: root/research/sample-data/mp3_psa_streamer.py
diff options
context:
space:
mode:
Diffstat (limited to 'research/sample-data/mp3_psa_streamer.py')
-rw-r--r--research/sample-data/mp3_psa_streamer.py513
1 files changed, 513 insertions, 0 deletions
diff --git a/research/sample-data/mp3_psa_streamer.py b/research/sample-data/mp3_psa_streamer.py
new file mode 100644
index 0000000..36d2974
--- /dev/null
+++ b/research/sample-data/mp3_psa_streamer.py
@@ -0,0 +1,513 @@
+#!/usr/bin/env python3
+import os
+import sys
+import time
+import socket
+import struct
+import argparse
+import random
+import tempfile
+import subprocess
+from tqdm import tqdm # progress bar sexxy
+
+# need that for lame transcoding
+try:
+ from mutagen.mp3 import MP3 # type: ignore
+ MUTAGEN_AVAILABLE = True
+except ImportError:
+ MUTAGEN_AVAILABLE = False
+
+# Default configuration values
+DEFAULT_MULTICAST_ADDR = "172.16.20.109"
+DEFAULT_PORT = 1681
+DEFAULT_ZONE_INFO = "01 1000 0000 0000 0000 0000 0000" # Zone 1
+DEFAULT_SEND_DUPLICATES = True # Do it like Bodet SIGMA does and just send duplicates "for redundancy" (UDP for emergency systems who the fuck though that was a good idea)
+DEFAULT_CHUNK_SIZE = 1000 # Need to use 1000 due to not understanding protocol fully yet
+DEFAULT_TTL = 3 # We want it going to Multicast and then to the client, nomore otherwise bodeter speaker gets confused
+DEFAULT_HEADER = "4d454c" # MEL goofy ah header
+DEFAULT_COMMAND = "0703" # Probably the stream command
+DEFAULT_STREAM_ID = "110c" # Write none without qoutes to use random stream ID
+
+# Prem prem maximum it supports
+HIGH_QUALITY = {
+ "bitrate": "256k", # 256k max
+ "sample_rate": 48000, # 48000 max
+ "channels": 1, # mono because uhm one speaker only
+ "sample_format": "s32", # max s32
+ "description": "Highest Quality (256kbps, 48kHz)"
+}
+
+# Bodeter Shitsy Sigma software defaults
+LOW_QUALITY = {
+ "bitrate": "64k", # Bodet Defaults
+ "sample_rate": 32000, # Bodet MP3 file default
+ "channels": 1,
+ "sample_format": "s16", # bodet MP3 file default
+ "description": "Normal Quality (64kbps, 32kHz)"
+}
+
+DEFAULT_TIMING_FACTOR = 1 # Used back when chat gpt suggested i should adjust the timing... but it was really just that I have to use 1000 byte chunks ...
+
+def compute_psa_checksum(data: bytes) -> bytes:
+ """Compute PSA checksum for packet data."""
+ var_e = 0x0000 # Starting seed value
+ for i in range(len(data)):
+ var_e ^= (data[i] + i) & 0xFFFF
+ return var_e.to_bytes(2, 'big') # 2-byte checksum, big-endian
+
+def create_mp3_packet(sequence_num, zone_info, stream_id, mp3_chunk, show_packet_length=False):
+ """Create a PSA packet containing MP3 data."""
