minor improvements

pkg/wyze/client.go

@@ -211,7 +211,7 @@ func (c *Client) StartAudio() error {
 }
 
 func (c *Client) StartIntercom() error {
-	if c.conn.IsBackchannelReady() {
+	if c.conn == nil || !c.conn.IsBackchannelReady() {
 		return nil
 	}
 
@@ -223,6 +223,17 @@ func (c *Client) StartIntercom() error {
 	return c.conn.AVServStart()
 }
 
+func (c *Client) StopIntercom() error {
+	if c.conn == nil || !c.conn.IsBackchannelReady() {
+		return nil
+	}
+
+	k10010 := c.buildK10010(MediaTypeReturnAudio, false)
+	c.conn.WriteAndWaitIOCtrl(KCmdControlChannel, k10010, KCmdControlChannelResp, 5*time.Second)
+
+	return c.conn.AVServStop()
+}
+
 func (c *Client) ReadPacket() (*tutk.Packet, error) {
 	return c.conn.AVRecvFrameData()
 }
@@ -270,10 +281,16 @@ func (c *Client) Close() error {
 		fmt.Printf("[Wyze] Closing connection\n")
 	}
 
+	c.StopIntercom()
+
 	if c.conn != nil {
 		c.conn.Close()
 	}
 
+	if c.verbose {
+		fmt.Printf("[Wyze] Connection closed\n")
+	}
+
 	return nil
 }
 

pkg/wyze/tutk/README.md

@@ -443,38 +443,10 @@ Offset  Size  Field            Description
 [31]    1     TwoWayAudio      0x01 if intercom supported
 [32-35] 4     Reserved
 [36-39] 4     BufferConfig     0x00000004
-[40-43] 4     Capabilities     0x001F07FB (see below)
+[40-43] 4     Capabilities     0x001F07FB
 [44-57] 14    Reserved
 ```
 
-### Capabilities Bitmask (0x001F07FB)
-
-```
-Bit   Hex        Name                    Description
-──────────────────────────────────────────────────────────────
-0     0x00000001 CYCLIC_FRAME_NUMBERING  Frame numbers wrap around
-1     0x00000002 CLEAN_BUF_ON_RESET      Clear buffer on stream reset
-3     0x00000008 TIMESTAMP_IN_FRAMEINFO  Timestamps in FRAMEINFO struct
-4     0x00000010 MULTI_CHANNEL           Multiple AV channels supported
-5     0x00000020 EXTENDED_FRAMEINFO      40-byte FRAMEINFO (vs 16-byte SDK)
-6     0x00000040 RESEND_TIMEOUT          Packet resend with timeout
-7     0x00000080 DTLS_SUPPORT            DTLS encryption supported
-8     0x00000100 SPEAKER_CHANNEL         Two-way audio / intercom
-9     0x00000200 PTZ_CHANNEL             PTZ control channel
-10    0x00000400 PLAYBACK_CHANNEL        SD card playback channel
-16    0x00010000 AV_SECURITY_ENABLED     Encrypted AV stream
-17    0x00020000 RESEND_ENABLED          Packet resend mechanism
-18    0x00040000 DTLS_PSK                DTLS with Pre-Shared Key
-19    0x00080000 DTLS_ECDHE              DTLS with ECDHE key exchange
-20    0x00100000 CHACHA20_POLY1305       ChaCha20-Poly1305 cipher support
-```
-
-**0x001F07FB breakdown:**
-```
-0x001F07FB = 0b0000_0000_0001_1111_0000_0111_1111_1011
-           = Bits: 0,1,3,4,5,6,7,8,9,10,16,17,18,19,20
-```
-
 ---
 
 ## 7. K-Command Authentication
@@ -515,9 +487,21 @@ Offset  Size  Field          Description
 [16+]   var   Payload        Command-specific data
 ```
 
-### K10000 - Auth Request (16 bytes)
+### K10000 - Auth Request (16 + JSON bytes)
 
-Header only, no payload. Initiates authentication.
+```
+Offset  Size  Field          Description
+──────────────────────────────────────────────────────────────
+[0-15]  16    HLHeader       CommandID = 10000, PayloadLen = len(JSON)
+[16+]   var   JSONPayload    Audio codec preferences
+```
+
+**JSON Payload:**
+```json
+{"cameraInfo":{"audioEncoderList":[137,138,140]}}
+```
+
+Where audioEncoderList contains supported codec IDs: 137=PCMU, 138=PCMA, 140=PCM.
 
