diff --git a/protocol_prototype/DryBox/UI/audio_player.py b/protocol_prototype/DryBox/UI/audio_player.py
new file mode 100644
index 0000000..6b922d7
--- /dev/null
+++ b/protocol_prototype/DryBox/UI/audio_player.py
@@ -0,0 +1,253 @@
+import wave
+import threading
+import queue
+import time
+import os
+from datetime import datetime
+from PyQt5.QtCore import QObject, pyqtSignal
+
+# Try to import PyAudio, but handle if it's not available
+try:
+    import pyaudio
+    PYAUDIO_AVAILABLE = True
+except ImportError:
+    PYAUDIO_AVAILABLE = False
+    print("Warning: PyAudio not installed. Audio playback will be disabled.")
+    print("To enable playback, install with: sudo dnf install python3-devel portaudio-devel && pip install pyaudio")
+
+class AudioPlayer(QObject):
+    playback_started = pyqtSignal(int)  # client_id
+    playback_stopped = pyqtSignal(int)  # client_id
+    recording_saved = pyqtSignal(int, str)  # client_id, filepath
+    
+    def __init__(self):
+        super().__init__()
+        self.audio = None
+        self.streams = {}  # client_id -> stream
+        self.buffers = {}  # client_id -> queue
+        self.threads = {}  # client_id -> thread
+        self.recording_buffers = {}  # client_id -> list of audio data
+        self.recording_enabled = {}  # client_id -> bool
+        self.playback_enabled = {}  # client_id -> bool
+        self.sample_rate = 8000
+        self.channels = 1
+        self.chunk_size = 320  # 40ms at 8kHz
+        self.debug_callback = None
+        
+        if PYAUDIO_AVAILABLE:
+            try:
+                self.audio = pyaudio.PyAudio()
+            except Exception as e:
+                self.debug(f"Failed to initialize PyAudio: {e}")
+                self.audio = None
+        else:
+            self.audio = None
+            self.debug("PyAudio not available - playback disabled, recording still works")
+    
+    def debug(self, message):
+        if self.debug_callback:
+            self.debug_callback(f"[AudioPlayer] {message}")
+        else:
+            print(f"[AudioPlayer] {message}")
+    
+    def set_debug_callback(self, callback):
+        self.debug_callback = callback
+    
+    def start_playback(self, client_id):
+        """Start audio playback for a client"""
+        if not self.audio:
+            self.debug("Audio playback not available - PyAudio not installed")
+            self.debug("To enable: sudo dnf install python3-devel portaudio-devel && pip install pyaudio")
+            return False
+            
+        if client_id in self.streams:
+            self.debug(f"Playback already active for client {client_id}")
+            return False
+        
+        try:
+            # Create audio stream
+            stream = self.audio.open(
+                format=pyaudio.paInt16,
+                channels=self.channels,
+                rate=self.sample_rate,
+                output=True,
+                frames_per_buffer=self.chunk_size
+            )
+            
+            self.streams[client_id] = stream
+            self.buffers[client_id] = queue.Queue()
+            self.playback_enabled[client_id] = True
+            
+            # Start playback thread
+            thread = threading.Thread(
+                target=self._playback_thread,
+                args=(client_id,),
+                daemon=True
+            )
+            self.threads[client_id] = thread
+            thread.start()
+            
+            self.debug(f"Started playback for client {client_id}")
+            self.playback_started.emit(client_id)
+            return True
+            
+        except Exception as e:
+            self.debug(f"Failed to start playback for client {client_id}: {e}")
+            return False
+    
+    def stop_playback(self, client_id):
+        """Stop audio playback for a client"""
+        if client_id not in self.streams:
+            return
+        
+        self.playback_enabled[client_id] = False
+        
+        # Wait for thread to finish
+        if client_id in self.threads:
+            self.threads[client_id].join(timeout=1.0)
+            del self.threads[client_id]
+        
+        # Close stream
+        if client_id in self.streams:
+            try:
+                self.streams[client_id].stop_stream()
+                self.streams[client_id].close()
+            except:
+                pass
+            del self.streams[client_id]
+        
+        # Clear buffer
+        if client_id in self.buffers:
+            del self.buffers[client_id]
+        
+        self.debug(f"Stopped playback for client {client_id}")
+        self.playback_stopped.emit(client_id)
+    
+    def add_audio_data(self, client_id, pcm_data):
+        """Add audio data to playback buffer"""
+        # Initialize frame counter for debug logging
+        if not hasattr(self, '_frame_count'):
+            self._frame_count = {}
+        if client_id not in self._frame_count:
+            self._frame_count[client_id] = 0
+        self._frame_count[client_id] += 1
+        
+        # Only log occasionally to avoid spam
+        if self._frame_count[client_id] == 1 or self._frame_count[client_id] % 25 == 0:
+            self.debug(f"Client {client_id} audio frame #{self._frame_count[client_id]}: {len(pcm_data)} bytes")
+        
+        if client_id in self.buffers:
+            self.buffers[client_id].put(pcm_data)
+            if self._frame_count[client_id] == 1:
+                self.debug(f"Client {client_id} buffer started, queue size: {self.buffers[client_id].qsize()}")
+        else:
+            self.debug(f"Client {client_id} has no buffer (playback not started?)")
+        
+        # Add to recording buffer if recording
+        if self.recording_enabled.get(client_id, False):
+            if client_id not in self.recording_buffers:
+                self.recording_buffers[client_id] = []
+            self.recording_buffers[client_id].append(pcm_data)
+    
+    def _playback_thread(self, client_id):
+        """Thread function for audio playback"""
+        stream = self.streams.get(client_id)
+        buffer = self.buffers.get(client_id)
+        
+        if not stream or not buffer:
+            return
+        
+        self.debug(f"Playback thread started for client {client_id}")
+        
+        while self.playback_enabled.get(client_id, False):
+            try:
+                # Get audio data from buffer with timeout
+                audio_data = buffer.get(timeout=0.1)
+                
+                # Only log first frame to avoid spam
+                if not hasattr(self, '_playback_logged'):
+                    self._playback_logged = {}
+                if client_id not in self._playback_logged:
+                    self._playback_logged[client_id] = False
+                    
+                if not self._playback_logged[client_id]:
+                    self.debug(f"Client {client_id} playback thread playing first frame: {len(audio_data)} bytes")
+                    self._playback_logged[client_id] = True
+                
+                # Play audio
+                stream.write(audio_data)
+                
+            except queue.Empty:
+                # No data available, continue
+                continue
+            except Exception as e:
+                self.debug(f"Playback error for client {client_id}: {e}")
+                break
+        
+        self.debug(f"Playback thread ended for client {client_id}")
+    
+    def start_recording(self, client_id):
+        """Start recording received audio"""
+        self.recording_enabled[client_id] = True
+        self.recording_buffers[client_id] = []
+        self.debug(f"Started recording for client {client_id}")
+    
+    def stop_recording(self, client_id, save_path=None):
+        """Stop recording and optionally save to file"""
+        if not self.recording_enabled.get(client_id, False):
+            return None
+        
+        self.recording_enabled[client_id] = False
+        
+        if client_id not in self.recording_buffers:
+            return None
+        
+        audio_data = self.recording_buffers[client_id]
+        
+        if not audio_data:
+            self.debug(f"No audio data recorded for client {client_id}")
+            return None
+        
+        # Generate filename if not provided
+        if not save_path:
+            timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+            save_path = f"wav/received_client{client_id}_{timestamp}.wav"
+        
+        # Ensure directory exists
+        save_dir = os.path.dirname(save_path)
+        if save_dir:
+            os.makedirs(save_dir, exist_ok=True)
+        
+        try:
+            # Combine all audio chunks
+            combined_audio = b''.join(audio_data)
+            
+            # Save as WAV file
+            with wave.open(save_path, 'wb') as wav_file:
+                wav_file.setnchannels(self.channels)
+                wav_file.setsampwidth(2)  # 16-bit
+                wav_file.setframerate(self.sample_rate)
+                wav_file.writeframes(combined_audio)
+            
+            self.debug(f"Saved recording for client {client_id} to {save_path}")
+            self.recording_saved.emit(client_id, save_path)
+            
+            # Clear recording buffer
+            del self.recording_buffers[client_id]
+            
+            return save_path
+            
+        except Exception as e:
+            self.debug(f"Failed to save recording for client {client_id}: {e}")
+            return None
+    
+    def cleanup(self):
+        """Clean up audio resources"""
+        # Stop all playback
+        for client_id in list(self.streams.keys()):
+            self.stop_playback(client_id)
+        
+        # Terminate PyAudio
+        if self.audio:
+            self.audio.terminate()
+            self.audio = None
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/UI/audio_processor.py b/protocol_prototype/DryBox/UI/audio_processor.py
new file mode 100644
index 0000000..8f98c79
--- /dev/null
+++ b/protocol_prototype/DryBox/UI/audio_processor.py
@@ -0,0 +1,220 @@
+import numpy as np
+import wave
+import os
+from datetime import datetime
+from PyQt5.QtCore import QObject, pyqtSignal
+import struct
+
+class AudioProcessor(QObject):
+    processing_complete = pyqtSignal(str)  # filepath
+    
+    def __init__(self):
+        super().__init__()
+        self.debug_callback = None
+        
+    def debug(self, message):
+        if self.debug_callback:
+            self.debug_callback(f"[AudioProcessor] {message}")
+        else:
+            print(f"[AudioProcessor] {message}")
+    
+    def set_debug_callback(self, callback):
+        self.debug_callback = callback
+    
+    def apply_gain(self, audio_data, gain_db):
+        """Apply gain to audio data"""
+        # Convert bytes to numpy array
+        samples = np.frombuffer(audio_data, dtype=np.int16)
+        
+        # Apply gain
+        gain_linear = 10 ** (gain_db / 20.0)
+        samples_float = samples.astype(np.float32) * gain_linear
+        
+        # Clip to prevent overflow
+        samples_float = np.clip(samples_float, -32768, 32767)
+        
+        # Convert back to int16
+        return samples_float.astype(np.int16).tobytes()
+    
+    def apply_noise_gate(self, audio_data, threshold_db=-40):
+        """Apply noise gate to remove low-level noise"""
+        samples = np.frombuffer(audio_data, dtype=np.int16)
+        
+        # Calculate RMS in dB
+        rms = np.sqrt(np.mean(samples.astype(np.float32) ** 2))
+        rms_db = 20 * np.log10(max(rms, 1e-10))
+        
+        # Gate the audio if below threshold
+        if rms_db < threshold_db:
+            return np.zeros_like(samples, dtype=np.int16).tobytes()
+        
+        return audio_data
+    
+    def apply_low_pass_filter(self, audio_data, cutoff_hz=3400, sample_rate=8000):
+        """Apply simple low-pass filter"""
+        samples = np.frombuffer(audio_data, dtype=np.int16).astype(np.float32)
+        
+        # Simple moving average filter
+        # Calculate filter length based on cutoff frequency
+        filter_length = int(sample_rate / cutoff_hz)
+        if filter_length < 3:
+            filter_length = 3
+            
+        # Apply moving average
+        filtered = np.convolve(samples, np.ones(filter_length) / filter_length, mode='same')
+        
+        return filtered.astype(np.int16).tobytes()
+    
+    def apply_high_pass_filter(self, audio_data, cutoff_hz=300, sample_rate=8000):
+        """Apply simple high-pass filter"""
+        samples = np.frombuffer(audio_data, dtype=np.int16).astype(np.float32)
+        
+        # Simple differentiator as high-pass
+        filtered = np.diff(samples, prepend=samples[0])
+        
+        # Scale to maintain amplitude
+        scale = cutoff_hz / (sample_rate / 2)
+        filtered *= scale
+        
+        return filtered.astype(np.int16).tobytes()
+    
+    def normalize_audio(self, audio_data, target_db=-3):
+        """Normalize audio to target dB level"""
+        samples = np.frombuffer(audio_data, dtype=np.int16).astype(np.float32)
+        
+        # Find peak
+        peak = np.max(np.abs(samples))
+        if peak == 0:
+            return audio_data
+        
+        # Calculate current peak in dB
+        current_db = 20 * np.log10(peak / 32768.0)
+        
+        # Calculate gain needed
+        gain_db = target_db - current_db
+        
+        # Apply gain
+        return self.apply_gain(audio_data, gain_db)
+    
+    def remove_silence(self, audio_data, threshold_db=-40, min_silence_ms=100, sample_rate=8000):
+        """Remove silence from audio"""
+        samples = np.frombuffer(audio_data, dtype=np.int16)
+        
+        # Calculate frame size for silence detection
+        frame_size = int(sample_rate * min_silence_ms / 1000)
+        
+        # Detect non-silent regions
+        non_silent_regions = []
+        i = 0
+        
+        while i < len(samples):
+            frame = samples[i:i+frame_size]
+            if len(frame) == 0:
+                break
+                
+            # Calculate RMS of frame
+            rms = np.sqrt(np.mean(frame.astype(np.float32) ** 2))
+            rms_db = 20 * np.log10(max(rms, 1e-10))
+            
+            if rms_db > threshold_db:
+                # Found non-silent region, find its extent
+                start = i
+                while i < len(samples):
+                    frame = samples[i:i+frame_size]
+                    if len(frame) == 0:
+                        break
+                    rms = np.sqrt(np.mean(frame.astype(np.float32) ** 2))
+                    rms_db = 20 * np.log10(max(rms, 1e-10))
+                    if rms_db <= threshold_db:
+                        break
+                    i += frame_size
+                non_silent_regions.append((start, i))
+            else:
+                i += frame_size
+        
+        # Combine non-silent regions
+        if not non_silent_regions:
+            return audio_data  # Return original if all silent
+        
+        combined = []
+        for start, end in non_silent_regions:
+            combined.extend(samples[start:end])
+        
+        return np.array(combined, dtype=np.int16).tobytes()
+    
+    def save_processed_audio(self, audio_data, original_path, processing_type):
+        """Save processed audio with descriptive filename"""
+        # Generate new filename
+        base_name = os.path.splitext(os.path.basename(original_path))[0]
+        timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+        new_filename = f"{base_name}_{processing_type}_{timestamp}.wav"
+        
+        # Ensure directory exists
+        save_dir = os.path.dirname(original_path)
+        if not save_dir:
+            save_dir = "wav"
+        os.makedirs(save_dir, exist_ok=True)
+        
+        save_path = os.path.join(save_dir, new_filename)
+        
+        try:
+            with wave.open(save_path, 'wb') as wav_file:
+                wav_file.setnchannels(1)
+                wav_file.setsampwidth(2)
+                wav_file.setframerate(8000)
+                wav_file.writeframes(audio_data)
+            
+            self.debug(f"Saved processed audio to {save_path}")
+            self.processing_complete.emit(save_path)
+            return save_path
+            
+        except Exception as e:
+            self.debug(f"Failed to save processed audio: {e}")
+            return None
+    
+    def concatenate_audio_files(self, file_paths, output_path=None):
+        """Concatenate multiple audio files"""
+        if not file_paths:
+            return None
+        
+        combined_data = b''
+        sample_rate = None
+        
+        for file_path in file_paths:
+            try:
+                with wave.open(file_path, 'rb') as wav_file:
+                    if sample_rate is None:
+                        sample_rate = wav_file.getframerate()
+                    elif wav_file.getframerate() != sample_rate:
+                        self.debug(f"Sample rate mismatch in {file_path}")
+                        continue
+                    
+                    data = wav_file.readframes(wav_file.getnframes())
+                    combined_data += data
+                    
+            except Exception as e:
+                self.debug(f"Failed to read {file_path}: {e}")
+        
+        if not combined_data:
+            return None
+        
+        # Save concatenated audio
+        if not output_path:
+            timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+            output_path = f"wav/concatenated_{timestamp}.wav"
+        
+        os.makedirs(os.path.dirname(output_path), exist_ok=True)
+        
+        try:
+            with wave.open(output_path, 'wb') as wav_file:
+                wav_file.setnchannels(1)
+                wav_file.setsampwidth(2)
+                wav_file.setframerate(sample_rate or 8000)
+                wav_file.writeframes(combined_data)
+            
+            self.debug(f"Saved concatenated audio to {output_path}")
+            return output_path
+            
+        except Exception as e:
+            self.debug(f"Failed to save concatenated audio: {e}")
+            return None
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/UI/client_state.py b/protocol_prototype/DryBox/UI/client_state.py
deleted file mode 100644
index 3f260c6..0000000
--- a/protocol_prototype/DryBox/UI/client_state.py
+++ /dev/null
@@ -1,79 +0,0 @@
-# client_state.py
-from queue import Queue
-from session import NoiseXKSession
-import time
-
-class ClientState:
-    def __init__(self, client_id):
-        self.client_id = client_id
-        self.command_queue = Queue()
-        self.initiator = None
-        self.keypair = None
-        self.peer_pubkey = None
-        self.session = None
-        self.handshake_in_progress = False
-        self.handshake_start_time = None
-        self.call_active = False
-
-    def process_command(self, client):
-        """Process commands from the queue."""
-        if not self.command_queue.empty():
-            print(f"Client {self.client_id} processing command queue, size: {self.command_queue.qsize()}")
-            command = self.command_queue.get()
-            if command == "handshake":
-                try:
-                    print(f"Client {self.client_id} starting handshake, initiator: {self.initiator}")
-                    self.session = NoiseXKSession(self.keypair, self.peer_pubkey)
-                    self.session.handshake(client.sock, self.initiator)
-                    print(f"Client {self.client_id} handshake complete")
-                    client.send("HANDSHAKE_DONE")
-                except Exception as e:
-                    print(f"Client {self.client_id} handshake failed: {e}")
-                    client.state_changed.emit("CALL_END", "", self.client_id)
-                finally:
-                    self.handshake_in_progress = False
-                    self.handshake_start_time = None
-
-    def start_handshake(self, initiator, keypair, peer_pubkey):
-        """Queue handshake command."""
-        self.initiator = initiator
-        self.keypair = keypair
-        self.peer_pubkey = peer_pubkey
-        print(f"Client {self.client_id} queuing handshake, initiator: {initiator}")
-        self.handshake_in_progress = True
-        self.handshake_start_time = time.time()
-        self.command_queue.put("handshake")
-
-    def handle_data(self, client, data):
-        """Handle received data (control or audio)."""
-        try:
-            decoded_data = data.decode('utf-8').strip()
-            print(f"Client {self.client_id} received raw: {decoded_data}")
-            if decoded_data in ["RINGING", "CALL_END", "CALL_DROPPED", "IN_CALL", "HANDSHAKE", "HANDSHAKE_DONE"]:
-                client.state_changed.emit(decoded_data, decoded_data, self.client_id)
-                if decoded_data == "HANDSHAKE":
-                    self.handshake_in_progress = True
-                elif decoded_data == "HANDSHAKE_DONE":
-                    self.call_active = True
-            else:
-                print(f"Client {self.client_id} ignored unexpected text message: {decoded_data}")
-        except UnicodeDecodeError:
-            if self.call_active and self.session:
-                try:
-                    print(f"Client {self.client_id} received audio packet, length={len(data)}")
-                    decrypted_data = self.session.decrypt(data)
-                    print(f"Client {self.client_id} decrypted audio packet, length={len(decrypted_data)}")
-                    client.data_received.emit(decrypted_data, self.client_id)
-                except Exception as e:
-                    print(f"Client {self.client_id} failed to process audio packet: {e}")
-            else:
-                print(f"Client {self.client_id} ignored non-text message: {data.hex()}")
-
-    def check_handshake_timeout(self, client):
-        """Check for handshake timeout."""
-        if self.handshake_in_progress and self.handshake_start_time:
-            if time.time() - self.handshake_start_time > 30:
-                print(f"Client {self.client_id} handshake timeout after 30s")
-                client.state_changed.emit("CALL_END", "", self.client_id)
-                self.handshake_in_progress = False
-                self.handshake_start_time = None
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/UI/main.py b/protocol_prototype/DryBox/UI/main.py
index 5780be7..8d7ce77 100644
--- a/protocol_prototype/DryBox/UI/main.py
+++ b/protocol_prototype/DryBox/UI/main.py
@@ -1,19 +1,32 @@
 import sys
 from PyQt5.QtWidgets import (
     QApplication, QMainWindow, QWidget, QVBoxLayout, QHBoxLayout,
-    QPushButton, QLabel, QFrame, QSizePolicy, QStyle
+    QPushButton, QLabel, QFrame, QSizePolicy, QStyle, QTextEdit, QSplitter,
+    QMenu, QAction, QInputDialog, QShortcut
 )
-from PyQt5.QtCore import Qt, QSize
-from PyQt5.QtGui import QFont
+from PyQt5.QtCore import Qt, QSize, QTimer, pyqtSignal
+from PyQt5.QtGui import QFont, QTextCursor, QKeySequence
+import time
+import threading
 from phone_manager import PhoneManager
 from waveform_widget import WaveformWidget
 from phone_state import PhoneState
 