+ # Part 1: Header (static "MEL")
+ header = bytes.fromhex("4d454c")
+
+ # Prepare all the components that will go into the payload
+ # Start marker (0100)
+ start_marker = bytes.fromhex("0100")
+
+ # Sequence number (1 byte, Little Endian with FF padding)
+ seq = sequence_num.to_bytes(1, 'little') + b'\xff'
+
+ # Command (static 070301)
+ command = bytes.fromhex(DEFAULT_COMMAND)
+
+ # Zone info
+ zones = bytes.fromhex(zone_info)
+
+ # Stream ID (2 bytes)
+ stream_id_bytes = stream_id.to_bytes(2, 'little')
+
+ # Constant data (appears in all original packets)
+ constant_data = bytes.fromhex("05080503e8")
+
+
+ # Build the payload (everything that comes after start_marker)
+ payload = seq + command + zones + stream_id_bytes + constant_data + mp3_chunk
+
+ # Calculate the length for the header - IMPORTANT: This matches what the PSA software does
+ # Length is the total packet length MINUS the header (4d454c) and the length field itself (2 bytes)
+ length_value = len(start_marker) + len(payload) + 7 # +2 for checksum
+
+ # Insert the length as a 2-byte value (big endian)
+ length_bytes = length_value.to_bytes(2, 'big')
+
+ # Assemble the packet without checksum
+ packet_data = header + length_bytes + start_marker + payload
+
+ # Calculate and append checksum
+ checksum = compute_psa_checksum(packet_data)
+
+ # Create final packet
+ final_packet = packet_data + checksum
+
+ if show_packet_length:
+ print(f"Packet length: {len(final_packet)} bytes, Length field: {length_value} (0x{length_value:04x})")
+
+ return final_packet
+
+def send_multicast_packet(sock, packet, multicast_addr, port, send_duplicates=True):
+ """Send a packet to the multicast address."""
+ try:
+ # Send the packet
+ sock.sendto(packet, (multicast_addr, port))
+
+ # Send a duplicate (common in multicast to improve "reliability")
+ if send_duplicates:
+ sock.sendto(packet, (multicast_addr, port))
+
+ return True
+ except Exception as e:
+ print(f"Error sending packet: {e}")
+ return False
+
+def setup_multicast_socket(ttl=DEFAULT_TTL):
+ """Set up a socket for sending multicast UDP."""
+ sock = socket.socket(socket.AF_INET, socket.SOCK_DGRAM, socket.IPPROTO_UDP)
+ sock.setsockopt(socket.IPPROTO_IP, socket.IP_MULTICAST_TTL, ttl)
+ return sock
+
+def check_dependencies():
+ """Check if required dependencies are installed."""
+ try:
+ subprocess.run(['lame', '--version'], stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+ except FileNotFoundError:
+ print("Error: LAME encoder is required but not found.")
+ print("Please install LAME and make sure it's available in your PATH.")
+ sys.exit(1)
+
+ if not MUTAGEN_AVAILABLE:
+ print("Warning: mutagen library not found. Limited audio file analysis available.")
+ print("Install with: pip install mutagen")
+
+def get_audio_info(mp3_file):
+ """Get audio file information using mutagen if available."""
+ try:
+ if MUTAGEN_AVAILABLE:
+ # Use mutagen to analyze the file
+ audio = MP3(mp3_file)
+
+ # Extract relevant information
+ sample_rate = audio.info.sample_rate
+ bit_rate = int(audio.info.bitrate / 1000) # Convert to kbps
+ channels = audio.info.channels
+ codec_name = "mp3" if audio.info.layer == 3 else f"mpeg-{audio.info.layer}"
+
+ return {
+ 'bit_rate': bit_rate,
+ 'sample_rate': sample_rate,
+ 'channels': channels,
+ 'codec_name': codec_name
+ }
+ else:
+ # Fallback method: use LAME to identify file info
+ cmd = ['lame', '--decode', mp3_file, '-t', '--brief', '-']
+ result = subprocess.run(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.PIPE, text=True)
+
+ # LAME outputs limited info to stderr when using --brief
+ info_text = result.stderr
+
+ # Very basic parsing - this is not ideal but works as a fallback
+ codec_name = "mp3" # Assume it's MP3
+ sample_rate = 44100 # Default assumption
+ channels = 2 # Default assumption
+ bit_rate = 0 # Unknown
+
+ for line in info_text.split('\n'):
+ if "MPEG" in line and "Layer" in line:
+ codec_name = "mp3"
+ if "Hz" in line:
+ parts = line.split()
+ for part in parts:
+ if "Hz" in part:
+ try:
+ sample_rate = int(part.replace("Hz", ""))
+ except ValueError:
+ pass
+ if "stereo" in line.lower():
+ channels = 2
+ elif "mono" in line.lower():
+ channels = 1
+ if "kbps" in line:
+ parts = line.split()
+ for part in parts:
+ if "kbps" in part:
+ try:
+ bit_rate = int(part.replace("kbps", ""))
+ except ValueError:
+ pass
+
+ return {
+ 'bit_rate': bit_rate,
+ 'sample_rate': sample_rate,
+ 'channels': channels,
+ 'codec_name': codec_name
+ }
+
+ except Exception as e:
+ print(f"Error getting audio information: {e}")
+ return None
+
+def transcode_mp3(input_file, quality_preset, include_metadata=False):
+ """Transcode MP3 file using LAME encoder."""