 ### K10001 - Challenge (33+ bytes)
 
@@ -543,7 +527,7 @@ Offset  Size  Field          Description
 ──────────────────────────────────────────────────────────────
 [0-15]  16    HLHeader       CommandID = 10002, PayloadLen = 22
 [16-31] 16    Response       XXTEA-decrypted challenge
-[32-35] 4     UIDPrefix      First 4 bytes of UID
+[32-35] 4     SessionID      Random 4-byte session identifier
 [36]    1     VideoFlag      1 = enable video stream
 [37]    1     AudioFlag      1 = enable audio stream
 ```
@@ -620,6 +604,22 @@ Offset  Size  Field          Description
 | 0xF0 (240) | Maximum |
 | 0x3C (60) | SD quality |
 
+### K10052 - Set Resolution Doorbell (22 bytes)
+
+Used by doorbell models (WYZEDB3, WVOD1, HL_WCO2, WYZEC1) instead of K10056:
+
+```
+Offset  Size  Field          Description
+──────────────────────────────────────────────────────────────
+[0-15]  16    HLHeader       CommandID = 10052, PayloadLen = 6
+[16-17] 2     Bitrate        KB/s value (LE)
+[18]    1     FrameSize      Resolution + 1 (see table above)
+[19]    1     FPS            Frames per second, 0 = auto
+[20-21] 2     Reserved       Zero-filled
+```
+
+**Note:** K10052 has a different field order than K10056 (bitrate before frameSize).
+
 ---
 
 ## 9. AV Frame Structure
@@ -1155,12 +1155,14 @@ authkey = b64.replace('+', 'Z').replace('/', '9').replace('=', 'A')
 
 | Command | ID | Description |
 |---------|-----|-------------|
-| KCmdAuth | 10000 | Auth request |
+| KCmdAuth | 10000 | Auth request (with JSON) |
 | KCmdChallenge | 10001 | Challenge from camera |
 | KCmdChallengeResp | 10002 | Challenge response |
 | KCmdAuthResult | 10003 | Auth result (JSON) |
 | KCmdControlChannel | 10010 | Start/stop media |
 | KCmdControlChannelResp | 10011 | Control response |
+| KCmdSetResolutionDB | 10052 | Set resolution (doorbell) |
+| KCmdSetResolutionDBResp | 10053 | Resolution response (doorbell) |
 | KCmdSetResolution | 10056 | Set resolution/bitrate |
 | KCmdSetResolutionResp | 10057 | Resolution response |
 

pkg/wyze/tutk/conn.go

@@ -222,17 +222,19 @@ func (c *Conn) AVServStart() error {
 
 func (c *Conn) AVServStop() error {
 	c.mu.Lock()
-	defer c.mu.Unlock()
+	serverConn := c.serverConn
-
+	c.serverConn = nil
 	// Reset audio TX state
 	c.audioSeq = 0
 	c.audioFrameNo = 0
+	c.mu.Unlock()
 
-	if c.serverConn != nil {
+	if serverConn == nil {
-		err := c.serverConn.Close()
+		return nil
-		c.serverConn = nil
-		return err
 	}
+
+	go serverConn.Close()
+
 	return nil
 }
 
@@ -339,8 +341,13 @@ func (c *Conn) WriteAndWaitIOCtrl(cmd uint16, payload []byte, expectCmd uint16,
 	frame := c.buildIOCtrlFrame(payload)
 	var t *time.Timer
 	t = time.AfterFunc(1, func() {
-		if _, err := c.clientConn.Write(frame); err == nil && t != nil {
+		c.mu.RLock()
-			t.Reset(time.Second)
+		conn := c.clientConn
+		c.mu.RUnlock()
+		if conn != nil {
+			if _, err := conn.Write(frame); err == nil && t != nil {
+				t.Reset(time.Second)
+			}
 		}
 	})
 	defer t.Stop()
@@ -399,10 +406,6 @@ func (c *Conn) Close() error {
 		c.clientConn.Close()
 		c.clientConn = nil
 	}
-	if c.serverConn != nil {
-		c.serverConn.Close()
-		c.serverConn = nil
-	}
 	if c.frames != nil {
 		c.frames.Close()
 	}
@@ -705,7 +708,7 @@ func (c *Conn) handleSpeakerAVLogin() error {
 	}
 
 	buf := make([]byte, 1024)
-	c.serverConn.SetReadDeadline(time.Now().Add(5 * time.Second))