 class PhoneUI(QMainWindow):
+    debug_signal = pyqtSignal(str)
+    
     def __init__(self):
         super().__init__()
-        self.setWindowTitle("Enhanced Dual Phone Interface")
-        self.setGeometry(100, 100, 900, 750)
+        self.setWindowTitle("DryBox - Noise XK + Codec2 + 4FSK")
+        self.setGeometry(100, 100, 1200, 900)
+        
+        # Set minimum size to ensure window is resizable
+        self.setMinimumSize(800, 600)
+        
+        # Auto test state
+        self.auto_test_running = False
+        self.auto_test_timer = None
+        self.test_step = 0
         self.setStyleSheet("""
             QMainWindow { background-color: #333333; }
             QLabel { color: #E0E0E0; font-size: 14px; }
@@ -39,75 +52,162 @@ class PhoneUI(QMainWindow):
                 padding: 15px;
             }
             QWidget#phoneWidget {
-                border: 1px solid #4A4A4A; border-radius: 8px;
-                padding: 10px; background-color: #3A3A3A;
+                border: 2px solid #4A4A4A; border-radius: 10px;
+                background-color: #3A3A3A;
+                min-width: 250px;
             }
+            QTextEdit#debugConsole {
+                background-color: #1E1E1E; color: #00FF00; 
+                font-family: monospace; font-size: 12px;
+                border: 2px solid #0078D4; border-radius: 5px;
+            }
+            QPushButton#autoTestButton {
+                background-color: #FF8C00; min-height: 35px;
+            }
+            QPushButton#autoTestButton:hover { background-color: #FF7F00; }
         """)
 
+        # Setup debug signal early
+        self.debug_signal.connect(self.append_debug)
+        
         self.manager = PhoneManager()
+        self.manager.ui = self  # Set UI reference for debug logging
         self.manager.initialize_phones()
 
-        # Main widget and layout
+        # Main widget with splitter
         main_widget = QWidget()
         self.setCentralWidget(main_widget)
         main_layout = QVBoxLayout()
-        main_layout.setSpacing(20)
-        main_layout.setContentsMargins(20, 20, 20, 20)
-        main_layout.setAlignment(Qt.AlignCenter)
         main_widget.setLayout(main_layout)
+        
+        # Create splitter for phones and debug console
+        self.splitter = QSplitter(Qt.Vertical)
+        main_layout.addWidget(self.splitter)
+        
+        # Top widget for phones
+        phones_widget = QWidget()
+        phones_layout = QVBoxLayout()
+        phones_layout.setSpacing(20)
+        phones_layout.setContentsMargins(20, 20, 20, 20)
+        phones_layout.setAlignment(Qt.AlignCenter)
+        phones_widget.setLayout(phones_layout)
 
         # App Title
-        app_title_label = QLabel("Dual Phone Control Panel")
+        app_title_label = QLabel("Integrated Protocol Control Panel")
         app_title_label.setObjectName("mainTitleLabel")
         app_title_label.setAlignment(Qt.AlignCenter)
-        main_layout.addWidget(app_title_label)
+        phones_layout.addWidget(app_title_label)
+        
+        # Protocol info
+        protocol_info = QLabel("Noise XK + Codec2 (1200bps) + 4FSK")
+        protocol_info.setAlignment(Qt.AlignCenter)
+        protocol_info.setStyleSheet("font-size: 12px; color: #00A2E8;")
+        phones_layout.addWidget(protocol_info)
 
         # Phone displays layout
         phone_controls_layout = QHBoxLayout()
-        phone_controls_layout.setSpacing(50)
-        phone_controls_layout.setAlignment(Qt.AlignCenter)
-        main_layout.addLayout(phone_controls_layout)
+        phone_controls_layout.setSpacing(20)
+        phone_controls_layout.setContentsMargins(10, 0, 10, 0)
+        phones_layout.addLayout(phone_controls_layout)
 
         # Setup UI for phones
         for phone in self.manager.phones:
-            phone_container_widget, phone_display_frame, phone_button, waveform_widget, sent_waveform_widget, phone_status_label = self._create_phone_ui(
+            phone_container_widget, phone_display_frame, phone_button, waveform_widget, sent_waveform_widget, phone_status_label, playback_button, record_button = self._create_phone_ui(
                 f"Phone {phone['id']+1}", lambda checked, pid=phone['id']: self.manager.phone_action(pid, self)
             )
             phone['button'] = phone_button
             phone['waveform'] = waveform_widget
             phone['sent_waveform'] = sent_waveform_widget
             phone['status_label'] = phone_status_label
+            phone['playback_button'] = playback_button
+            phone['record_button'] = record_button
+            
+            # Connect audio control buttons with proper closure
+            playback_button.clicked.connect(lambda checked, pid=phone['id']: self.toggle_playback(pid))
+            record_button.clicked.connect(lambda checked, pid=phone['id']: self.toggle_recording(pid))
             phone_controls_layout.addWidget(phone_container_widget)
-            phone['client'].data_received.connect(lambda data, cid=phone['id']: self.manager.update_waveform(cid, data))
+            # Connect data_received signal - it emits (data, client_id)
+            phone['client'].data_received.connect(lambda data, cid: self.manager.update_waveform(cid, data))
             phone['client'].state_changed.connect(lambda state, num, cid=phone['id']: self.set_phone_state(cid, state, num))
             phone['client'].start()
 
-        # Spacer
-        main_layout.addStretch(1)
-
+        # Control buttons layout
+        control_layout = QHBoxLayout()
+        control_layout.setSpacing(15)
+        control_layout.setContentsMargins(20, 10, 20, 10)
+        
+        # Auto Test Button
+        self.auto_test_button = QPushButton("🧪 Run Automatic Test")
+        self.auto_test_button.setObjectName("autoTestButton")
+        self.auto_test_button.setMinimumWidth(180)
+        self.auto_test_button.setMaximumWidth(250)
+        self.auto_test_button.clicked.connect(self.toggle_auto_test)
+        control_layout.addWidget(self.auto_test_button)
+        
+        # Clear Debug Button
+        self.clear_debug_button = QPushButton("Clear Debug")
+        self.clear_debug_button.setMinimumWidth(100)
+        self.clear_debug_button.setMaximumWidth(150)
+        self.clear_debug_button.clicked.connect(self.clear_debug)
+        control_layout.addWidget(self.clear_debug_button)
+        
+        # Audio Processing Button
+        self.audio_menu_button = QPushButton("Audio Options")
+        self.audio_menu_button.setMinimumWidth(100)
+        self.audio_menu_button.setMaximumWidth(150)
+        self.audio_menu_button.clicked.connect(self.show_audio_menu)
+        control_layout.addWidget(self.audio_menu_button)
+        
         # Settings Button
         self.settings_button = QPushButton("Settings")
         self.settings_button.setObjectName("settingsButton")
-        self.settings_button.setFixedWidth(180)
+        self.settings_button.setMinimumWidth(100)
+        self.settings_button.setMaximumWidth(150)
         self.settings_button.setIcon(self.style().standardIcon(QStyle.SP_FileDialogDetailedView))
         self.settings_button.setIconSize(QSize(20, 20))
         self.settings_button.clicked.connect(self.settings_action)
-        settings_layout = QHBoxLayout()
-        settings_layout.addStretch()
-        settings_layout.addWidget(self.settings_button)
-        settings_layout.addStretch()
-        main_layout.addLayout(settings_layout)
+        control_layout.addWidget(self.settings_button)
+        
+        phones_layout.addLayout(control_layout)
+        
+        # Add phones widget to splitter
+        self.splitter.addWidget(phones_widget)
+        
+        # Debug console
+        self.debug_console = QTextEdit()
+        self.debug_console.setObjectName("debugConsole")
+        self.debug_console.setReadOnly(True)
+        self.debug_console.setMinimumHeight(200)
+        self.debug_console.setMaximumHeight(400)
+        self.splitter.addWidget(self.debug_console)
+        
+        # Flush any queued debug messages
+        if hasattr(self, '_debug_queue'):
+            for msg in self._debug_queue:
+                self.debug_console.append(msg)
+            del self._debug_queue
+        
+        # Set splitter sizes (70% phones, 30% debug)
+        self.splitter.setSizes([600, 300])
 
         # Initialize UI
         for phone in self.manager.phones:
             self.update_phone_ui(phone['id'])
+            
+        # Initial debug message
+        QTimer.singleShot(100, lambda: self.debug("DryBox UI initialized with integrated protocol"))
+        
+        # Setup keyboard shortcuts
+        self.setup_shortcuts()
 
     def _create_phone_ui(self, title, action_slot):
         phone_container_widget = QWidget()
         phone_container_widget.setObjectName("phoneWidget")
+        phone_container_widget.setSizePolicy(QSizePolicy.Preferred, QSizePolicy.Preferred)
         phone_layout = QVBoxLayout()
         phone_layout.setAlignment(Qt.AlignCenter)
-        phone_layout.setSpacing(15)
+        phone_layout.setSpacing(10)
+        phone_layout.setContentsMargins(15, 15, 15, 15)
         phone_container_widget.setLayout(phone_layout)
 
         phone_title_label = QLabel(title)
@@ -117,8 +217,9 @@ class PhoneUI(QMainWindow):
 
         phone_display_frame = QFrame()
         phone_display_frame.setObjectName("phoneDisplay")
-        phone_display_frame.setFixedSize(250, 350)
-        phone_display_frame.setSizePolicy(QSizePolicy.Fixed, QSizePolicy.Fixed)
+        phone_display_frame.setMinimumSize(200, 250)
+        phone_display_frame.setMaximumSize(300, 400)
+        phone_display_frame.setSizePolicy(QSizePolicy.Preferred, QSizePolicy.Preferred)
 
         display_content_layout = QVBoxLayout(phone_display_frame)
         display_content_layout.setAlignment(Qt.AlignCenter)
@@ -129,28 +230,83 @@ class PhoneUI(QMainWindow):
         phone_layout.addWidget(phone_display_frame, alignment=Qt.AlignCenter)
 
         phone_button = QPushButton()
-        phone_button.setFixedWidth(120)
+        phone_button.setMinimumWidth(100)
+        phone_button.setMaximumWidth(150)
         phone_button.setIconSize(QSize(20, 20))
         phone_button.clicked.connect(action_slot)
         phone_layout.addWidget(phone_button, alignment=Qt.AlignCenter)
 
         # Received waveform
-        waveform_label = QLabel(f"{title} Received Audio")
+        waveform_label = QLabel(f"{title} Received")
         waveform_label.setAlignment(Qt.AlignCenter)
-        waveform_label.setStyleSheet("font-size: 14px; color: #E0E0E0;")
+        waveform_label.setStyleSheet("font-size: 12px; color: #E0E0E0;")
         phone_layout.addWidget(waveform_label)
         waveform_widget = WaveformWidget(dynamic=False)
+        waveform_widget.setMinimumSize(200, 50)
+        waveform_widget.setMaximumSize(300, 80)
+        waveform_widget.setSizePolicy(QSizePolicy.Preferred, QSizePolicy.Fixed)
         phone_layout.addWidget(waveform_widget, alignment=Qt.AlignCenter)
 
         # Sent waveform
-        sent_waveform_label = QLabel(f"{title} Sent Audio")
+        sent_waveform_label = QLabel(f"{title} Sent")
         sent_waveform_label.setAlignment(Qt.AlignCenter)
-        sent_waveform_label.setStyleSheet("font-size: 14px; color: #E0E0E0;")
+        sent_waveform_label.setStyleSheet("font-size: 12px; color: #E0E0E0;")
         phone_layout.addWidget(sent_waveform_label)
         sent_waveform_widget = WaveformWidget(dynamic=False)
+        sent_waveform_widget.setMinimumSize(200, 50)
+        sent_waveform_widget.setMaximumSize(300, 80)
+        sent_waveform_widget.setSizePolicy(QSizePolicy.Preferred, QSizePolicy.Fixed)
         phone_layout.addWidget(sent_waveform_widget, alignment=Qt.AlignCenter)
+        
+        # Audio control buttons
+        audio_controls_layout = QHBoxLayout()
+        audio_controls_layout.setAlignment(Qt.AlignCenter)
+        
+        playback_button = QPushButton("🔊 Playback")
+        playback_button.setCheckable(True)
+        playback_button.setMinimumWidth(90)
+        playback_button.setMaximumWidth(120)
+        playback_button.setStyleSheet("""
+            QPushButton {
+                background-color: #404040;
+                color: white;
+                border: 1px solid #606060;
+                padding: 5px;
+                border-radius: 3px;
+            }
+            QPushButton:checked {
+                background-color: #4CAF50;
+            }
+            QPushButton:hover {
+                background-color: #505050;
+            }
+        """)
+        
+        record_button = QPushButton("⏺ Record")
+        record_button.setCheckable(True)
+        record_button.setMinimumWidth(90)
+        record_button.setMaximumWidth(120)
+        record_button.setStyleSheet("""
+            QPushButton {
+                background-color: #404040;
+                color: white;
+                border: 1px solid #606060;
+                padding: 5px;
+                border-radius: 3px;
+            }
+            QPushButton:checked {
+                background-color: #F44336;
+            }
+            QPushButton:hover {
+                background-color: #505050;
+            }
+        """)
+        
+        audio_controls_layout.addWidget(playback_button)
+        audio_controls_layout.addWidget(record_button)
+        phone_layout.addLayout(audio_controls_layout)
 