+ print(f"Transcoding audio to {quality_preset['description']}...")
+
+ # Create temporary output file
+ fd, temp_output = tempfile.mkstemp(suffix='.mp3')
+ os.close(fd)
+
+ try:
+ # Build LAME command
+ cmd = [
+ 'lame',
+ '--quiet', # Less output
+ '-m', 'm' if quality_preset['channels'] == 1 else 's', # m=mono, s=stereo
+ '--resample', str(quality_preset['sample_rate'] // 1000), # Convert to kHz
+ '-b', quality_preset['bitrate'].replace('k', ''), # Remove 'k' suffix
+ '--cbr', # Use constant bitrate
+ ]
+
+ # Add options to minimize metadata if requested
+ if not include_metadata:
+ cmd.append('--noreplaygain')
+ # Use more compatible tag options instead of --id3v2-none
+ cmd.append('--tt')
+ cmd.append('') # Empty title
+ cmd.append('--tc')
+ cmd.append('') # Empty comment
+ cmd.append('--nohist')
+
+ # Input and output files
+ cmd.extend([input_file, temp_output])
+
+ result = subprocess.run(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.PIPE, text=True)
+
+ if result.returncode != 0:
+ print(f"Error transcoding audio: {result.stderr}")
+ os.remove(temp_output)
+ return None
+
+ return temp_output
+
+ except Exception as e:
+ print(f"Error during transcoding: {e}")
+ if os.path.exists(temp_output):
+ os.remove(temp_output)
+ return None
+
+def prepare_mp3_file(mp3_file, quality, include_metadata=False, sample_format=None):
+ """Ensure the MP3 file meets the required specifications."""
+ # Check if quality is valid
+ quality_preset = HIGH_QUALITY if quality == "high" else LOW_QUALITY
+
+ # Override sample format if specified
+ if sample_format:
+ quality_preset = quality_preset.copy()
+ quality_preset["sample_format"] = sample_format
+
+ # Get audio file information
+ info = get_audio_info(mp3_file)
+ if not info:
+ return None
+
+ print(f"Audio file details: {info['codec_name']}, {info['bit_rate']}kbps, "
+ f"{info['sample_rate']}Hz, {info['channels']} channel(s)")
+
+ # Check if transcoding is needed
+ needs_transcode = False
+
+ if info['codec_name'].lower() != 'mp3':
+ print("Non-MP3 format detected. Transcoding required.")
+ needs_transcode = True
+ elif info['bit_rate'] is None or abs(info['bit_rate'] - int(quality_preset['bitrate'][:-1])) > 5:
+ print(f"Bitrate mismatch: {info['bit_rate']}kbps vs {quality_preset['bitrate']}. Transcoding required.")
+ needs_transcode = True
+ elif info['sample_rate'] != quality_preset['sample_rate']:
+ print(f"Sample rate mismatch: {info['sample_rate']}Hz vs {quality_preset['sample_rate']}Hz. Transcoding required.")
+ needs_transcode = True
+ elif info['channels'] != quality_preset['channels']:
+ print(f"Channel count mismatch: {info['channels']} vs {quality_preset['channels']} (mono). Transcoding required.")
+ needs_transcode = True
+
+ # If transcoding is needed, do it
+ if needs_transcode:
+ return transcode_mp3(mp3_file, quality_preset, include_metadata)
+ else:
+ print("Audio file already meets required specifications.")
+ return mp3_file
+
+def calculate_delay_ms(chunk_size, bitrate, timing_factor=DEFAULT_TIMING_FACTOR):
+ """
+ Calculate the appropriate delay between chunks for real-time streaming.
+
+ Args:
+ chunk_size: Size of each audio chunk in bytes
+ bitrate: Audio bitrate (string like '128k' or integer)
+ timing_factor: Factor to adjust timing (lower = faster playback)
+
+ Returns:
+ Delay in milliseconds between packets
+ """
+ if isinstance(bitrate, str) and bitrate.endswith('k'):
+ bitrate_kbps = int(bitrate[:-1])
+ else:
+ bitrate_kbps = int(bitrate)
+
+ bytes_per_second = bitrate_kbps * 1000 // 8
+ delay_ms = (chunk_size / bytes_per_second) * 1000 * timing_factor
+ return delay_ms
+
+class StreamTimer:
+ """Maintains proper timing for streaming audio packets."""