+	c.serverConn.SetReadDeadline(time.Now().Add(2 * time.Second))
 	n, err := c.serverConn.Read(buf)
 	if err != nil {
 		return fmt.Errorf("read av login: %w", err)

pkg/wyze/tutk/dtls.go

@@ -2,6 +2,7 @@ package tutk
 
 import (
 	"net"
+	"sync"
 	"time"
 
 	"github.com/pion/dtls/v3"
@@ -42,6 +43,9 @@ func buildDtlsConfig(psk []byte, isServer bool) *dtls.Config {
 type ChannelAdapter struct {
 	conn    *Conn
 	channel uint8
+
+	mu           sync.Mutex
+	readDeadline time.Time
 }
 
 func (a *ChannelAdapter) ReadFrom(p []byte) (n int, addr net.Addr, err error) {
@@ -52,6 +56,29 @@ func (a *ChannelAdapter) ReadFrom(p []byte) (n int, addr net.Addr, err error) {
 		buf = a.conn.serverBuf
 	}
 
+	a.mu.Lock()
+	deadline := a.readDeadline
+	a.mu.Unlock()
+
+	if !deadline.IsZero() {
+		timeout := time.Until(deadline)
+		if timeout <= 0 {
+			return 0, nil, &timeoutError{}
+		}
+
+		timer := time.NewTimer(timeout)
+		defer timer.Stop()
+
+		select {
+		case data := <-buf:
+			return copy(p, data), a.conn.addr, nil
+		case <-timer.C:
+			return 0, nil, &timeoutError{}
+		case <-a.conn.ctx.Done():
+			return 0, nil, net.ErrClosed
+		}
+	}
+
 	select {
 	case data := <-buf:
 		return copy(p, data), a.conn.addr, nil
@@ -67,8 +94,27 @@ func (a *ChannelAdapter) WriteTo(p []byte, _ net.Addr) (int, error) {
 	return len(p), nil
 }
 
-func (a *ChannelAdapter) Close() error                     { return nil }
+func (a *ChannelAdapter) Close() error        { return nil }
-func (a *ChannelAdapter) LocalAddr() net.Addr              { return &net.UDPAddr{} }
+func (a *ChannelAdapter) LocalAddr() net.Addr { return &net.UDPAddr{} }
-func (a *ChannelAdapter) SetDeadline(time.Time) error      { return nil }
+
-func (a *ChannelAdapter) SetReadDeadline(time.Time) error  { return nil }
+func (a *ChannelAdapter) SetDeadline(t time.Time) error {
+	a.mu.Lock()
+	a.readDeadline = t
+	a.mu.Unlock()
+	return nil
+}
+
+func (a *ChannelAdapter) SetReadDeadline(t time.Time) error {
+	a.mu.Lock()
+	a.readDeadline = t
+	a.mu.Unlock()
+	return nil
+}
+
 func (a *ChannelAdapter) SetWriteDeadline(time.Time) error { return nil }
+
+type timeoutError struct{}
+
+func (e *timeoutError) Error() string   { return "i/o timeout" }
+func (e *timeoutError) Timeout() bool   { return true }
+func (e *timeoutError) Temporary() bool { return true }

pkg/wyze/tutk/frame.go

@@ -4,6 +4,7 @@ import (
 	"encoding/binary"
 	"encoding/hex"
 	"fmt"
+	"sync"
 
 	"github.com/AlexxIT/go2rtc/pkg/aac"
 )
@@ -282,6 +283,8 @@ type FrameHandler struct {
 	audioTS  tsTracker
 	output   chan *Packet
 	verbose  bool
+	closed   bool
+	closeMu  sync.Mutex
 }
 
 func NewFrameHandler(verbose bool) *FrameHandler {
@@ -297,6 +300,13 @@ func (h *FrameHandler) Recv() <-chan *Packet {
 }
 
 func (h *FrameHandler) Close() {
+	h.closeMu.Lock()
+	defer h.closeMu.Unlock()
+
+	if h.closed {
+		return
+	}
+	h.closed = true
 	close(h.output)
 }
 
@@ -540,6 +550,13 @@ func (h *FrameHandler) handleAudio(payload []byte, fi *FrameInfo) {
 }
 
 func (h *FrameHandler) queue(pkt *Packet) {
+	h.closeMu.Lock()
+	defer h.closeMu.Unlock()
+
+	if h.closed {
+		return
+	}
+
 	select {
 	case h.output <- pkt:
 	default:
@@ -548,7 +565,11 @@ func (h *FrameHandler) queue(pkt *Packet) {
 		case <-h.output:
 		default:
 		}
-		h.output <- pkt
+		select {
+		case h.output <- pkt:
+		default:
+			// Queue still full, drop this packet
+		}
 	}
 }