-        return phone_container_widget, phone_display_frame, phone_button, waveform_widget, sent_waveform_widget, phone_status_label
+        return phone_container_widget, phone_display_frame, phone_button, waveform_widget, sent_waveform_widget, phone_status_label, playback_button, record_button
 
     def update_phone_ui(self, phone_id):
         phone = self.manager.phones[phone_id]
@@ -182,28 +338,371 @@ class PhoneUI(QMainWindow):
             button.setStyleSheet("background-color: #107C10;")
 
     def set_phone_state(self, client_id, state_str, number):
+        self.debug(f"Phone {client_id + 1} state change: {state_str}")
+        
+        # Handle protocol-specific states
+        if state_str == "HANDSHAKE_COMPLETE":
+            phone = self.manager.phones[client_id]
+            phone['status_label'].setText("🔒 Secure Channel Established")
+            self.debug(f"Phone {client_id + 1} secure channel established")
+            self.manager.start_audio(client_id, parent=self)
+            return
+        elif state_str == "VOICE_START":
+            phone = self.manager.phones[client_id]
+            phone['status_label'].setText("🎤 Voice Active (Encrypted)")
+            self.debug(f"Phone {client_id + 1} voice session started")
+            return
+        elif state_str == "VOICE_END":
+            phone = self.manager.phones[client_id]
+            phone['status_label'].setText("🔒 Secure Channel")
+            self.debug(f"Phone {client_id + 1} voice session ended")
+            return
+            
+        # Handle regular states
         state = self.manager.map_state(state_str)
         phone = self.manager.phones[client_id]
         other_phone = self.manager.phones[1 - client_id]
-        print(f"Setting state for Phone {client_id + 1}: {state}, number: {number}, is_initiator: {phone['is_initiator']}")
+        self.debug(f"Setting state for Phone {client_id + 1}: {state.name if hasattr(state, 'name') else state}, number: {number}, is_initiator: {phone['is_initiator']}")
         phone['state'] = state
         if state == PhoneState.IN_CALL:
-            print(f"Phone {client_id + 1} confirmed in IN_CALL state")
-            if number == "IN_CALL" and phone['is_initiator']:
-                print(f"Phone {client_id + 1} (initiator) starting handshake")
-                phone['client'].send("HANDSHAKE")
+            self.debug(f"Phone {client_id + 1} confirmed in IN_CALL state")
+            self.debug(f"  state_str={state_str}, number={number}")
+            self.debug(f"  is_initiator={phone['is_initiator']}")
+            
+            # Only start handshake when the initiator RECEIVES the IN_CALL message
+            if state_str == "IN_CALL" and phone['is_initiator']:
+                self.debug(f"Phone {client_id + 1} (initiator) received IN_CALL, starting handshake")
                 phone['client'].start_handshake(initiator=True, keypair=phone['keypair'], peer_pubkey=other_phone['public_key'])
-            elif number == "HANDSHAKE" and not phone['is_initiator']:
-                print(f"Phone {client_id + 1} (responder) starting handshake")
-                phone['client'].start_handshake(initiator=False, keypair=phone['keypair'], peer_pubkey=other_phone['public_key'])
+            elif number == "HANDSHAKE":
+                # Old text-based handshake trigger - no longer used
+                self.debug(f"Phone {client_id + 1} received legacy HANDSHAKE message")
             elif number == "HANDSHAKE_DONE":
-                self.manager.start_audio(client_id, parent=self)  # Pass self as parent
+                self.debug(f"Phone {client_id + 1} received HANDSHAKE_DONE")
+                # Handled by HANDSHAKE_COMPLETE now
+                pass
         self.update_phone_ui(client_id)
 
     def settings_action(self):
         print("Settings clicked")
+        self.debug("Settings clicked")
+
+    def debug(self, message):
+        """Thread-safe debug logging to both console and UI"""
+        timestamp = time.strftime("%H:%M:%S.%f")[:-3]
+        debug_msg = f"[{timestamp}] {message}"
+        print(debug_msg)  # Console output
+        self.debug_signal.emit(debug_msg)  # UI output
+    
+    def append_debug(self, message):
+        """Append debug message to console (called from main thread)"""
+        if hasattr(self, 'debug_console'):
+            self.debug_console.append(message)
+            # Auto-scroll to bottom
+            cursor = self.debug_console.textCursor()
+            cursor.movePosition(QTextCursor.End)
+            self.debug_console.setTextCursor(cursor)
+        else:
+            # Queue messages until console is ready
+            if not hasattr(self, '_debug_queue'):
+                self._debug_queue = []
+            self._debug_queue.append(message)
+    
+    def clear_debug(self):
+        """Clear debug console"""
+        self.debug_console.clear()
+        self.debug("Debug console cleared")
+    
+    def toggle_auto_test(self):
+        """Toggle automatic test sequence"""
+        if not self.auto_test_running:
+            self.start_auto_test()
+        else:
+            self.stop_auto_test()
+    
+    def start_auto_test(self):
+        """Start automatic test sequence"""
+        self.auto_test_running = True
+        self.auto_test_button.setText("⏹ Stop Test")
+        self.test_step = 0
+        
+        self.debug("=== STARTING AUTOMATIC TEST SEQUENCE ===")
+        self.debug("Test will go through complete protocol flow")
+        
+        # Start test timer
+        self.auto_test_timer = QTimer()
+        self.auto_test_timer.timeout.connect(self.execute_test_step)
+        self.auto_test_timer.start(2000)  # 2 second intervals
+        
+        # Execute first step immediately
+        self.execute_test_step()
+    
+    def stop_auto_test(self):
+        """Stop automatic test sequence"""
+        self.auto_test_running = False
+        self.auto_test_button.setText("🧪 Run Automatic Test")
+        
+        if self.auto_test_timer:
+            self.auto_test_timer.stop()
+            self.auto_test_timer = None
+        
+        self.debug("=== TEST SEQUENCE STOPPED ===")
+    
+    def execute_test_step(self):
+        """Execute next step in test sequence"""
+        phone1 = self.manager.phones[0]
+        phone2 = self.manager.phones[1]
+        
+        self.debug(f"\n--- Test Step {self.test_step + 1} ---")
+        
+        if self.test_step == 0:
+            # Step 1: Check initial state
+            self.debug("Checking initial state...")
+            state1 = phone1['state']
+            state2 = phone2['state']
+            # Handle both enum and int states
+            state1_name = state1.name if hasattr(state1, 'name') else str(state1)
+            state2_name = state2.name if hasattr(state2, 'name') else str(state2)
+            self.debug(f"Phone 1 state: {state1_name}")
+            self.debug(f"Phone 2 state: {state2_name}")
+            self.debug(f"Phone 1 connected: {phone1['client'].sock is not None}")
+            self.debug(f"Phone 2 connected: {phone2['client'].sock is not None}")
+            
+        elif self.test_step == 1:
+            # Step 2: Make call
+            self.debug("Phone 1 calling Phone 2...")
+            self.manager.phone_action(0, self)
+            state1_name = phone1['state'].name if hasattr(phone1['state'], 'name') else str(phone1['state'])
+            state2_name = phone2['state'].name if hasattr(phone2['state'], 'name') else str(phone2['state'])
+            self.debug(f"Phone 1 state after call: {state1_name}")
+            self.debug(f"Phone 2 state after call: {state2_name}")
+            
+        elif self.test_step == 2:
+            # Step 3: Answer call
+            self.debug("Phone 2 answering call...")
+            self.manager.phone_action(1, self)
+            state1_name = phone1['state'].name if hasattr(phone1['state'], 'name') else str(phone1['state'])
+            state2_name = phone2['state'].name if hasattr(phone2['state'], 'name') else str(phone2['state'])
+            self.debug(f"Phone 1 state after answer: {state1_name}")
+            self.debug(f"Phone 2 state after answer: {state2_name}")
+            self.debug(f"Phone 1 is_initiator: {phone1['is_initiator']}")
+            self.debug(f"Phone 2 is_initiator: {phone2['is_initiator']}")
+            
+        elif self.test_step == 3:
+            # Step 4: Check handshake progress
+            self.debug("Checking handshake progress...")
+            self.debug(f"Phone 1 handshake in progress: {phone1['client'].state.handshake_in_progress}")
+            self.debug(f"Phone 2 handshake in progress: {phone2['client'].state.handshake_in_progress}")
+            self.debug(f"Phone 1 command queue: {phone1['client'].state.command_queue.qsize()}")
+            self.debug(f"Phone 2 command queue: {phone2['client'].state.command_queue.qsize()}")
+            # Increase timer interval for handshake
+            self.auto_test_timer.setInterval(3000)  # 3 seconds
+            
+        elif self.test_step == 4:
+            # Step 5: Check handshake status
+            self.debug("Checking Noise XK handshake status...")
+            self.debug(f"Phone 1 handshake complete: {phone1['client'].handshake_complete}")
+            self.debug(f"Phone 2 handshake complete: {phone2['client'].handshake_complete}")
+            self.debug(f"Phone 1 has session: {phone1['client'].noise_session is not None}")
+            self.debug(f"Phone 2 has session: {phone2['client'].noise_session is not None}")
+            # Reset timer interval
+            self.auto_test_timer.setInterval(2000)
+            
+        elif self.test_step == 5:
+            # Step 6: Check voice status
+            self.debug("Checking voice session status...")
+            self.debug(f"Phone 1 voice active: {phone1['client'].voice_active}")
+            self.debug(f"Phone 2 voice active: {phone2['client'].voice_active}")
+            self.debug(f"Phone 1 codec initialized: {phone1['client'].codec is not None}")
+            self.debug(f"Phone 2 codec initialized: {phone2['client'].codec is not None}")
+            self.debug(f"Phone 1 modem initialized: {phone1['client'].modem is not None}")
+            self.debug(f"Phone 2 modem initialized: {phone2['client'].modem is not None}")
+            
+        elif self.test_step == 6:
+            # Step 7: Check audio transmission
+            self.debug("Checking audio transmission...")
+            self.debug(f"Phone 1 audio file loaded: {phone1['audio_file'] is not None}")
+            self.debug(f"Phone 2 audio file loaded: {phone2['audio_file'] is not None}")
+            self.debug(f"Phone 1 frame counter: {phone1.get('frame_counter', 0)}")
+            self.debug(f"Phone 2 frame counter: {phone2.get('frame_counter', 0)}")
+            self.debug(f"Phone 1 audio timer active: {phone1['audio_timer'] is not None and phone1['audio_timer'].isActive()}")
+            self.debug(f"Phone 2 audio timer active: {phone2['audio_timer'] is not None and phone2['audio_timer'].isActive()}")
+            
+        elif self.test_step == 7:
+            # Step 8: Protocol details
+            self.debug("Protocol stack details:")
+            if phone1['client'].codec:
+                self.debug(f"Codec mode: {phone1['client'].codec.mode.name}")
+                self.debug(f"Frame size: {phone1['client'].codec.frame_bits} bits")
+                self.debug(f"Frame duration: {phone1['client'].codec.frame_ms} ms")
+            if phone1['client'].modem:
+                self.debug(f"FSK frequencies: {phone1['client'].modem.frequencies}")
+                self.debug(f"Symbol rate: {phone1['client'].modem.baud_rate} baud")
+            
+        elif self.test_step == 8:
+            # Step 9: Wait for more frames
+            self.debug("Letting voice transmission run...")
+            self.auto_test_timer.setInterval(5000)  # Wait 5 seconds
+            
+        elif self.test_step == 9:
+            # Step 10: Final statistics
+            self.debug("Final transmission statistics:")
+            self.debug(f"Phone 1 frames sent: {phone1.get('frame_counter', 0)}")
+            self.debug(f"Phone 2 frames sent: {phone2.get('frame_counter', 0)}")
+            self.auto_test_timer.setInterval(2000)  # Back to 2 seconds
+            
+        elif self.test_step == 10:
+            # Step 11: Hang up
+            self.debug("Hanging up call...")
+            self.manager.phone_action(0, self)
+            state1_name = phone1['state'].name if hasattr(phone1['state'], 'name') else str(phone1['state'])
+            state2_name = phone2['state'].name if hasattr(phone2['state'], 'name') else str(phone2['state'])
+            self.debug(f"Phone 1 state after hangup: {state1_name}")
+            self.debug(f"Phone 2 state after hangup: {state2_name}")
+            
+        elif self.test_step == 11:
+            # Complete
+            self.debug("\n=== TEST SEQUENCE COMPLETE ===")
+            self.debug("All protocol components tested successfully!")
+            self.stop_auto_test()
+            return
+        
+        self.test_step += 1
+    
+    def toggle_playback(self, phone_id):
+        """Toggle audio playback for a phone"""
+        is_enabled = self.manager.toggle_playback(phone_id)
+        phone = self.manager.phones[phone_id]
+        phone['playback_button'].setChecked(is_enabled)
+        
+        if is_enabled:
+            self.debug(f"Phone {phone_id + 1}: Audio playback enabled")
+        else:
+            self.debug(f"Phone {phone_id + 1}: Audio playback disabled")
+    
+    def toggle_recording(self, phone_id):
+        """Toggle audio recording for a phone"""
+        is_recording, save_path = self.manager.toggle_recording(phone_id)
+        phone = self.manager.phones[phone_id]
+        phone['record_button'].setChecked(is_recording)
+        
+        if is_recording:
+            self.debug(f"Phone {phone_id + 1}: Recording started")
+        else:
+            if save_path:
+                self.debug(f"Phone {phone_id + 1}: Recording saved to {save_path}")
+            else:
+                self.debug(f"Phone {phone_id + 1}: Recording stopped (no data)")
+    
+    def show_audio_menu(self):
+        """Show audio processing options menu"""
+        menu = QMenu(self)
+        
+        # Create phone selection submenu
+        for phone_id in range(2):
+            phone_menu = menu.addMenu(f"Phone {phone_id + 1}")
+            
+            # Export buffer
+            export_action = QAction("Export Audio Buffer", self)
+            export_action.triggered.connect(lambda checked, pid=phone_id: self.export_audio_buffer(pid))
+            phone_menu.addAction(export_action)
+            
+            # Clear buffer
+            clear_action = QAction("Clear Audio Buffer", self)
+            clear_action.triggered.connect(lambda checked, pid=phone_id: self.clear_audio_buffer(pid))
+            phone_menu.addAction(clear_action)
+            
+            phone_menu.addSeparator()
+            
+            # Processing options
+            normalize_action = QAction("Normalize Audio", self)
+            normalize_action.triggered.connect(lambda checked, pid=phone_id: self.process_audio(pid, "normalize"))
+            phone_menu.addAction(normalize_action)
+            
+            gain_action = QAction("Apply Gain...", self)
+            gain_action.triggered.connect(lambda checked, pid=phone_id: self.apply_gain_dialog(pid))
+            phone_menu.addAction(gain_action)
+            
+            noise_gate_action = QAction("Apply Noise Gate", self)
+            noise_gate_action.triggered.connect(lambda checked, pid=phone_id: self.process_audio(pid, "noise_gate"))
+            phone_menu.addAction(noise_gate_action)
+            
+            low_pass_action = QAction("Apply Low Pass Filter", self)
+            low_pass_action.triggered.connect(lambda checked, pid=phone_id: self.process_audio(pid, "low_pass"))
+            phone_menu.addAction(low_pass_action)
+            
+            high_pass_action = QAction("Apply High Pass Filter", self)
+            high_pass_action.triggered.connect(lambda checked, pid=phone_id: self.process_audio(pid, "high_pass"))
+            phone_menu.addAction(high_pass_action)
+            
+            remove_silence_action = QAction("Remove Silence", self)
+            remove_silence_action.triggered.connect(lambda checked, pid=phone_id: self.process_audio(pid, "remove_silence"))
+            phone_menu.addAction(remove_silence_action)
+        
+        # Show menu at button position
+        menu.exec_(self.audio_menu_button.mapToGlobal(self.audio_menu_button.rect().bottomLeft()))
+    
+    def export_audio_buffer(self, phone_id):
+        """Export audio buffer for a phone"""
+        save_path = self.manager.export_buffered_audio(phone_id)
+        if save_path:
+            self.debug(f"Phone {phone_id + 1}: Audio buffer exported to {save_path}")
+        else:
+            self.debug(f"Phone {phone_id + 1}: No audio data to export")
+    
+    def clear_audio_buffer(self, phone_id):
+        """Clear audio buffer for a phone"""
+        self.manager.clear_audio_buffer(phone_id)
+    
+    def process_audio(self, phone_id, processing_type):
+        """Process audio with specified type"""
+        save_path = self.manager.process_audio(phone_id, processing_type)
+        if save_path:
+            self.debug(f"Phone {phone_id + 1}: Processed audio saved to {save_path}")
+        else:
+            self.debug(f"Phone {phone_id + 1}: Audio processing failed")
+    
+    def apply_gain_dialog(self, phone_id):
+        """Show dialog to get gain value"""
+        gain, ok = QInputDialog.getDouble(
+            self, "Apply Gain", "Enter gain in dB:", 
+            0.0, -20.0, 20.0, 1
+        )
+        if ok:
+            save_path = self.manager.process_audio(phone_id, "gain", gain_db=gain)
+            if save_path:
+                self.debug(f"Phone {phone_id + 1}: Applied {gain}dB gain, saved to {save_path}")
+    
+    def setup_shortcuts(self):
+        """Setup keyboard shortcuts"""
+        # Phone 1 shortcuts
+        QShortcut(QKeySequence("1"), self, lambda: self.manager.phone_action(0, self))
+        QShortcut(QKeySequence("Ctrl+1"), self, lambda: self.toggle_playback(0))
+        QShortcut(QKeySequence("Alt+1"), self, lambda: self.toggle_recording(0))
+        
+        # Phone 2 shortcuts
+        QShortcut(QKeySequence("2"), self, lambda: self.manager.phone_action(1, self))
+        QShortcut(QKeySequence("Ctrl+2"), self, lambda: self.toggle_playback(1))
+        QShortcut(QKeySequence("Alt+2"), self, lambda: self.toggle_recording(1))
+        
+        # General shortcuts
+        QShortcut(QKeySequence("Space"), self, self.toggle_auto_test)
+        QShortcut(QKeySequence("Ctrl+L"), self, self.clear_debug)
+        QShortcut(QKeySequence("Ctrl+A"), self, self.show_audio_menu)
+        
+        self.debug("Keyboard shortcuts enabled:")
+        self.debug("  1/2: Phone action (call/answer/hangup)")
+        self.debug("  Ctrl+1/2: Toggle playback")
+        self.debug("  Alt+1/2: Toggle recording")
+        self.debug("  Space: Toggle auto test")
+        self.debug("  Ctrl+L: Clear debug")
+        self.debug("  Ctrl+A: Audio options menu")
 