+
+ def __init__(self, chunk_size, bitrate, timing_factor=DEFAULT_TIMING_FACTOR):
+ """Initialize the stream timer with audio parameters."""
+ self.chunk_size = chunk_size
+ self.bitrate = bitrate
+ self.timing_factor = timing_factor
+ self.start_time = None
+ self.packets_sent = 0
+ self.delay_per_packet = calculate_delay_ms(chunk_size, bitrate, timing_factor) / 1000.0
+
+ def start(self):
+ """Start the stream timer."""
+ self.start_time = time.time()
+ self.packets_sent = 0
+
+ def wait_for_next_packet(self):
+ """Wait until it's time to send the next packet."""
+ if self.start_time is None:
+ self.start()
+ return
+
+ self.packets_sent += 1
+
+ # Calculate when this packet should be sent
+ target_time = self.start_time + (self.packets_sent * self.delay_per_packet)
+
+ # Calculate how long to wait
+ now = time.time()
+ wait_time = target_time - now
+
+ # Only wait if we're ahead of schedule
+ if wait_time > 0:
+ time.sleep(wait_time)
+
+def stream_mp3_to_psa(mp3_file, multicast_addr, port, zone_info, send_duplicates,
+ chunk_size, quality, include_metadata=False, sample_format=None,
+ show_packet_length=False, timing_factor=DEFAULT_TIMING_FACTOR, fixed_stream_id=None):
+ """Stream an MP3 file to PSA system."""
+ try:
+ # Check dependencies
+ check_dependencies()
+
+ # Prepare MP3 file (transcode if needed)
+ prepared_file = prepare_mp3_file(mp3_file, quality, include_metadata, sample_format)
+ if not prepared_file:
+ print("Failed to prepare MP3 file.")
+ return
+
+ using_temp_file = prepared_file != mp3_file
+
+ try:
+ # Use fixed stream ID if provided, otherwise generate a random one
+ if fixed_stream_id is not None:
+ stream_id = fixed_stream_id
+ else:
+ stream_id = random.randint(0, 0xFFFF) # 16-bit (2 bytes) stream ID
+
+ # Open the MP3 file
+ with open(prepared_file, 'rb') as f:
+ mp3_data = f.read()
+
+ # Calculate number of chunks
+ chunks = chunk_mp3_data(mp3_data, chunk_size)
+ total_chunks = len(chunks)
+
+ # Get the appropriate quality preset
+ quality_preset = HIGH_QUALITY if quality == "high" else LOW_QUALITY
+
+ # Calculate the appropriate delay for real-time streaming using the actual bitrate
+ delay_ms = calculate_delay_ms(chunk_size, quality_preset['bitrate'])
+
+ print(f"Streaming {os.path.basename(mp3_file)} ({len(mp3_data)/1024:.2f} KB)")
+ print(f"Split into {total_chunks} packets of {chunk_size} bytes each")
+ print(f"Sending to multicast {multicast_addr}:{port}")
+ print(f"Stream ID: {stream_id:04x}, Zone info: {zone_info}")
+ print(f"Duplicate packets: {'Yes' if send_duplicates else 'No'}")
+ print(f"Quality preset: {HIGH_QUALITY['description'] if quality == 'high' else LOW_QUALITY['description']}")
+ print(f"Including metadata: {'Yes' if include_metadata else 'No'}")
+ print(f"Real-time streaming delay: {delay_ms:.2f}ms per packet")
+
+ # Setup multicast socket
+ sock = setup_multicast_socket()
+
+ # Initialize the stream timer with the appropriate parameters
+ timer = StreamTimer(chunk_size, quality_preset['bitrate'], timing_factor)
+ timer.start()
+
+ # Process and send each chunk
+ sequence_num = 0
+ with tqdm(total=total_chunks, desc="Streaming progress") as pbar:
+ try:
+ for i, chunk in enumerate(chunks):
+ # Create packet with current sequence number
+ packet = create_mp3_packet(sequence_num, zone_info, stream_id, chunk, show_packet_length)
+
+ # Send the packet
+ success = send_multicast_packet(sock, packet, multicast_addr, port, send_duplicates)
+
+ # Increment sequence num
+ sequence_num = (sequence_num + 1) % 256
+
+ # Update progress bar
+ pbar.update(1)
+
+ # Wait for the next packet timing using our timer
+ if i < total_chunks - 1: # No need to wait after the last chunk
+ timer.wait_for_next_packet()
+ except Exception as e:
+ print(f"Error during streaming: {e}")
+
+ print("\nStreaming completed successfully!")