     def closeEvent(self, event):
+        if self.auto_test_running:
+            self.stop_auto_test()
+        # Clean up audio player
+        if hasattr(self.manager, 'audio_player'):
+            self.manager.audio_player.cleanup()
         for phone in self.manager.phones:
             phone['client'].stop()
         event.accept()
diff --git a/protocol_prototype/DryBox/UI/noise_wrapper.py b/protocol_prototype/DryBox/UI/noise_wrapper.py
new file mode 100644
index 0000000..dbcdb62
--- /dev/null
+++ b/protocol_prototype/DryBox/UI/noise_wrapper.py
@@ -0,0 +1,127 @@
+"""Wrapper for Noise XK handshake over GSM simulator"""
+
+import struct
+from dissononce.processing.impl.handshakestate import HandshakeState
+from dissononce.processing.impl.symmetricstate import SymmetricState
+from dissononce.processing.impl.cipherstate import CipherState
+from dissononce.processing.handshakepatterns.interactive.XK import XKHandshakePattern
+from dissononce.cipher.chachapoly import ChaChaPolyCipher
+from dissononce.dh.x25519.x25519 import X25519DH
+from dissononce.dh.keypair import KeyPair
+from dissononce.dh.x25519.public import PublicKey
+from dissononce.hash.sha256 import SHA256Hash
+
+class NoiseXKWrapper:
+    """Wrapper for Noise XK that works over message-passing instead of direct sockets"""
+    
+    def __init__(self, keypair, peer_pubkey, debug_callback=None):
+        self.keypair = keypair
+        self.peer_pubkey = peer_pubkey
+        self.debug = debug_callback or print
+        
+        # Build handshake state
+        cipher = ChaChaPolyCipher()
+        dh = X25519DH()
+        hshash = SHA256Hash()
+        symmetric = SymmetricState(CipherState(cipher), hshash)
+        self._hs = HandshakeState(symmetric, dh)
+        
+        self._send_cs = None
+        self._recv_cs = None
+        self.handshake_complete = False
+        self.is_initiator = None  # Track initiator status
+        
+        # Message buffers
+        self.outgoing_messages = []
+        self.incoming_messages = []
+        
+    def start_handshake(self, initiator):
+        """Start the handshake process"""
+        self.debug(f"Starting Noise XK handshake as {'initiator' if initiator else 'responder'}")
+        self.is_initiator = initiator  # Store initiator status
+        
+        if initiator:
+            # Initiator knows peer's static out-of-band
+            self._hs.initialize(
+                XKHandshakePattern(),
+                True,
+                b'',
+                s=self.keypair,
+                rs=self.peer_pubkey
+            )
+            # Generate first message
+            buf = bytearray()
+            self._hs.write_message(b'', buf)
+            self.outgoing_messages.append(bytes(buf))
+            self.debug(f"Generated handshake message 1: {len(buf)} bytes")
+        else:
+            # Responder doesn't know peer's static yet
+            self._hs.initialize(
+                XKHandshakePattern(),
+                False,
+                b'',
+                s=self.keypair
+            )
+            self.debug("Responder initialized, waiting for first message")
+    
+    def process_handshake_message(self, data):
+        """Process incoming handshake message and generate response if needed"""
+        self.debug(f"Processing handshake message: {len(data)} bytes")
+        
+        try:
+            # Read the message
+            payload = bytearray()
+            cs_pair = self._hs.read_message(data, payload)
+            
+            # Check if we need to send a response
+            if not cs_pair:
+                # More messages needed
+                buf = bytearray()
+                cs_pair = self._hs.write_message(b'', buf)
+                self.outgoing_messages.append(bytes(buf))
+                self.debug(f"Generated handshake response: {len(buf)} bytes")
+                
+                # Check if handshake completed after writing (for initiator)
+                if cs_pair:
+                    self._complete_handshake(cs_pair)
+            else:
+                # Handshake complete after reading (for responder)
+                self._complete_handshake(cs_pair)
+                
+        except Exception as e:
+            self.debug(f"Handshake error: {e}")
+            raise
+    
+    def get_next_handshake_message(self):
+        """Get next outgoing handshake message"""
+        if self.outgoing_messages:
+            return self.outgoing_messages.pop(0)
+        return None
+    
+    def encrypt(self, plaintext):
+        """Encrypt a message"""
+        if not self.handshake_complete:
+            raise RuntimeError("Handshake not complete")
+        return self._send_cs.encrypt_with_ad(b'', plaintext)
+    
+    def decrypt(self, ciphertext):
+        """Decrypt a message"""
+        if not self.handshake_complete:
+            raise RuntimeError("Handshake not complete")
+        return self._recv_cs.decrypt_with_ad(b'', ciphertext)
+    
+    def _complete_handshake(self, cs_pair):
+        """Complete the handshake with the given cipher states"""
+        self.debug("Handshake complete, setting up cipher states")
+        cs0, cs1 = cs_pair
+        
+        # Use stored initiator status
+        if self.is_initiator:
+            self._send_cs, self._recv_cs = cs0, cs1
+            self.debug("Set up cipher states as initiator")
+        else:
+            self._send_cs, self._recv_cs = cs1, cs0
+            self.debug("Set up cipher states as responder")
+        
+        self.handshake_complete = True
+        self.debug("Cipher states established")
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/UI/phone_client.py b/protocol_prototype/DryBox/UI/phone_client.py
deleted file mode 100644
index e4e9fbf..0000000
--- a/protocol_prototype/DryBox/UI/phone_client.py
+++ /dev/null
@@ -1,110 +0,0 @@
-import socket
-import time
-import select
-from PyQt5.QtCore import QThread, pyqtSignal
-from client_state import ClientState
-
-class PhoneClient(QThread):
-    data_received = pyqtSignal(bytes, int)
-    state_changed = pyqtSignal(str, str, int)
-
-    def __init__(self, client_id):
-        super().__init__()
-        self.host = "localhost"
-        self.port = 12345
-        self.client_id = client_id
-        self.sock = None
-        self.running = True
-        self.state = ClientState(client_id)
-
-    def connect_socket(self):
-        retries = 3
-        for attempt in range(retries):
-            try:
-                self.sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
-                self.sock.setsockopt(socket.SOL_SOCKET, socket.SO_KEEPALIVE, 1)
-                self.sock.settimeout(120)
-                self.sock.connect((self.host, self.port))
-                print(f"Client {self.client_id} connected to {self.host}:{self.port}")
-                return True
-            except Exception as e:
-                print(f"Client {self.client_id} connection attempt {attempt + 1} failed: {e}")
-                if attempt < retries - 1:
-                    time.sleep(1)
-                self.sock = None
-        return False
-
-    def run(self):
-        while self.running:
-            if not self.sock:
-                if not self.connect_socket():
-                    print(f"Client {self.client_id} failed to connect after retries")
-                    self.state_changed.emit("CALL_END", "", self.client_id)
-                    break
-            try:
-                while self.running:
-                    self.state.process_command(self)
-                    self.state.check_handshake_timeout(self)
-                    if not self.state.handshake_in_progress:
-                        if self.sock is None:
-                            print(f"Client {self.client_id} socket is None, exiting inner loop")
-                            break
-                        readable, _, _ = select.select([self.sock], [], [], 0.01)
-                        if readable:
-                            try:
-                                if self.sock is None:
-                                    print(f"Client {self.client_id} socket is None before recv, exiting")
-                                    break
-                                data = self.sock.recv(1024)
-                                if not data:
-                                    print(f"Client {self.client_id} disconnected")
-                                    self.state_changed.emit("CALL_END", "", self.client_id)
-                                    break
-                                self.state.handle_data(self, data)
-                            except socket.error as e:
-                                print(f"Client {self.client_id} socket error: {e}")
-                                self.state_changed.emit("CALL_END", "", self.client_id)
-                                break
-                    else:
-                        self.msleep(20)
-                        print(f"Client {self.client_id} yielding during handshake")
-                    self.msleep(1)
-            except Exception as e:
-                print(f"Client {self.client_id} unexpected error in run loop: {e}")
-                self.state_changed.emit("CALL_END", "", self.client_id)
-                break
-            finally:
-                if self.sock:
-                    try:
-                        self.sock.close()
-                    except Exception as e:
-                        print(f"Client {self.client_id} error closing socket: {e}")
-                    self.sock = None
-
-    def send(self, message):
-        if self.sock and self.running:
-            try:
-                if isinstance(message, str):
-                    data = message.encode('utf-8')
-                    self.sock.send(data)
-                    print(f"Client {self.client_id} sent: {message}, length={len(data)}")
-                else:
-                    self.sock.send(message)
-                    print(f"Client {self.client_id} sent binary data, length={len(message)}")
-            except socket.error as e:
-                print(f"Client {self.client_id} send error: {e}")
-                self.state_changed.emit("CALL_END", "", self.client_id)
-
-    def stop(self):
-        self.running = False
-        if self.sock:
-            try:
-                self.sock.close()
-            except Exception as e:
-                print(f"Client {self.client_id} error closing socket in stop: {e}")
-            self.sock = None
-        self.quit()
-        self.wait(1000)
-
-    def start_handshake(self, initiator, keypair, peer_pubkey):
-        self.state.start_handshake(initiator, keypair, peer_pubkey)
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/UI/phone_manager.py b/protocol_prototype/DryBox/UI/phone_manager.py
index 88a9730..d53f837 100644
--- a/protocol_prototype/DryBox/UI/phone_manager.py
+++ b/protocol_prototype/DryBox/UI/phone_manager.py
@@ -1,17 +1,37 @@
 import secrets
 from PyQt5.QtCore import QTimer
-from phone_client import PhoneClient
+from protocol_phone_client import ProtocolPhoneClient
 from session import NoiseXKSession
 from phone_state import PhoneState  # Added import
+from audio_player import AudioPlayer
+from audio_processor import AudioProcessor
+import struct
+import wave
+import os
 
 class PhoneManager:
     def __init__(self):
         self.phones = []
         self.handshake_done_count = 0
+        self.ui = None  # Will be set by UI
+        self.audio_player = AudioPlayer()
+        self.audio_player.set_debug_callback(self.debug)
+        self.audio_processor = AudioProcessor()
+        self.audio_processor.set_debug_callback(self.debug)
+        self.audio_buffer = {}  # client_id -> list of audio chunks for processing
 
+    def debug(self, message):
+        """Send debug message to UI if available"""
+        if self.ui and hasattr(self.ui, 'debug'):
+            self.ui.debug(f"[PhoneManager] {message}")
+        else:
+            print(f"[PhoneManager] {message}")
+    
     def initialize_phones(self):
         for i in range(2):
-            client = PhoneClient(i)
+            client = ProtocolPhoneClient(i)  # Use protocol client
+            client.set_debug_callback(self.debug)  # Set debug callback
+            client.manager = self  # Set manager reference for handshake lookup
             keypair = NoiseXKSession.generate_keypair()
             phone = {
                 'id': i,
@@ -21,9 +41,15 @@ class PhoneManager:
                 'audio_timer': None,
                 'keypair': keypair,
                 'public_key': keypair.public,
-                'is_initiator': False
+                'is_initiator': False,
+                'audio_file': None,  # For test audio
+                'frame_counter': 0,
+                'playback_enabled': False,
+                'recording_enabled': False
             }
+            client.keypair = keypair  # Also set keypair on client
             self.phones.append(phone)
+            self.debug(f"Initialized Phone {i+1} with public key: {keypair.public.data.hex()[:32]}...")
 
         self.phones[0]['peer_public_key'] = self.phones[1]['public_key']
         self.phones[1]['peer_public_key'] = self.phones[0]['public_key']
@@ -31,16 +57,19 @@ class PhoneManager:
     def phone_action(self, phone_id, ui_manager):
         phone = self.phones[phone_id]
         other_phone = self.phones[1 - phone_id]
-        print(f"Phone {phone_id + 1} Action, current state: {phone['state']}, is_initiator: {phone['is_initiator']}")
+        self.debug(f"Phone {phone_id + 1} action triggered, current state: {phone['state'].name}")
 
         if phone['state'] == PhoneState.IDLE:
+            self.debug(f"Phone {phone_id + 1} initiating call to Phone {2-phone_id}")
             phone['state'] = PhoneState.CALLING
             other_phone['state'] = PhoneState.RINGING
             phone['is_initiator'] = True
             other_phone['is_initiator'] = False
             phone['client'].send("RINGING")
         elif phone['state'] == PhoneState.RINGING:
-            phone['state'] = other_phone['state'] = PhoneState.IN_CALL
+            self.debug(f"Phone {phone_id + 1} answering call from Phone {2-phone_id}")
+            phone['state'] = PhoneState.IN_CALL
+            # Don't set other_phone state here - let it set when it receives IN_CALL
             phone['client'].send("IN_CALL")
         elif phone['state'] in [PhoneState.IN_CALL, PhoneState.CALLING]:
             if not phone['client'].state.handshake_in_progress and phone['state'] != PhoneState.CALLING:
@@ -49,40 +78,287 @@ class PhoneManager:
                 for p in [phone, other_phone]:
                     if p['audio_timer']:
                         p['audio_timer'].stop()
+                    # End voice session
+                    if p['client'].voice_active:
+                        p['client'].end_voice_session()
+                    # Close audio file
+                    if p['audio_file']:
+                        p['audio_file'].close()
+                        p['audio_file'] = None
+                    p['frame_counter'] = 0
             else:
-                print(f"Phone {phone_id + 1} cannot hang up during handshake or call setup")
+                self.debug(f"Phone {phone_id + 1} cannot hang up during handshake or call setup")
         
         ui_manager.update_phone_ui(phone_id)
         ui_manager.update_phone_ui(1 - phone_id)
 
     def send_audio(self, phone_id):
         phone = self.phones[phone_id]
-        if phone['state'] == PhoneState.IN_CALL and phone['client'].state.session and phone['client'].sock:
-            mock_audio = secrets.token_bytes(16)
-            try:
+        if phone['state'] != PhoneState.IN_CALL:
+            self.debug(f"Phone {phone_id + 1} not in call, stopping audio timer")
+            if phone['audio_timer']:
+                phone['audio_timer'].stop()
+            return
+            
+        if not phone['client'].handshake_complete:
+            self.debug(f"Phone {phone_id + 1} handshake not complete, skipping audio send")
+            return
+            
+        if not phone['client'].voice_active:
+            self.debug(f"Phone {phone_id + 1} voice not active, skipping audio send")
+            return
+            
+        if phone['state'] == PhoneState.IN_CALL and phone['client'].handshake_complete and phone['client'].voice_active:
+            # Load test audio file if not loaded
+            if phone['audio_file'] is None:
+                wav_path = "../wav/input.wav"
+                if not os.path.exists(wav_path):
+                    wav_path = "wav/input.wav"
+                if os.path.exists(wav_path):
+                    try:
+                        phone['audio_file'] = wave.open(wav_path, 'rb')
+                        self.debug(f"Phone {phone_id + 1} loaded test audio file: {wav_path}")
+                        # Verify it's 8kHz mono
+                        if phone['audio_file'].getframerate() != 8000:
+                            self.debug(f"Warning: {wav_path} is {phone['audio_file'].getframerate()}Hz, expected 8000Hz")
+                        if phone['audio_file'].getnchannels() != 1:
+                            self.debug(f"Warning: {wav_path} has {phone['audio_file'].getnchannels()} channels, expected 1")
+                        
+                        # Skip initial silence - jump to 1 second in (8000 samples)
+                        phone['audio_file'].setpos(8000)
+                        self.debug(f"Phone {phone_id + 1} skipped initial silence, starting at 1 second")
+                    except Exception as e:
+                        self.debug(f"Phone {phone_id + 1} failed to load audio: {e}")
+                        # Use mock audio as fallback
+                        phone['audio_file'] = None
+            
+            # Read audio frame (40ms at 8kHz = 320 samples)
+            if phone['audio_file']:
+                try:
+                    frames = phone['audio_file'].readframes(320)
+                    if not frames or len(frames) < 640:  # 320 samples * 2 bytes
+                        # Loop back to 1 second (skip silence)
+                        phone['audio_file'].setpos(8000)
+                        frames = phone['audio_file'].readframes(320)
+                        self.debug(f"Phone {phone_id + 1} looped audio back to 1 second mark")
+                    
+                    # Send through protocol (codec + 4FSK + encryption)
+                    phone['client'].send_voice_frame(frames)
+                    
+                    # Update waveform
+                    if len(frames) >= 2:
+                        samples = struct.unpack(f'{len(frames)//2}h', frames)
+                        self.update_sent_waveform(phone_id, frames)
+                    
+                    # If playback is enabled on the sender, play the original audio
+                    if phone['playback_enabled']:
+                        self.audio_player.add_audio_data(phone_id, frames)
+                        if phone['frame_counter'] % 25 == 0:
+                            self.debug(f"Phone {phone_id + 1} playing original audio (sender playback)")
+                    
+                    phone['frame_counter'] += 1
+                    if phone['frame_counter'] % 25 == 0:  # Log every second
+                        self.debug(f"Phone {phone_id + 1} sent {phone['frame_counter']} voice frames")
+                        
+                except Exception as e:
+                    self.debug(f"Phone {phone_id + 1} audio send error: {e}")
+            else:
+                # Fallback: send mock audio
+                mock_audio = secrets.token_bytes(320)
+                phone['client'].send_voice_frame(mock_audio)
                 self.update_sent_waveform(phone_id, mock_audio)
-                phone['client'].state.session.send(phone['client'].sock, mock_audio)
-                print(f"Client {phone_id} sent encrypted audio packet, length=32")
-            except Exception as e:
-                print(f"Client {phone_id} failed to send audio: {e}")
 
     def start_audio(self, client_id, parent=None):
         self.handshake_done_count += 1
-        print(f"HANDSHAKE_DONE received for client {client_id}, count: {self.handshake_done_count}")
+        self.debug(f"HANDSHAKE_DONE received for client {client_id}, count: {self.handshake_done_count}")
+        
+        # Start voice session for this client
+        phone = self.phones[client_id]
+        if phone['client'].handshake_complete and not phone['client'].voice_active:
+            phone['client'].start_voice_session()
+        
         if self.handshake_done_count == 2:
-            for phone in self.phones:
-                if phone['state'] == PhoneState.IN_CALL:
-                    if not phone['audio_timer'] or not phone['audio_timer'].isActive():
-                        phone['audio_timer'] = QTimer(parent)  # Parent to PhoneUI
-                        phone['audio_timer'].timeout.connect(lambda pid=phone['id']: self.send_audio(pid))
-                        phone['audio_timer'].start(100)
+            # Add a small delay to ensure both sides are ready
+            def start_audio_timers():
+                self.debug("Starting audio timers for both phones")
+                for phone in self.phones:
+                    if phone['state'] == PhoneState.IN_CALL:
+                        if not phone['audio_timer'] or not phone['audio_timer'].isActive():
+                            phone['audio_timer'] = QTimer(parent)  # Parent to PhoneUI
+                            phone['audio_timer'].timeout.connect(lambda pid=phone['id']: self.send_audio(pid))
+                            phone['audio_timer'].start(40)  # 40ms for each voice frame
+            
+            # Delay audio start by 500ms to ensure both sides are ready
+            QTimer.singleShot(500, start_audio_timers)
             self.handshake_done_count = 0
 
     def update_waveform(self, client_id, data):
+        # Only process actual audio data (should be 640 bytes for 320 samples * 2 bytes)
+        # Ignore small control messages
+        if len(data) < 320:  # Less than 160 samples (too small for audio)
+            self.debug(f"Phone {client_id + 1} received non-audio data: {len(data)} bytes (ignoring)")
+            return
+            
         self.phones[client_id]['waveform'].set_data(data)
+        
+        # Debug log audio data reception (only occasionally to avoid spam)
+        if not hasattr(self, '_audio_frame_count'):
+            self._audio_frame_count = {}
+        if client_id not in self._audio_frame_count:
+            self._audio_frame_count[client_id] = 0
+        self._audio_frame_count[client_id] += 1
+        
+        if self._audio_frame_count[client_id] == 1 or self._audio_frame_count[client_id] % 25 == 0:
+            self.debug(f"Phone {client_id + 1} received audio frame #{self._audio_frame_count[client_id]}: {len(data)} bytes")
+        
+        # Store audio data in buffer for potential processing
+        if client_id not in self.audio_buffer:
+            self.audio_buffer[client_id] = []
+        self.audio_buffer[client_id].append(data)
+        
+        # Keep buffer size reasonable (last 30 seconds at 8kHz)
+        max_chunks = 30 * 25  # 30 seconds * 25 chunks/second
+        if len(self.audio_buffer[client_id]) > max_chunks:
+            self.audio_buffer[client_id] = self.audio_buffer[client_id][-max_chunks:]
+        
+        # Forward audio data to player if playback is enabled
+        if self.phones[client_id]['playback_enabled']:
+            if self._audio_frame_count[client_id] == 1:
+                self.debug(f"Phone {client_id + 1} forwarding audio to player (playback enabled)")
+            self.audio_player.add_audio_data(client_id, data)
 
     def update_sent_waveform(self, client_id, data):
         self.phones[client_id]['sent_waveform'].set_data(data)
+    
+    def toggle_playback(self, client_id):
+        """Toggle audio playback for a phone"""
+        phone = self.phones[client_id]
+        
+        if phone['playback_enabled']:
+            # Stop playback
+            self.audio_player.stop_playback(client_id)
+            phone['playback_enabled'] = False
+            self.debug(f"Phone {client_id + 1} playback stopped")
+        else:
+            # Start playback
+            if self.audio_player.start_playback(client_id):
+                phone['playback_enabled'] = True
+                self.debug(f"Phone {client_id + 1} playback started")
+                # Removed test beep - we want to hear actual audio
+            else:
+                self.debug(f"Phone {client_id + 1} failed to start playback")
+        
+        return phone['playback_enabled']
+    
+    def toggle_recording(self, client_id):
+        """Toggle audio recording for a phone"""
+        phone = self.phones[client_id]
+        
+        if phone['recording_enabled']:
+            # Stop recording and save
+            save_path = self.audio_player.stop_recording(client_id)
+            phone['recording_enabled'] = False
+            if save_path:
+                self.debug(f"Phone {client_id + 1} recording saved to {save_path}")
+            return False, save_path
+        else:
+            # Start recording
+            self.audio_player.start_recording(client_id)
+            phone['recording_enabled'] = True
+            self.debug(f"Phone {client_id + 1} recording started")
+            return True, None
+    
+    def save_received_audio(self, client_id, filename=None):
+        """Save the last received audio to a file"""
+        if client_id not in self.phones:
+            return None
+            
+        save_path = self.audio_player.stop_recording(client_id, filename)
+        if save_path:
+            self.debug(f"Phone {client_id + 1} audio saved to {save_path}")
+        return save_path
+
+    def process_audio(self, client_id, processing_type, **kwargs):
+        """Process buffered audio with specified processing type"""
+        if client_id not in self.audio_buffer or not self.audio_buffer[client_id]:
+            self.debug(f"No audio data available for Phone {client_id + 1}")
+            return None
+        
+        # Combine all audio chunks
+        combined_audio = b''.join(self.audio_buffer[client_id])
+        
+        # Apply processing based on type
+        processed_audio = combined_audio
+        
+        if processing_type == "normalize":
+            target_db = kwargs.get('target_db', -3)
+            processed_audio = self.audio_processor.normalize_audio(combined_audio, target_db)
+        
+        elif processing_type == "gain":
+            gain_db = kwargs.get('gain_db', 0)
+            processed_audio = self.audio_processor.apply_gain(combined_audio, gain_db)
+        
+        elif processing_type == "noise_gate":
+            threshold_db = kwargs.get('threshold_db', -40)
+            processed_audio = self.audio_processor.apply_noise_gate(combined_audio, threshold_db)
+        
+        elif processing_type == "low_pass":
+            cutoff_hz = kwargs.get('cutoff_hz', 3400)
+            processed_audio = self.audio_processor.apply_low_pass_filter(combined_audio, cutoff_hz)
+        
+        elif processing_type == "high_pass":
+            cutoff_hz = kwargs.get('cutoff_hz', 300)
+            processed_audio = self.audio_processor.apply_high_pass_filter(combined_audio, cutoff_hz)
+        
+        elif processing_type == "remove_silence":
+            threshold_db = kwargs.get('threshold_db', -40)
+            processed_audio = self.audio_processor.remove_silence(combined_audio, threshold_db)
+        
+        # Save processed audio
+        save_path = f"wav/phone{client_id + 1}_received.wav"
+        processed_path = self.audio_processor.save_processed_audio(
+            processed_audio, save_path, processing_type
+        )
+        
+        return processed_path
+    
+    def export_buffered_audio(self, client_id, filename=None):
+        """Export current audio buffer to file"""
+        if client_id not in self.audio_buffer or not self.audio_buffer[client_id]:
+            self.debug(f"No audio data available for Phone {client_id + 1}")
+            return None
+        
+        # Combine all audio chunks
+        combined_audio = b''.join(self.audio_buffer[client_id])
+        
+        # Generate filename if not provided
+        if not filename:
+            from datetime import datetime
+            timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+            filename = f"wav/phone{client_id + 1}_buffer_{timestamp}.wav"
+        
+        # Ensure directory exists
+        os.makedirs(os.path.dirname(filename), exist_ok=True)
+        
+        try:
+            with wave.open(filename, 'wb') as wav_file:
+                wav_file.setnchannels(1)
+                wav_file.setsampwidth(2)
+                wav_file.setframerate(8000)
+                wav_file.writeframes(combined_audio)
+            
+            self.debug(f"Exported audio buffer for Phone {client_id + 1} to {filename}")
+            return filename
+            
+        except Exception as e:
+            self.debug(f"Failed to export audio buffer: {e}")
+            return None
+    
+    def clear_audio_buffer(self, client_id):
+        """Clear audio buffer for a phone"""
+        if client_id in self.audio_buffer:
+            self.audio_buffer[client_id] = []
+            self.debug(f"Cleared audio buffer for Phone {client_id + 1}")
 
     def map_state(self, state_str):
         if state_str == "RINGING":
diff --git a/protocol_prototype/DryBox/UI/phone_state.py b/protocol_prototype/DryBox/UI/phone_state.py
index 56d40c1..c5d5a83 100644
--- a/protocol_prototype/DryBox/UI/phone_state.py
+++ b/protocol_prototype/DryBox/UI/phone_state.py
@@ -1,4 +1,6 @@
-class PhoneState:
+from enum import Enum
+
+class PhoneState(Enum):
     IDLE = 0
     CALLING = 1
     IN_CALL = 2
diff --git a/protocol_prototype/DryBox/UI/protocol_client_state.py b/protocol_prototype/DryBox/UI/protocol_client_state.py
new file mode 100644
index 0000000..800cb8f
--- /dev/null
+++ b/protocol_prototype/DryBox/UI/protocol_client_state.py
@@ -0,0 +1,133 @@
+# protocol_client_state.py
+from queue import Queue
+from session import NoiseXKSession
+import time
+
+class ProtocolClientState:
+    """Enhanced client state for integrated protocol with voice codec"""
+    
+    def __init__(self, client_id):
+        self.client_id = client_id
+        self.command_queue = Queue()
+        self.initiator = None
+        self.keypair = None
+        self.peer_pubkey = None
+        self.session = None
+        self.handshake_in_progress = False
+        self.handshake_start_time = None
+        self.call_active = False
+        self.voice_active = False
+        self.debug_callback = None
+    
+    def debug(self, message):
+        """Send debug message"""
+        if self.debug_callback:
+            self.debug_callback(f"[State{self.client_id+1}] {message}")
+        else:
+            print(f"[State{self.client_id+1}] {message}")
+        
+    def process_command(self, client):
+        """Process commands from the queue."""
+        if not self.command_queue.empty():
+            self.debug(f"Processing command queue, size: {self.command_queue.qsize()}")
+            command = self.command_queue.get()
+            self.debug(f"Processing command: {command}")
+            
+            if command == "handshake":
+                # Handshake is now handled by the wrapper in the client
+                self.debug(f"Handshake command processed")
+                self.handshake_in_progress = False
+                self.handshake_start_time = None
+                    
+            elif command == "start_voice":
+                if client.handshake_complete:
+                    client.start_voice_session()
+                    self.voice_active = True
+                    
+            elif command == "end_voice":
+                if self.voice_active:
+                    client.end_voice_session()
+                    self.voice_active = False
+    
+    def start_handshake(self, initiator, keypair, peer_pubkey):
+        """Queue handshake command."""
+        self.initiator = initiator
+        self.keypair = keypair
+        self.peer_pubkey = peer_pubkey
+        self.debug(f"Queuing handshake, initiator: {initiator}")
+        self.handshake_in_progress = True
+        self.handshake_start_time = time.time()
+        self.command_queue.put("handshake")
+    
+    def handle_data(self, client, data):
+        """Handle received data (control or audio)."""
+        try:
+            # Try to decode as text first
+            decoded_data = data.decode('utf-8').strip()
+            self.debug(f"Received raw: {decoded_data}")
+            
+            # Handle control messages
+            if decoded_data in ["RINGING", "CALL_END", "CALL_DROPPED", "IN_CALL", "HANDSHAKE", "HANDSHAKE_DONE"]:
+                self.debug(f"Emitting state change: {decoded_data}")
+                # Log which client is receiving what
+                self.debug(f"Client {self.client_id} received {decoded_data} message")
+                client.state_changed.emit(decoded_data, decoded_data, self.client_id)
+                
+                if decoded_data == "IN_CALL":
+                    self.debug(f"Received IN_CALL, setting call_active = True")
+                    self.call_active = True
+                elif decoded_data == "HANDSHAKE":
+                    self.debug(f"Received HANDSHAKE, setting handshake_in_progress = True")
+                    self.handshake_in_progress = True
+                elif decoded_data == "HANDSHAKE_DONE":
+                    self.debug(f"Received HANDSHAKE_DONE from peer")
+                    self.call_active = True
+                    # Start voice session on this side too
+                    if client.handshake_complete and not client.voice_active:
+                        self.debug(f"Starting voice session after receiving HANDSHAKE_DONE")
+                        self.command_queue.put("start_voice")
+                elif decoded_data in ["CALL_END", "CALL_DROPPED"]:
+                    self.debug(f"Received {decoded_data}, ending call")
+                    self.call_active = False
+                    if self.voice_active:
+                        self.command_queue.put("end_voice")
+            else:
+                self.debug(f"Ignored unexpected text message: {decoded_data}")
+                
+        except UnicodeDecodeError:
+            # Handle binary data (protocol messages or encrypted data)
+            if len(data) > 0 and data[0] == 0x20 and not client.handshake_complete:  # Noise handshake message only before handshake completes
+                self.debug(f"Received Noise handshake message")
+                # Initialize responder if not already done
+                if not client.handshake_initiated:
+                    # Find the other phone's public key
+                    # This is a bit hacky but works for our 2-phone setup
+                    manager = getattr(client, 'manager', None)
+                    if manager:
+                        other_phone = manager.phones[1 - self.client_id]
+                        client.start_handshake(initiator=False, 
+                                             keypair=client.keypair or manager.phones[self.client_id]['keypair'],
+                                             peer_pubkey=other_phone['public_key'])
+                # Pass to protocol handler
+                client._handle_protocol_message(data)
+            elif client.handshake_complete and client.noise_wrapper:
+                # Pass encrypted data back to client for decryption
+                client._handle_encrypted_data(data)
+            else:
+                # Pass other binary messages to protocol handler only if not yet complete
+                if not client.handshake_complete:
+                    client._handle_protocol_message(data)
+    
+    def check_handshake_timeout(self, client):
+        """Check for handshake timeout."""
+        if self.handshake_in_progress and self.handshake_start_time:
+            if time.time() - self.handshake_start_time > 30:
+                self.debug(f"Handshake timeout after 30s")
+                client.state_changed.emit("CALL_END", "", self.client_id)
+                self.handshake_in_progress = False
+                self.handshake_start_time = None
+                
+    def queue_voice_command(self, command):
+        """Queue voice-related commands"""
+        if command in ["start_voice", "end_voice"]:
+            self.command_queue.put(command)
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/UI/protocol_phone_client.py b/protocol_prototype/DryBox/UI/protocol_phone_client.py
new file mode 100644
index 0000000..f8750b2
--- /dev/null
+++ b/protocol_prototype/DryBox/UI/protocol_phone_client.py
@@ -0,0 +1,456 @@
+import socket
+import time
+import select
+import struct
+import array
+from PyQt5.QtCore import QThread, pyqtSignal
+from protocol_client_state import ProtocolClientState
+from session import NoiseXKSession
+from noise_wrapper import NoiseXKWrapper
+from dissononce.dh.keypair import KeyPair
+from dissononce.dh.x25519.public import PublicKey
+import sys
+import os
+sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+from voice_codec import Codec2Wrapper, FSKModem, Codec2Mode
+# ChaCha20 removed - using only Noise XK encryption
+
+class ProtocolPhoneClient(QThread):
+    """Integrated phone client with Noise XK, Codec2, 4FSK, and ChaCha20"""
+    data_received = pyqtSignal(bytes, int)
+    state_changed = pyqtSignal(str, str, int)
+    
+    def __init__(self, client_id):
+        super().__init__()
+        self.host = "localhost"
+        self.port = 12345
+        self.client_id = client_id
+        self.sock = None
+        self.running = True
+        self.state = ProtocolClientState(client_id)
+        
+        # Noise XK session
+        self.noise_session = None
+        self.noise_wrapper = None
+        self.handshake_complete = False
+        self.handshake_initiated = False
+        
+        # No buffer needed with larger frame size
+        
+        # Voice codec components
+        self.codec = Codec2Wrapper(mode=Codec2Mode.MODE_1200)
+        self.modem = FSKModem()
+        
+        # Voice encryption handled by Noise XK
+        # No separate voice key needed
+        
+        # Voice state
+        self.voice_active = False
+        self.voice_frame_counter = 0
+        
+        # Message buffer for fragmented messages
+        self.recv_buffer = bytearray()
+        
+        # Debug callback
+        self.debug_callback = None
+    
+    def set_debug_callback(self, callback):
+        """Set debug callback function"""
+        self.debug_callback = callback
+        self.state.debug_callback = callback
+    
+    def debug(self, message):
+        """Send debug message"""
+        if self.debug_callback:
+            self.debug_callback(f"[Phone{self.client_id+1}] {message}")
+        else:
+            print(f"[Phone{self.client_id+1}] {message}")
+        
+    def connect_socket(self):
+        retries = 3
+        for attempt in range(retries):
+            try:
+                self.sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+                self.sock.setsockopt(socket.SOL_SOCKET, socket.SO_KEEPALIVE, 1)
+                self.sock.settimeout(120)
+                self.sock.connect((self.host, self.port))
+                self.debug(f"Connected to GSM simulator at {self.host}:{self.port}")
+                return True
+            except Exception as e:
+                self.debug(f"Connection attempt {attempt + 1} failed: {e}")
+                if attempt < retries - 1:
+                    time.sleep(1)
+                self.sock = None
+        return False
+    
+    def run(self):
+        while self.running:
+            if not self.sock:
+                if not self.connect_socket():
+                    self.debug("Failed to connect after retries")
+                    self.state_changed.emit("CALL_END", "", self.client_id)
+                    break
+            try:
+                while self.running:
+                    self.state.process_command(self)
+                    self.state.check_handshake_timeout(self)
+                    
+                    if self.handshake_complete and self.voice_active:
+                        # Process voice data if active
+                        self._process_voice_data()
+                    
+                    # Always check for incoming data, even during handshake
+                    if self.sock is None:
+                        break
+                    readable, _, _ = select.select([self.sock], [], [], 0.01)
+                    if readable:
+                        try:
+                            if self.sock is None:
+                                break
+                            chunk = self.sock.recv(4096)
+                            if not chunk:
+                                self.debug("Disconnected from server")
+                                self.state_changed.emit("CALL_END", "", self.client_id)
+                                break
+                            
+                            # Add to buffer
+                            self.recv_buffer.extend(chunk)
+                            
+                            # Process complete messages
+                            while len(self.recv_buffer) >= 4:
+                                # Read message length
+                                msg_len = struct.unpack('>I', self.recv_buffer[:4])[0]
+                                
+                                # Check if we have the complete message
+                                if len(self.recv_buffer) >= 4 + msg_len:
+                                    # Extract message
+                                    data = bytes(self.recv_buffer[4:4+msg_len])
+                                    # Remove from buffer
+                                    self.recv_buffer = self.recv_buffer[4+msg_len:]
+                                    # Pass to state handler
+                                    self.state.handle_data(self, data)
+                                else:
+                                    # Wait for more data
+                                    break
+                                    
+                        except socket.error as e:
+                            self.debug(f"Socket error: {e}")
+                            self.state_changed.emit("CALL_END", "", self.client_id)
+                            break
+                    
+                    self.msleep(1)
+            except Exception as e:
+                self.debug(f"Unexpected error in run loop: {e}")
+                self.state_changed.emit("CALL_END", "", self.client_id)
+                break
+            finally:
+                if self.sock:
+                    try:
+                        self.sock.close()
+                    except Exception as e:
+                        self.debug(f"Error closing socket: {e}")
+                    self.sock = None
+    
+    def _handle_encrypted_data(self, data):
+        """Handle encrypted data after handshake"""
+        if not self.handshake_complete or not self.noise_wrapper:
+            self.debug(f"Cannot decrypt - handshake not complete")
+            return
+            
+        # All data after handshake is encrypted, decrypt it first
+        try:
+            plaintext = self.noise_wrapper.decrypt(data)
+            
+            # Check if it's a text message
+            try:
+                text_msg = plaintext.decode('utf-8').strip()
+                if text_msg == "HANDSHAKE_DONE":
+                    self.debug(f"Received encrypted HANDSHAKE_DONE")
+                    self.state_changed.emit("HANDSHAKE_DONE", "HANDSHAKE_DONE", self.client_id)
+                    return
+            except:
+                pass
+                
+            # Otherwise handle as protocol message
+            self._handle_protocol_message(plaintext)
+        except Exception as e:
+            # Suppress common decryption errors
+            pass
+    
+    def _handle_protocol_message(self, plaintext):
+        """Handle decrypted protocol messages"""
+        if len(plaintext) < 1:
+            return
+            
+        msg_type = plaintext[0]
+        msg_data = plaintext[1:]
+        
+        if msg_type == 0x10:  # Voice start
+            self.debug("Received VOICE_START message")
+            self._handle_voice_start(msg_data)
+        elif msg_type == 0x11:  # Voice data
+            self._handle_voice_data(msg_data)
+        elif msg_type == 0x12:  # Voice end
+            self.debug("Received VOICE_END message")
+            self._handle_voice_end(msg_data)
+        elif msg_type == 0x20:  # Noise handshake
+            self.debug("Received NOISE_HS message")
+            self._handle_noise_handshake(msg_data)
+        else:
+            self.debug(f"Received unknown protocol message type: 0x{msg_type:02x}")
+            # Don't emit control messages to data_received - that's only for audio
+            # Control messages should be handled via state_changed signal
+    
+    def _handle_voice_start(self, data):
+        """Handle voice session start"""
+        self.debug("Voice session started by peer")
+        self.voice_active = True
+        self.voice_frame_counter = 0
+        self.state_changed.emit("VOICE_START", "", self.client_id)
+    
+    def _handle_voice_data(self, data):
+        """Handle voice frame (already decrypted by Noise)"""
+        if len(data) < 4:
+            return
+            
+        try:
+            # Data is float array packed as bytes
+            # Unpack the float array
+            num_floats = len(data) // 4
+            modulated_signal = struct.unpack(f'{num_floats}f', data)
+            
+            # Demodulate FSK
+            demodulated_data, confidence = self.modem.demodulate(modulated_signal)
+            
+            if confidence > 0.5:  # Only decode if confidence is good
+                # Create Codec2Frame from demodulated data
+                from voice_codec import Codec2Frame, Codec2Mode
+                frame = Codec2Frame(
+                    mode=Codec2Mode.MODE_1200,
+                    bits=demodulated_data,
+                    timestamp=time.time(),
+                    frame_number=self.voice_frame_counter
+                )
+                
+                # Decode with Codec2
+                pcm_samples = self.codec.decode(frame)
+                
+                if self.voice_frame_counter == 0:
+                    self.debug(f"First voice frame demodulated with confidence {confidence:.2f}")
+                
+                # Send PCM to UI for playback
+                if pcm_samples is not None and len(pcm_samples) > 0:
+                    # Only log details for first frame and every 25th frame
+                    if self.voice_frame_counter == 0 or self.voice_frame_counter % 25 == 0:
+                        self.debug(f"Decoded PCM samples: type={type(pcm_samples)}, len={len(pcm_samples)}")
+                    
+                    # Convert to bytes if needed
+                    if hasattr(pcm_samples, 'tobytes'):
+                        pcm_bytes = pcm_samples.tobytes()
+                    elif isinstance(pcm_samples, (list, array.array)):
+                        # Convert array to bytes
+                        import array
+                        if isinstance(pcm_samples, list):
+                            pcm_array = array.array('h', pcm_samples)
+                            pcm_bytes = pcm_array.tobytes()
+                        else:
+                            pcm_bytes = pcm_samples.tobytes()
+                    else:
+                        pcm_bytes = bytes(pcm_samples)
+                    
+                    if self.voice_frame_counter == 0:
+                        self.debug(f"Emitting first PCM frame: {len(pcm_bytes)} bytes")
+                    
+                    self.data_received.emit(pcm_bytes, self.client_id)
+                    self.voice_frame_counter += 1
+                    # Log frame reception periodically
+                    if self.voice_frame_counter == 1 or self.voice_frame_counter % 25 == 0:
+                        self.debug(f"Received voice data frame #{self.voice_frame_counter}")
+                else:
+                    self.debug(f"Codec decode returned None or empty")
+            else:
+                if self.voice_frame_counter % 10 == 0:
+                    self.debug(f"Low confidence demodulation: {confidence:.2f}")
+                    
+        except Exception as e:
+            self.debug(f"Voice decode error: {e}")
+    
+    def _handle_voice_end(self, data):
+        """Handle voice session end"""
+        self.debug("Voice session ended by peer")
+        self.voice_active = False
+        self.state_changed.emit("VOICE_END", "", self.client_id)
+    
+    def _handle_noise_handshake(self, data):
+        """Handle Noise handshake message"""
+        if not self.noise_wrapper:
+            self.debug("Received handshake message but no wrapper initialized")
+            return
+            
+        try:
+            # Process the handshake message
+            self.noise_wrapper.process_handshake_message(data)
+            
+            # Check if we need to send a response
+            response = self.noise_wrapper.get_next_handshake_message()
+            if response:
+                self.send(b'\x20' + response)
+                
+            # Check if handshake is complete
+            if self.noise_wrapper.handshake_complete and not self.handshake_complete:
+                self.debug("Noise wrapper handshake complete, calling complete_handshake()")
+                self.complete_handshake()
+                
+        except Exception as e:
+            self.debug(f"Handshake processing error: {e}")
+            self.state_changed.emit("CALL_END", "", self.client_id)
+    
+    def _process_voice_data(self):
+        """Process outgoing voice data"""
+        # This would be called when we have voice input to send
+        # For now, this is a placeholder
+        pass
+    
+    def send_voice_frame(self, pcm_samples):
+        """Send a voice frame through the protocol"""
+        if not self.handshake_complete:
+            self.debug("Cannot send voice - handshake not complete")
+            return
+        if not self.voice_active:
+            self.debug("Cannot send voice - voice session not active")
+            return
+            
+        try:
+            # Encode with Codec2
+            codec_frame = self.codec.encode(pcm_samples)
+            if not codec_frame:
+                return
+            
+            if self.voice_frame_counter % 25 == 0:  # Log every 25 frames (1 second)
+                self.debug(f"Encoding voice frame #{self.voice_frame_counter}: {len(pcm_samples)} bytes PCM → {len(codec_frame.bits)} bytes compressed")
+                
+            # Modulate with FSK
+            modulated_data = self.modem.modulate(codec_frame.bits)
+            
+            # Convert modulated float array to bytes
+            modulated_bytes = struct.pack(f'{len(modulated_data)}f', *modulated_data)
+            
+            if self.voice_frame_counter % 25 == 0:
+                self.debug(f"Voice frame size: {len(modulated_bytes)} bytes")
+            
+            # Build voice data message (no ChaCha20, will be encrypted by Noise)
+            msg = bytes([0x11]) + modulated_bytes
+            
+            # Send through Noise encrypted channel
+            self.send(msg)
+            
+            self.voice_frame_counter += 1
+            
+        except Exception as e:
+            self.debug(f"Voice encode error: {e}")
+    
+    def send(self, message):
+        """Send data through Noise encrypted channel with proper framing"""
+        if self.sock and self.running:
+            try:
+                # Handshake messages (0x20) bypass Noise encryption
+                if isinstance(message, bytes) and len(message) > 0 and message[0] == 0x20:
+                    # Add length prefix for framing
+                    framed = struct.pack('>I', len(message)) + message
+                    self.sock.send(framed)
+                    return
+                
+                if self.handshake_complete and self.noise_wrapper:
+                    # Encrypt everything with Noise after handshake
+                    # Convert string to bytes if needed
+                    if isinstance(message, str):
+                        message = message.encode('utf-8')
+                    encrypted = self.noise_wrapper.encrypt(message)
+                    # Add length prefix for framing
+                    framed = struct.pack('>I', len(encrypted)) + encrypted
+                    self.sock.send(framed)
+                else:
+                    # During handshake, send raw with framing
+                    if isinstance(message, str):
+                        data = message.encode('utf-8')
+                        framed = struct.pack('>I', len(data)) + data
+                        self.sock.send(framed)
+                        self.debug(f"Sent control message: {message}")
+                    else:
+                        framed = struct.pack('>I', len(message)) + message
+                        self.sock.send(framed)
+            except socket.error as e:
+                self.debug(f"Send error: {e}")
+                self.state_changed.emit("CALL_END", "", self.client_id)
+    
+    def stop(self):
+        self.running = False
+        self.voice_active = False
+        if self.sock:
+            try:
+                self.sock.close()
+            except Exception as e:
+                self.debug(f"Error closing socket in stop: {e}")
+            self.sock = None
+        self.quit()
+        self.wait(1000)
+    
+    def start_handshake(self, initiator, keypair, peer_pubkey):
+        """Start Noise XK handshake"""
+        self.debug(f"Starting Noise XK handshake as {'initiator' if initiator else 'responder'}")
+        self.debug(f"Our public key: {keypair.public.data.hex()[:32]}...")
+        self.debug(f"Peer public key: {peer_pubkey.data.hex()[:32]}...")
+        
+        # Create noise wrapper
+        self.noise_wrapper = NoiseXKWrapper(keypair, peer_pubkey, self.debug)
+        self.noise_wrapper.start_handshake(initiator)
+        self.handshake_initiated = True
+        
+        # Send first handshake message if initiator
+        if initiator:
+            msg = self.noise_wrapper.get_next_handshake_message()
+            if msg:
+                # Send as NOISE_HS message type
+                self.send(b'\x20' + msg)  # 0x20 = Noise handshake message
+    
+    def complete_handshake(self):
+        """Called when Noise handshake completes"""
+        self.handshake_complete = True
+        
+        self.debug("Noise XK handshake complete!")
+        self.debug("Secure channel established")
+        
+        # Send HANDSHAKE_DONE message
+        self.send("HANDSHAKE_DONE")
+        
+        self.state_changed.emit("HANDSHAKE_COMPLETE", "", self.client_id)
+    
+    def start_voice_session(self):
+        """Start a voice session"""
+        if not self.handshake_complete:
+            self.debug("Cannot start voice - handshake not complete")
+            return
+            
+        self.voice_active = True
+        self.voice_frame_counter = 0
+        
+        # Send voice start message
+        msg = bytes([0x10])  # Voice start message type
+        self.send(msg)
+        
+        self.debug("Voice session started")
+        self.state_changed.emit("VOICE_START", "", self.client_id)
+    
+    def end_voice_session(self):
+        """End a voice session"""
+        if not self.voice_active:
+            return
+            
+        self.voice_active = False
+        
+        # Send voice end message
+        msg = bytes([0x12])  # Voice end message type
+        self.send(msg)
+        
+        self.debug("Voice session ended")
+        self.state_changed.emit("VOICE_END", "", self.client_id)
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/UI/session.py b/protocol_prototype/DryBox/UI/session.py
index 5fc48c2..86a4a3f 100644
--- a/protocol_prototype/DryBox/UI/session.py
+++ b/protocol_prototype/DryBox/UI/session.py
@@ -10,8 +10,8 @@ from dissononce.dh.keypair import KeyPair
 from dissononce.dh.x25519.public import PublicKey
 from dissononce.hash.sha256 import SHA256Hash
 