+
+ finally:
+ # Clean up temporary file if created
+ if using_temp_file and os.path.exists(prepared_file):
+ os.remove(prepared_file)
+
+ except FileNotFoundError:
+ print(f"Error: MP3 file {mp3_file} not found")
+ sys.exit(1)
+ except Exception as e:
+ print(f"Error during streaming: {e}")
+ sys.exit(1)
+
+def chunk_mp3_data(data, chunk_size):
+ """
+ Split MP3 data into chunks of specified size.
+
+ Args:
+ data: The complete MP3 data as bytes
+ chunk_size: Size of each chunk in bytes
+
+ Returns:
+ A list of data chunks
+ """
+ chunks = []
+ # Calculate how many chunks we'll have
+ num_chunks = (len(data) + chunk_size - 1) // chunk_size
+
+ for i in range(num_chunks):
+ start = i * chunk_size
+ end = min((i + 1) * chunk_size, len(data))
+ chunks.append(data[start:end])
+
+ return chunks
+
+def main():
+ # Set up command line arguments
+ parser = argparse.ArgumentParser(description="Stream MP3 to Bodet PSA System")
+ parser.add_argument("mp3_file", help="Path to MP3 file to stream")
+ parser.add_argument("-a", "--addr", default=DEFAULT_MULTICAST_ADDR,
+ help=f"Multicast address (default: {DEFAULT_MULTICAST_ADDR})")
+ parser.add_argument("-p", "--port", type=int, default=DEFAULT_PORT,
+ help=f"UDP port (default: {DEFAULT_PORT})")
+ parser.add_argument("-z", "--zone", default=DEFAULT_ZONE_INFO,
+ help=f"Hex zone info (default: Zone 1)")
+ parser.add_argument("-n", "--no-duplicates", action="store_true",
+ help="Don't send duplicate packets (default: send duplicates)")
+ parser.add_argument("-c", "--chunk-size", type=int, default=DEFAULT_CHUNK_SIZE,
+ help=f"MP3 chunk size in bytes (default: {DEFAULT_CHUNK_SIZE})")
+ parser.add_argument("-q", "--quality", choices=["high", "low"], default="high",
+ help="Audio quality preset: high (128kbps, 48kHz) or low (40kbps, 32kHz) (default: high)")
+ parser.add_argument("-m", "--include-metadata", action="store_true",
+ help="Include metadata in MP3 stream (default: no metadata)")
+ parser.add_argument("-s", "--sample-format", choices=["s16", "s24", "s32"], default=None,
+ help="Sample format to use for transcoding (default: s16)")
+ parser.add_argument("--show-packet-length", action="store_true",
+ help="Show packet length information during streaming")
+ parser.add_argument("-t", "--timing-factor", type=float, default=DEFAULT_TIMING_FACTOR,
+ help=f"Timing adjustment factor (lower = faster playback, default: {DEFAULT_TIMING_FACTOR})")
+ parser.add_argument("--fixed-stream-id", type=int,
+ help="Use a fixed stream ID instead of a random one (0-65535)")
+
+ args = parser.parse_args()
+
+ # Stream the MP3 file
+ stream_mp3_to_psa(
+ args.mp3_file,
+ args.addr,
+ args.port,
+ args.zone,
+ not args.no_duplicates,
+ args.chunk_size,
+ args.quality,
+ args.include_metadata,
+ args.sample_format,
+ args.show_packet_length,
+ args.timing_factor,
+ args.fixed_stream_id # Pass fixed_stream_id parameter
+ )
+
+# Remove duplicate main() call
+if __name__ == "__main__":
+ main()