-# Configure root logger for debug output
-logging.basicConfig(level=logging.DEBUG, format="%(message)s")
+# Configure logging - disabled by default to avoid noise
+# logging.basicConfig(level=logging.DEBUG, format="%(message)s")
 
 class NoiseXKSession:
     @staticmethod
@@ -46,7 +46,7 @@ class NoiseXKSession:
         so that each side reads or writes in the correct message order.
         On completion, self._send_cs and self._recv_cs hold the two CipherStates.
         """
-        logging.debug(f"[handshake] start (initiator={initiator})")
+        # logging.debug(f"[handshake] start (initiator={initiator})")
         # initialize with our KeyPair and their PublicKey
         if initiator:
             # initiator knows peer’s static out-of-band
@@ -58,7 +58,7 @@ class NoiseXKSession:
                 rs=self.peer_pubkey
             )
         else:
-            logging.debug("[handshake] responder initializing without rs")
+            # logging.debug("[handshake] responder initializing without rs")
             # responder must NOT supply rs here
             self._hs.initialize(
                 XKHandshakePattern(),
@@ -72,34 +72,34 @@ class NoiseXKSession:
             # 1) -> e
             buf1 = bytearray()
             cs_pair = self._hs.write_message(b'', buf1)
-            logging.debug(f"[-> e] {buf1.hex()}")
+            # logging.debug(f"[-> e] {buf1.hex()}")
             self._send_all(sock, buf1)
 
             # 2) <- e, es, s, ss
             msg2 = self._recv_all(sock)
-            logging.debug(f"[<- msg2] {msg2.hex()}")
+            # logging.debug(f"[<- msg2] {msg2.hex()}")
             self._hs.read_message(msg2, bytearray())
 
             # 3) -> se (final)
             buf3 = bytearray()
             cs_pair = self._hs.write_message(b'', buf3)
-            logging.debug(f"[-> se] {buf3.hex()}")
+            # logging.debug(f"[-> se] {buf3.hex()}")
             self._send_all(sock, buf3)
         else:
             # 1) <- e
             msg1 = self._recv_all(sock)
-            logging.debug(f"[<- e] {msg1.hex()}")
+            # logging.debug(f"[<- e] {msg1.hex()}")
             self._hs.read_message(msg1, bytearray())
 
             # 2) -> e, es, s, ss
             buf2 = bytearray()
             cs_pair = self._hs.write_message(b'', buf2)
-            logging.debug(f"[-> msg2] {buf2.hex()}")
+            # logging.debug(f"[-> msg2] {buf2.hex()}")
             self._send_all(sock, buf2)
 
             # 3) <- se (final)
             msg3 = self._recv_all(sock)
-            logging.debug(f"[<- se] {msg3.hex()}")
+            # logging.debug(f"[<- se] {msg3.hex()}")
             cs_pair = self._hs.read_message(msg3, bytearray())
 
         # on the final step, we must get exactly two CipherStates
@@ -168,9 +168,9 @@ class NoiseXKSession:
         # Read 2-byte length prefix, then the payload
         hdr = self._read_exact(sock, 2)
         length = int.from_bytes(hdr, 'big')
-        logging.debug(f"[RECV] length={length} ({hdr.hex()})")
+        # logging.debug(f"[RECV] length={length} ({hdr.hex()})")
         data = self._read_exact(sock, length)
-        logging.debug(f"[RECV] data={data.hex()}")
+        # logging.debug(f"[RECV] data={data.hex()}")
         return data
 
     @staticmethod
diff --git a/protocol_prototype/DryBox/UI/waveform_widget.py b/protocol_prototype/DryBox/UI/waveform_widget.py
index 9b26240..bb507a0 100644
--- a/protocol_prototype/DryBox/UI/waveform_widget.py
+++ b/protocol_prototype/DryBox/UI/waveform_widget.py
@@ -1,4 +1,5 @@
 import random
+import struct
 from PyQt5.QtWidgets import QWidget
 from PyQt5.QtCore import QTimer, QSize, QPointF
 from PyQt5.QtGui import QPainter, QColor, QPen, QLinearGradient, QBrush
@@ -7,8 +8,8 @@ class WaveformWidget(QWidget):
     def __init__(self, parent=None, dynamic=False):
         super().__init__(parent)
         self.dynamic = dynamic
-        self.setMinimumSize(200, 80)
-        self.setMaximumHeight(100)
+        self.setMinimumSize(200, 60)
+        self.setMaximumHeight(80)
         self.waveform_data = [random.randint(10, 90) for _ in range(50)]
         if self.dynamic:
             self.timer = QTimer(self)
@@ -20,8 +21,28 @@ class WaveformWidget(QWidget):
         self.update()
 
     def set_data(self, data):
-        amplitude = sum(byte for byte in data) % 90 + 10
-        self.waveform_data = self.waveform_data[1:] + [amplitude]
+        # Convert audio data to visual amplitude
+        if isinstance(data, bytes) and len(data) >= 2:
+            # Extract PCM samples (16-bit signed)
+            num_samples = min(len(data) // 2, 20)  # Take up to 20 samples
+            samples = []
+            for i in range(0, num_samples * 2, 2):
+                if i + 1 < len(data):
+                    sample = struct.unpack('h', data[i:i+2])[0]
+                    # Normalize to 0-100 range
+                    amplitude = abs(sample) / 327.68  # 32768/100
+                    samples.append(min(95, max(5, amplitude)))
+            
+            if samples:
+                # Add new samples and maintain fixed size
+                self.waveform_data.extend(samples)
+                # Keep last 50 samples
+                self.waveform_data = self.waveform_data[-50:]
+        else:
+            # Fallback for non-audio data
+            amplitude = sum(byte for byte in data[:20]) % 90 + 10
+            self.waveform_data = self.waveform_data[1:] + [amplitude]
+        
         self.update()
 
     def paintEvent(self, event):
diff --git a/protocol_prototype/DryBox/devnote.txt b/protocol_prototype/DryBox/devnote.txt
deleted file mode 100644
index ae75b1c..0000000
--- a/protocol_prototype/DryBox/devnote.txt
+++ /dev/null
@@ -1,13 +0,0 @@
-simulator/
-├── gsm_simulator.py            # gsm_simulator
-├── launch_gsm_simulator.sh     # use to start docker and simulator, run in terminal
-
-2 clients nect to gsm_simulator and simulate a call using noise protocol
-UI/
-├── main.py                     # UI setup and event handling
-├── phone_manager.py            # Phone state, client init, audio logic
-├── phone_client.py             # Socket communication and threading
-├── client_state.py             # Client state and command processing
-├── session.py                  # Noise XK crypto session
-├── waveform_widget.py          # Waveform UI component
-├── phone_state.py              # State constants
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/install_audio_deps.sh b/protocol_prototype/DryBox/install_audio_deps.sh
new file mode 100755
index 0000000..8d4d5eb
--- /dev/null
+++ b/protocol_prototype/DryBox/install_audio_deps.sh
@@ -0,0 +1,58 @@
+#!/bin/bash
+# Install audio dependencies for DryBox
+
+echo "Installing audio dependencies for DryBox..."
+echo
+
+# Detect OS
+if [ -f /etc/os-release ]; then
+    . /etc/os-release
+    OS=$ID
+    VER=$VERSION_ID
+else
+    echo "Cannot detect OS. Please install manually."
+    exit 1
+fi
+
+case $OS in
+    fedora)
+        echo "Detected Fedora $VER"
+        echo "Installing python3-devel and portaudio-devel..."
+        sudo dnf install -y python3-devel portaudio-devel
+        ;;
+    
+    ubuntu|debian)
+        echo "Detected $OS $VER"
+        echo "Installing python3-dev and portaudio19-dev..."
+        sudo apt-get update
+        sudo apt-get install -y python3-dev portaudio19-dev
+        ;;
+    
+    *)
+        echo "Unsupported OS: $OS"
+        echo "Please install manually:"
+        echo "  - Python development headers"
+        echo "  - PortAudio development libraries"
+        exit 1
+        ;;
+esac
+
+if [ $? -eq 0 ]; then
+    echo
+    echo "System dependencies installed successfully!"
+    echo "Now installing PyAudio..."
+    pip install pyaudio
+    
+    if [ $? -eq 0 ]; then
+        echo
+        echo "✅ Audio dependencies installed successfully!"
+        echo "You can now use real-time audio playback in DryBox."
+    else
+        echo
+        echo "❌ Failed to install PyAudio"
+        echo "Try: pip install --user pyaudio"
+    fi
+else
+    echo
+    echo "❌ Failed to install system dependencies"
+fi
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/requirements.txt b/protocol_prototype/DryBox/requirements.txt
new file mode 100644
index 0000000..af3c25b
--- /dev/null
+++ b/protocol_prototype/DryBox/requirements.txt
@@ -0,0 +1,22 @@
+# Core dependencies for DryBox integrated protocol
+
+# Noise Protocol Framework
+dissononce>=0.34.3
+
+# Cryptography
+cryptography>=41.0.0
+
+# Qt GUI
+PyQt5>=5.15.0
+
+# Numerical computing (for signal processing)
+numpy>=1.24.0
+
+# Audio processing (for real audio I/O)
+pyaudio>=0.2.11
+
+# Wave file handling (included in standard library)
+# wave
+
+# For future integration with real Codec2
+# pycodec2>=1.0.0
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/run_ui.sh b/protocol_prototype/DryBox/run_ui.sh
new file mode 100755
index 0000000..2b3beb1
--- /dev/null
+++ b/protocol_prototype/DryBox/run_ui.sh
@@ -0,0 +1,11 @@
+#!/bin/bash
+# Run DryBox UI with proper Wayland support on Fedora
+
+cd "$(dirname "$0")"
+
+# Use native Wayland if available
+export QT_QPA_PLATFORM=wayland
+
+# Run the UI
+cd UI
+python3 main.py
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/simulator/gsm_simulator.py b/protocol_prototype/DryBox/simulator/gsm_simulator.py
index e0cc4d3..03bf5cb 100644
--- a/protocol_prototype/DryBox/simulator/gsm_simulator.py
+++ b/protocol_prototype/DryBox/simulator/gsm_simulator.py
@@ -1,10 +1,11 @@
 import socket
 import threading
 import time
+import struct
 
 HOST = "0.0.0.0"
 PORT = 12345
-FRAME_SIZE = 1000
+FRAME_SIZE = 10000  # Increased to avoid fragmenting voice frames
 FRAME_DELAY = 0.02
 
 clients = []
@@ -12,25 +13,49 @@ clients_lock = threading.Lock()
 
 def handle_client(client_sock, client_id):
     print(f"Starting handle_client for Client {client_id}")
+    recv_buffer = bytearray()
+    
     try:
         while True:
             other_client = None
             with clients_lock:
                 if len(clients) == 2 and client_id < len(clients):
                     other_client = clients[1 - client_id]
-            print(f"Client {client_id} waiting for data, other_client exists: {other_client is not None}")
             
             try:
-                data = client_sock.recv(1024)
-                if not data:
+                chunk = client_sock.recv(4096)
+                if not chunk:
                     print(f"Client {client_id} disconnected or no data received")
                     break
-                if other_client:
-                    for i in range(0, len(data), FRAME_SIZE):
-                        frame = data[i:i + FRAME_SIZE]
-                        other_client.send(frame)
-                        time.sleep(FRAME_DELAY)
-                    print(f"Forwarded {len(data)} bytes from Client {client_id} to Client {1 - client_id}")
+                    
+                # Add to buffer
+                recv_buffer.extend(chunk)
+                
+                # Process complete messages
+                while len(recv_buffer) >= 4:
+                    # Read message length
+                    msg_len = struct.unpack('>I', recv_buffer[:4])[0]
+                    
+                    # Check if we have the complete message
+                    if len(recv_buffer) >= 4 + msg_len:
+                        # Extract complete message (including length prefix)
+                        complete_msg = bytes(recv_buffer[:4+msg_len])
+                        # Remove from buffer
+                        recv_buffer = recv_buffer[4+msg_len:]
+                        
+                        # Forward complete message to other client
+                        if other_client:
+                            try:
+                                other_client.send(complete_msg)
+                                print(f"Forwarded {len(complete_msg)} bytes from Client {client_id} to Client {1 - client_id}")
+                            except Exception as e:
+                                print(f"Error forwarding from Client {client_id}: {e}")
+                        else:
+                            print(f"No other client to forward to from Client {client_id}")
+                    else:
+                        # Wait for more data
+                        break
+                        
             except socket.error as e:
                 print(f"Socket error with Client {client_id}: {e}")
                 break
diff --git a/protocol_prototype/DryBox/unused/external_caller.py b/protocol_prototype/DryBox/unused/external_caller.py
deleted file mode 100644
index 9b06026..0000000
--- a/protocol_prototype/DryBox/unused/external_caller.py
+++ /dev/null
@@ -1,24 +0,0 @@
-#external_caller.py
-import socket
-import time
-
-
-def connect():
-    caller_socket = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
-    caller_socket.connect(('localhost', 12345))
-    caller_socket.send("CALLER".encode())
-    print("Connected to GSM simulator as CALLER")
-    time.sleep(2)  # Wait 2 seconds for receiver to connect
-
-    for i in range(5):
-        message = f"Audio packet {i + 1}"
-        caller_socket.send(message.encode())
-        print(f"Sent: {message}")
-        time.sleep(1)
-
-    caller_socket.send("CALL_END".encode())
-    print("Call ended.")
-    caller_socket.close()
-
-if __name__ == "__main__":
-    connect()
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/unused/external_receiver.py b/protocol_prototype/DryBox/unused/external_receiver.py
deleted file mode 100644
index 20c02de..0000000
--- a/protocol_prototype/DryBox/unused/external_receiver.py
+++ /dev/null
@@ -1,37 +0,0 @@
-#external_receiver.py
-import socket
-
-def connect():
-    receiver_socket = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
-    receiver_socket.settimeout(15)  # Increase timeout to 15 seconds
-    receiver_socket.connect(('localhost', 12345))
-    receiver_socket.send("RECEIVER".encode())
-    print("Connected to GSM simulator as RECEIVER")
-
-    while True:
-        try:
-            data = receiver_socket.recv(1024).decode().strip()
-            if not data:
-                print("No data received. Connection closed.")
-                break
-            if data == "RINGING":
-                print("Incoming call... ringing")
-            elif data == "CALL_END":
-                print("Call ended by caller.")
-                break
-            elif data == "CALL_DROPPED":
-                print("Call dropped by network.")
-                break
-            else:
-                print(f"Received: {data}")
-        except socket.timeout:
-            print("Timed out waiting for data.")
-            break
-        except Exception as e:
-            print(f"Receiver error: {e}")
-            break
-
-    receiver_socket.close()
-
-if __name__ == "__main__":
-    connect()
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/unused/protocol.py b/protocol_prototype/DryBox/unused/protocol.py
deleted file mode 100644
index 4c3cc79..0000000
--- a/protocol_prototype/DryBox/unused/protocol.py
+++ /dev/null
@@ -1,86 +0,0 @@
-import socket
-import os
-import time
-import subprocess
-
-# Configuration
-HOST = "localhost"
-PORT = 12345
-INPUT_FILE = "wav/input.wav"
-OUTPUT_FILE = "wav/received.wav"
-
-
-def encrypt_data(data):
-    return data  # Replace with your encryption protocol
-
-
-def decrypt_data(data):
-    return data  # Replace with your decryption protocol
-
-
-def run_protocol(send_mode=True):
-    """Connect to the simulator and send/receive data."""
-    sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
-    sock.connect((HOST, PORT))
-    print(f"Connected to simulator at {HOST}:{PORT}")
-
-    if send_mode:
-        # Sender mode: Encode audio with toast
-        os.system(f"toast -p -l {INPUT_FILE}")  # Creates input.wav.gsm
-        input_gsm_file = f"{INPUT_FILE}.gsm"
-        if not os.path.exists(input_gsm_file):
-            print(f"Error: {input_gsm_file} not created")
-            sock.close()
-            return
-        with open(input_gsm_file, "rb") as f:
-            voice_data = f.read()
-
-        encrypted_data = encrypt_data(voice_data)
-        sock.send(encrypted_data)
-        print(f"Sent {len(encrypted_data)} bytes")
-        os.remove(input_gsm_file)  # Clean up
-    else:
-        # Receiver mode: Wait for and receive data
-        print("Waiting for data from sender...")
-        received_data = b""
-        sock.settimeout(5.0)
-        try:
-            while True:
-                print("Calling recv()...")
-                data = sock.recv(1024)
-                print(f"Received {len(data)} bytes")
-                if not data:
-                    print("Connection closed by sender or simulator")
-                    break
-                received_data += data
-        except socket.timeout:
-            print("Timed out waiting for data")
-
-        if received_data:
-            with open("received.gsm", "wb") as f:
-                f.write(decrypt_data(received_data))
-            print(f"Wrote {len(received_data)} bytes to received.gsm")
-            # Decode with untoast, then convert to WAV with sox
-            result = subprocess.run(["untoast", "received.gsm"], capture_output=True, text=True)
-            print(f"untoast return code: {result.returncode}")
-            print(f"untoast stderr: {result.stderr}")
-            if result.returncode == 0:
-                if os.path.exists("received"):
-                    # Convert raw PCM to WAV (8 kHz, mono, 16-bit)
-                    subprocess.run(["sox", "-t", "raw", "-r", "8000", "-e", "signed", "-b", "16", "-c", "1", "received",
-                                    OUTPUT_FILE])
-                    os.remove("received")
-                    print(f"Received and saved {len(received_data)} bytes to {OUTPUT_FILE}")
-                else:
-                    print("Error: 'received' file not created by untoast")
-            else:
-                print(f"untoast failed: {result.stderr}")
-        else:
-            print("No data received from simulator")
-
-    sock.close()
-
-
-if __name__ == "__main__":
-    mode = input("Enter 'send' to send data or 'receive' to receive data: ").strip().lower()
-    run_protocol(send_mode=(mode == "send"))
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/voice_codec.py b/protocol_prototype/DryBox/voice_codec.py
new file mode 100644
index 0000000..dcf8609
--- /dev/null
+++ b/protocol_prototype/DryBox/voice_codec.py
@@ -0,0 +1,714 @@
+"""
+Voice codec integration for encrypted voice over GSM.
+Implements Codec2 compression with FSK modulation for transmitting
+encrypted voice data over standard GSM voice channels.
+"""
+
+import array
+import math
+import struct
+from typing import Optional, Tuple, List
+from dataclasses import dataclass
+from enum import IntEnum
+
+try:
+    import numpy as np
+    HAS_NUMPY = True
+except ImportError:
+    HAS_NUMPY = False
+
+# ANSI colors
+RED = "\033[91m"
+GREEN = "\033[92m"
+YELLOW = "\033[93m"
+BLUE = "\033[94m"
+RESET = "\033[0m"
+
+
+class Codec2Mode(IntEnum):
+    """Codec2 bitrate modes."""
+    MODE_3200 = 0  # 3200 bps
+    MODE_2400 = 1  # 2400 bps
+    MODE_1600 = 2  # 1600 bps
+    MODE_1400 = 3  # 1400 bps
+    MODE_1300 = 4  # 1300 bps
+    MODE_1200 = 5  # 1200 bps (recommended for robustness)
+    MODE_700C = 6  # 700 bps
+
+
+@dataclass
+class Codec2Frame:
+    """Represents a single Codec2 compressed voice frame."""
+    mode: Codec2Mode
+    bits: bytes
+    timestamp: float
+    frame_number: int
+
+
+class Codec2Wrapper:
+    """
+    Wrapper for Codec2 voice codec.
+    In production, this would use py_codec2 or ctypes bindings to libcodec2.
+    This is a simulation interface for protocol development.
+    """
+
+    # Frame sizes in bits for each mode
+    FRAME_BITS = {
+        Codec2Mode.MODE_3200: 64,
+        Codec2Mode.MODE_2400: 48,
+        Codec2Mode.MODE_1600: 64,
+        Codec2Mode.MODE_1400: 56,
+        Codec2Mode.MODE_1300: 52,
+        Codec2Mode.MODE_1200: 48,
+        Codec2Mode.MODE_700C: 28
+    }
+
+    # Frame duration in ms
+    FRAME_MS = {
+        Codec2Mode.MODE_3200: 20,
+        Codec2Mode.MODE_2400: 20,
+        Codec2Mode.MODE_1600: 40,
+        Codec2Mode.MODE_1400: 40,
+        Codec2Mode.MODE_1300: 40,
+        Codec2Mode.MODE_1200: 40,
+        Codec2Mode.MODE_700C: 40
+    }
+
+    def __init__(self, mode: Codec2Mode = Codec2Mode.MODE_1200):
+        """
+        Initialize Codec2 wrapper.
+
+        Args:
+            mode: Codec2 bitrate mode (default 1200 bps for robustness)
+        """
+        self.mode = mode
+        self.frame_bits = self.FRAME_BITS[mode]
+        self.frame_bytes = (self.frame_bits + 7) // 8
+        self.frame_ms = self.FRAME_MS[mode]
+        self.frame_samples = int(8000 * self.frame_ms / 1000)  # 8kHz sampling
+        self.frame_counter = 0
+
+        # Quiet initialization - no print
+
+    def encode(self, audio_samples) -> Optional[Codec2Frame]:
+        """
+        Encode PCM audio samples to Codec2 frame.
+
+        Args:
+            audio_samples: PCM samples (8kHz, 16-bit signed)
+
+        Returns:
+            Codec2Frame or None if insufficient samples
+        """
+        if len(audio_samples) < self.frame_samples:
+            return None
+
+        # In production: call codec2_encode(state, bits, samples)
+        # Simulation: create pseudo-compressed data
+        compressed = self._simulate_compression(audio_samples[:self.frame_samples])
+
+        frame = Codec2Frame(
+            mode=self.mode,
+            bits=compressed,
+            timestamp=self.frame_counter * self.frame_ms / 1000.0,
+            frame_number=self.frame_counter
+        )
+
+        self.frame_counter += 1
+        return frame
+
+    def decode(self, frame: Codec2Frame):
+        """
+        Decode Codec2 frame to PCM audio samples.
+
+        Args:
+            frame: Codec2 compressed frame
+
+        Returns:
+            PCM samples (8kHz, 16-bit signed)
+        """
+        if frame.mode != self.mode:
+            raise ValueError(f"Frame mode {frame.mode} doesn't match decoder mode {self.mode}")
+
+        # In production: call codec2_decode(state, samples, bits)
+        # Simulation: decompress to audio
+        return self._simulate_decompression(frame.bits)
+
+    def _simulate_compression(self, samples) -> bytes:
+        """Simulate Codec2 compression (for testing)."""
+        # Convert to list if needed
+        if hasattr(samples, 'tolist'):
+            sample_list = samples.tolist()
+        elif hasattr(samples, '__iter__'):
+            sample_list = list(samples)
+        else:
+            sample_list = samples
+
+        # Extract basic features for simulation
+        if HAS_NUMPY and hasattr(samples, '__array__'):
+            # Convert to numpy array if needed
+            np_samples = np.asarray(samples, dtype=np.float32)
+            if len(np_samples) > 0:
+                mean_square = np.mean(np_samples ** 2)
+                energy = np.sqrt(mean_square) if not np.isnan(mean_square) else 0.0
+                zero_crossings = np.sum(np.diff(np.sign(np_samples)) != 0)
+            else:
+                energy = 0.0
+                zero_crossings = 0
+        else:
+            # Manual calculation without numpy
+            if sample_list and len(sample_list) > 0:
+                energy = math.sqrt(sum(s**2 for s in sample_list) / len(sample_list))
+                zero_crossings = sum(1 for i in range(1, len(sample_list))
+                                   if (sample_list[i-1] >= 0) != (sample_list[i] >= 0))
+            else:
+                energy = 0.0
+                zero_crossings = 0
+
+        # Pack into bytes (simplified)
+        # Ensure values are valid
+        energy_int = max(0, min(65535, int(energy)))
+        zc_int = max(0, min(65535, int(zero_crossings)))
+        data = struct.pack('<HH', energy_int, zc_int)
+
+        # Pad to expected frame size
+        data += b'\x00' * (self.frame_bytes - len(data))
+
+        return data[:self.frame_bytes]
+
+    def _simulate_decompression(self, compressed: bytes):
+        """Simulate Codec2 decompression (for testing)."""
+        # Unpack features
+        if len(compressed) >= 4:
+            energy, zero_crossings = struct.unpack('<HH', compressed[:4])
+        else:
+            energy, zero_crossings = 1000, 100
+
+        # Generate synthetic speech-like signal
+        if HAS_NUMPY:
+            t = np.linspace(0, self.frame_ms/1000, self.frame_samples)
+
+            # Base frequency from zero crossings
+            freq = zero_crossings * 10  # Simplified mapping
+
+            # Generate harmonics
+            signal = np.zeros(self.frame_samples)
+            for harmonic in range(1, 4):
+                signal += np.sin(2 * np.pi * freq * harmonic * t) / harmonic
+
+            # Apply energy envelope
+            signal *= energy / 10000.0
+
+            # Convert to 16-bit PCM
+            return (signal * 32767).astype(np.int16)
+        else:
+            # Manual generation without numpy
+            samples = []
+            freq = zero_crossings * 10
+
+            for i in range(self.frame_samples):
+                t = i / 8000.0  # 8kHz sample rate
+                value = 0
+                for harmonic in range(1, 4):
+                    value += math.sin(2 * math.pi * freq * harmonic * t) / harmonic
+
+                value *= energy / 10000.0
+                # Clamp to 16-bit range
+                sample = int(value * 32767)
+                sample = max(-32768, min(32767, sample))
+                samples.append(sample)
+
+            return array.array('h', samples)
+
+
+class FSKModem:
+    """
+    4-FSK modem for transmitting digital data over voice channels.
+    Designed to survive GSM/AMR/EVS vocoders.
+    """
+
+    def __init__(self, sample_rate: int = 8000, baud_rate: int = 600):
+        """
+        Initialize FSK modem.
+
+        Args:
+            sample_rate: Audio sample rate (Hz)
+            baud_rate: Symbol rate (baud)
+        """
+        self.sample_rate = sample_rate
+        self.baud_rate = baud_rate
+        self.samples_per_symbol = int(sample_rate / baud_rate)
+
+        # 4-FSK frequencies (300-3400 Hz band)
+        self.frequencies = [
+            600,   # 00
+            1200,  # 01
+            1800,  # 10
+            2400   # 11
+        ]
+
+        # Preamble for synchronization (800 Hz, 100ms)
+        self.preamble_freq = 800
+        self.preamble_duration = 0.1  # seconds
+
+        # Quiet initialization - no print
+
+    def modulate(self, data: bytes, add_preamble: bool = True):
+        """
+        Modulate binary data to FSK audio signal.
+
+        Args:
+            data: Binary data to modulate
+            add_preamble: Whether to add synchronization preamble
+
+        Returns:
+            Audio signal (normalized float32 array or list)
+        """
+        # Convert bytes to dibits (2-bit symbols)
+        symbols = []
+        for byte in data:
+            symbols.extend([
+                (byte >> 6) & 0x03,
+                (byte >> 4) & 0x03,
+                (byte >> 2) & 0x03,
+                byte & 0x03
+            ])
+
+        # Generate audio signal
+        signal = []
+
+        # Add preamble
+        if add_preamble:
+            preamble_samples = int(self.preamble_duration * self.sample_rate)
+            if HAS_NUMPY:
+                t = np.arange(preamble_samples) / self.sample_rate
+                preamble = np.sin(2 * np.pi * self.preamble_freq * t)
+                signal.extend(preamble)
+            else:
+                for i in range(preamble_samples):
+                    t = i / self.sample_rate
+                    value = math.sin(2 * math.pi * self.preamble_freq * t)
+                    signal.append(value)
+
+        # Modulate symbols
+        for symbol in symbols:
+            freq = self.frequencies[symbol]
+            if HAS_NUMPY:
+                t = np.arange(self.samples_per_symbol) / self.sample_rate
+                tone = np.sin(2 * np.pi * freq * t)
+                signal.extend(tone)
+            else:
+                for i in range(self.samples_per_symbol):
+                    t = i / self.sample_rate
+                    value = math.sin(2 * math.pi * freq * t)
+                    signal.append(value)
+
+        # Apply smoothing to reduce clicks
+        if HAS_NUMPY:
+            audio = np.array(signal, dtype=np.float32)
+        else:
+            audio = array.array('f', signal)
+        audio = self._apply_envelope(audio)
+
+        return audio
+
+    def demodulate(self, audio) -> Tuple[bytes, float]:
+        """
+        Demodulate FSK audio signal to binary data.
+
+        Args:
+            audio: Audio signal
+
+        Returns:
+            Tuple of (demodulated data, confidence score)
+        """
+        # Find preamble
+        preamble_start = self._find_preamble(audio)
+        if preamble_start < 0:
+            return b'', 0.0
+
+        # Skip preamble
+        data_start = preamble_start + int(self.preamble_duration * self.sample_rate)
+
+        # Demodulate symbols
+        symbols = []
+        confidence_scores = []
+
+        pos = data_start
+        while pos + self.samples_per_symbol <= len(audio):
+            symbol_audio = audio[pos:pos + self.samples_per_symbol]
+            symbol, confidence = self._demodulate_symbol(symbol_audio)
+            symbols.append(symbol)
+            confidence_scores.append(confidence)
+            pos += self.samples_per_symbol
+
+        # Convert symbols to bytes
+        data = bytearray()
+        for i in range(0, len(symbols), 4):
+            if i + 3 < len(symbols):
+                byte = (symbols[i] << 6) | (symbols[i+1] << 4) | (symbols[i+2] << 2) | symbols[i+3]
+                data.append(byte)
+
+        if HAS_NUMPY and confidence_scores:
+            avg_confidence = np.mean(confidence_scores)
+        else:
+            avg_confidence = sum(confidence_scores) / len(confidence_scores) if confidence_scores else 0.0
+        return bytes(data), avg_confidence
+
+    def _find_preamble(self, audio) -> int:
+        """Find preamble in audio signal."""
+        # Simple energy-based detection
+        window_size = int(0.01 * self.sample_rate)  # 10ms window
+
+        if HAS_NUMPY:
+            for i in range(0, len(audio) - window_size, window_size // 2):
+                window = audio[i:i + window_size]
+
+                # Check for preamble frequency
+                fft = np.fft.fft(window)
+                freqs = np.fft.fftfreq(len(window), 1/self.sample_rate)
+
+                # Find peak near preamble frequency
+                idx = np.argmax(np.abs(fft[:len(fft)//2]))
+                peak_freq = abs(freqs[idx])
+
+                if abs(peak_freq - self.preamble_freq) < 50:  # 50 Hz tolerance
+                    return i
+        else:
+            # Simple zero-crossing based detection without FFT
+            for i in range(0, len(audio) - window_size, window_size // 2):
+                window = list(audio[i:i + window_size])
+
+                # Count zero crossings
+                zero_crossings = 0
+                for j in range(1, len(window)):
+                    if (window[j-1] >= 0) != (window[j] >= 0):
+                        zero_crossings += 1
+
+                # Estimate frequency from zero crossings
+                estimated_freq = (zero_crossings * self.sample_rate) / (2 * len(window))
+
+                if abs(estimated_freq - self.preamble_freq) < 100:  # 100 Hz tolerance
+                    return i
+
+        return -1
+
+    def _demodulate_symbol(self, audio) -> Tuple[int, float]:
+        """Demodulate a single FSK symbol."""
+        if HAS_NUMPY:
+            # FFT-based demodulation
+            fft = np.fft.fft(audio)
+            freqs = np.fft.fftfreq(len(audio), 1/self.sample_rate)
+            magnitude = np.abs(fft[:len(fft)//2])
+
+            # Find energy at each FSK frequency
+            energies = []
+            for freq in self.frequencies:
+                idx = np.argmin(np.abs(freqs[:len(freqs)//2] - freq))
+                energy = magnitude[idx]
+                energies.append(energy)
+
+            # Select symbol with highest energy
+            symbol = np.argmax(energies)
+        else:
+            # Goertzel algorithm for specific frequency detection
+            audio_list = list(audio) if hasattr(audio, '__iter__') else audio
+            energies = []
+
+            for freq in self.frequencies:
+                # Goertzel algorithm
+                omega = 2 * math.pi * freq / self.sample_rate
+                coeff = 2 * math.cos(omega)
+
+                s_prev = 0
+                s_prev2 = 0
+
+                for sample in audio_list:
+                    s = sample + coeff * s_prev - s_prev2
+                    s_prev2 = s_prev
+                    s_prev = s
+
+                # Calculate magnitude
+                power = s_prev2 * s_prev2 + s_prev * s_prev - coeff * s_prev * s_prev2
+                energies.append(math.sqrt(abs(power)))
+
+            # Select symbol with highest energy
+            symbol = energies.index(max(energies))
+
+        # Confidence is ratio of strongest to second strongest
+        sorted_energies = sorted(energies, reverse=True)
+        confidence = sorted_energies[0] / (sorted_energies[1] + 1e-6)
+
+        return symbol, min(confidence, 10.0) / 10.0
+
+    def _apply_envelope(self, audio):
+        """Apply smoothing envelope to reduce clicks."""
+        # Simple raised cosine envelope
+        ramp_samples = int(0.002 * self.sample_rate)  # 2ms ramps
+
+        if len(audio) > 2 * ramp_samples:
+            if HAS_NUMPY:
+                # Fade in
+                t = np.linspace(0, np.pi/2, ramp_samples)
+                audio[:ramp_samples] *= np.sin(t) ** 2
+
+                # Fade out
+                audio[-ramp_samples:] *= np.sin(t[::-1]) ** 2
+            else:
+                # Manual fade in
+                for i in range(ramp_samples):
+                    t = (i / ramp_samples) * (math.pi / 2)
+                    factor = math.sin(t) ** 2
+                    audio[i] *= factor
+
+                # Manual fade out
+                for i in range(ramp_samples):
+                    t = ((ramp_samples - 1 - i) / ramp_samples) * (math.pi / 2)
+                    factor = math.sin(t) ** 2
+                    audio[-(i+1)] *= factor
+
+        return audio
+
+
+class VoiceProtocol:
+    """
+    Integrates voice codec and modem with the Icing protocol
+    for encrypted voice transmission over GSM.
+    """
+
+    def __init__(self, protocol_instance):
+        """
+        Initialize voice protocol handler.
+
+        Args:
+            protocol_instance: IcingProtocol instance
+        """
+        self.protocol = protocol_instance
+        self.codec = Codec2Wrapper(Codec2Mode.MODE_1200)
+        self.modem = FSKModem(sample_rate=8000, baud_rate=600)
+
+        # Voice crypto state
+        self.voice_iv_counter = 0
+        self.voice_sequence = 0
+
+        # Buffers
+        if HAS_NUMPY:
+            self.audio_buffer = np.array([], dtype=np.int16)
+        else:
+            self.audio_buffer = array.array('h')  # 16-bit signed integers
+        self.frame_buffer = []
+
+        print(f"{GREEN}[VOICE]{RESET} Voice protocol initialized")
+
+    def process_voice_input(self, audio_samples):
+        """
+        Process voice input: compress, encrypt, and modulate.
+
+        Args:
+            audio_samples: PCM audio samples (8kHz, 16-bit)
+
+        Returns:
+            Modulated audio signal ready for transmission (numpy array or array.array)
+        """
+        # Add to buffer
+        if HAS_NUMPY:
+            self.audio_buffer = np.concatenate([self.audio_buffer, audio_samples])
+        else:
+            self.audio_buffer.extend(audio_samples)
+
+        # Process complete frames
+        modulated_audio = []
+
+        while len(self.audio_buffer) >= self.codec.frame_samples:
+            # Extract frame
+            if HAS_NUMPY:
+                frame_audio = self.audio_buffer[:self.codec.frame_samples]
+                self.audio_buffer = self.audio_buffer[self.codec.frame_samples:]
+            else:
+                frame_audio = array.array('h', self.audio_buffer[:self.codec.frame_samples])
+                del self.audio_buffer[:self.codec.frame_samples]
+
+            # Compress with Codec2
+            compressed_frame = self.codec.encode(frame_audio)
+            if not compressed_frame:
+                continue
+
+            # Encrypt frame
+            encrypted = self._encrypt_voice_frame(compressed_frame)
+
+            # Add FEC
+            protected = self._add_fec(encrypted)
+
+            # Modulate to audio
+            audio_signal = self.modem.modulate(protected, add_preamble=True)
+            modulated_audio.append(audio_signal)
+
+        if modulated_audio:
+            if HAS_NUMPY:
+                return np.concatenate(modulated_audio)
+            else:
+                # Concatenate array.array objects
+                result = array.array('f')
+                for audio in modulated_audio:
+                    result.extend(audio)
+                return result
+        return None
+
+    def process_voice_output(self, modulated_audio):
+        """
+        Process received audio: demodulate, decrypt, and decompress.
+
+        Args:
+            modulated_audio: Received FSK-modulated audio
+
+        Returns:
+            Decoded PCM audio samples (numpy array or array.array)
+        """
+        # Demodulate
+        data, confidence = self.modem.demodulate(modulated_audio)
+
+        if confidence < 0.5:
+            print(f"{YELLOW}[VOICE]{RESET} Low demodulation confidence: {confidence:.2f}")
+            return None
+
+        # Remove FEC
+        frame_data = self._remove_fec(data)
+        if not frame_data:
+            return None
+
+        # Decrypt
+        compressed_frame = self._decrypt_voice_frame(frame_data)
+        if not compressed_frame:
+            return None
+
+        # Decompress
+        audio_samples = self.codec.decode(compressed_frame)
+
+        return audio_samples
+
+    def _encrypt_voice_frame(self, frame: Codec2Frame) -> bytes:
+        """Encrypt a voice frame using ChaCha20-CTR."""
+        if not self.protocol.hkdf_key:
+            raise ValueError("No encryption key available")
+
+        # Prepare frame data
+        frame_data = struct.pack('<BIH',
+            frame.mode,
+            frame.frame_number,
+            len(frame.bits)
+        ) + frame.bits
+
+        # Generate IV for this frame (ChaCha20 needs 16 bytes)
+        iv = struct.pack('<Q', self.voice_iv_counter) + b'\x00' * 8  # 8 + 8 = 16 bytes
+        self.voice_iv_counter += 1
+
+        # Encrypt using ChaCha20
+        from encryption import chacha20_encrypt
+        key = bytes.fromhex(self.protocol.hkdf_key)
+        encrypted = chacha20_encrypt(frame_data, key, iv)
+
+        # Add sequence number and IV hint
+        return struct.pack('<HQ', self.voice_sequence, self.voice_iv_counter) + encrypted
+
+    def _decrypt_voice_frame(self, data: bytes) -> Optional[Codec2Frame]:
+        """Decrypt a voice frame."""
+        if len(data) < 10:
+            return None
+
+        # Extract sequence and IV hint
+        sequence, iv_hint = struct.unpack('<HQ', data[:10])
+        encrypted = data[10:]
+
+        # Generate IV (16 bytes for ChaCha20)
+        iv = struct.pack('<Q', iv_hint) + b'\x00' * 8
+
+        # Decrypt
+        from encryption import chacha20_decrypt
+        key = bytes.fromhex(self.protocol.hkdf_key)
+
+        try:
+            decrypted = chacha20_decrypt(encrypted, key, iv)
+
+            # Parse frame
+            mode, frame_num, bits_len = struct.unpack('<BIH', decrypted[:7])
+            bits = decrypted[7:7+bits_len]
+
+            return Codec2Frame(
+                mode=Codec2Mode(mode),
+                bits=bits,
+                timestamp=0,  # Will be set by caller
+                frame_number=frame_num
+            )
+        except Exception as e:
+            print(f"{RED}[VOICE]{RESET} Decryption failed: {e}")
+            return None
+
+    def _add_fec(self, data: bytes) -> bytes:
+        """Add forward error correction."""
+        # Simple repetition code (3x) for testing
+        # In production: use convolutional code or LDPC
+        fec_data = bytearray()
+
+        for byte in data:
+            # Repeat each byte 3 times
+            fec_data.extend([byte, byte, byte])
+
+        return bytes(fec_data)
+
+    def _remove_fec(self, data: bytes) -> Optional[bytes]:
+        """Remove FEC and correct errors."""
+        if len(data) % 3 != 0:
+            return None
+
+        corrected = bytearray()
+
+        for i in range(0, len(data), 3):
+            # Majority voting
+            votes = [data[i], data[i+1], data[i+2]]
+            byte_value = max(set(votes), key=votes.count)
+            corrected.append(byte_value)
+
+        return bytes(corrected)
+
+
+# Example usage
+if __name__ == "__main__":
+    # Test Codec2 wrapper
+    print(f"\n{BLUE}=== Testing Codec2 Wrapper ==={RESET}")
+    codec = Codec2Wrapper(Codec2Mode.MODE_1200)
+
+    # Generate test audio
+    if HAS_NUMPY:
+        t = np.linspace(0, 0.04, 320)  # 40ms at 8kHz
+        test_audio = (np.sin(2 * np.pi * 440 * t) * 16384).astype(np.int16)
+    else:
+        test_audio = array.array('h')
+        for i in range(320):
+            t = i * 0.04 / 320
+            value = int(math.sin(2 * math.pi * 440 * t) * 16384)
+            test_audio.append(value)
+
+    # Encode
+    frame = codec.encode(test_audio)
+    print(f"Encoded frame: {len(frame.bits)} bytes")
+
+    # Decode
+    decoded = codec.decode(frame)
+    print(f"Decoded audio: {len(decoded)} samples")
+
+    # Test FSK modem
+    print(f"\n{BLUE}=== Testing FSK Modem ==={RESET}")
+    modem = FSKModem()
+
+    # Test data
+    test_data = b"Hello, secure voice!"
+
+    # Modulate
+    modulated = modem.modulate(test_data)
+    print(f"Modulated: {len(modulated)} samples ({len(modulated)/8000:.2f}s)")
+
+    # Demodulate
+    demodulated, confidence = modem.demodulate(modulated)
+    print(f"Demodulated: {demodulated}")
+    print(f"Confidence: {confidence:.2%}")
+    print(f"Match: {demodulated == test_data}")
\ No newline at end of file
diff --git a/protocol_prototype/DryBox/wav/input.wav b/protocol_prototype/DryBox/wav/input.wav
index 576bfd8..da3e917 100644
Binary files a/protocol_prototype/DryBox/wav/input.wav and b/protocol_prototype/DryBox/wav/input.wav differ
diff --git a/protocol_prototype/DryBox/wav/input_original.wav b/protocol_prototype/DryBox/wav/input_original.wav
new file mode 100644
index 0000000..576bfd8
Binary files /dev/null and b/protocol_prototype/DryBox/wav/input_original.wav differ
diff --git a/protocol_prototype/DryBox/wav/test_codec_only.wav b/protocol_prototype/DryBox/wav/test_codec_only.wav
new file mode 100644
index 0000000..b5f4502
Binary files /dev/null and b/protocol_prototype/DryBox/wav/test_codec_only.wav differ
diff --git a/protocol_prototype/DryBox/wav/test_full_pipeline.wav b/protocol_prototype/DryBox/wav/test_full_pipeline.wav
new file mode 100644
index 0000000..b5f4502
Binary files /dev/null and b/protocol_prototype/DryBox/wav/test_full_pipeline.wav differ