diff --git a/internal/rtmp/conn.go b/internal/rtmp/conn.go
index af2d3d12..673b66bb 100644
--- a/internal/rtmp/conn.go
+++ b/internal/rtmp/conn.go
@@ -584,7 +584,7 @@ func (c *Conn) WriteMessage(msg message.Message) error {
 	return c.mrw.Write(msg)
 }
 
-func trackFromH264DecoderConfig(data []byte) (*format.H264, error) {
+func trackFromH264DecoderConfig(data []byte) (format.Format, error) {
 	var conf h264conf.Conf
 	err := conf.Unmarshal(data)
 	if err != nil {
@@ -766,9 +766,9 @@ func (c *Conn) readTracksFromMetadata(payload []interface{}) (format.Format, *fo
 	}
 }
 
-func (c *Conn) readTracksFromMessages(msg message.Message) (*format.H264, *format.MPEG4Audio, error) {
+func (c *Conn) readTracksFromMessages(msg message.Message) (format.Format, *format.MPEG4Audio, error) {
 	var startTime *time.Duration
-	var videoTrack *format.H264
+	var videoTrack format.Format
 	var audioTrack *format.MPEG4Audio
 
 	// analyze 1 second of packets
diff --git a/internal/rtmp/conn_test.go b/internal/rtmp/conn_test.go
index 7c969ef7..80a72a0a 100644
--- a/internal/rtmp/conn_test.go
+++ b/internal/rtmp/conn_test.go
@@ -5,6 +5,7 @@ import (
 	"net"
 	"net/url"
 	"testing"
+	"time"
 
 	"github.com/aler9/gortsplib/v2/pkg/codecs/h264"
 	"github.com/aler9/gortsplib/v2/pkg/codecs/mpeg4audio"
@@ -537,7 +538,7 @@ func TestReadTracks(t *testing.T) {
 			},
 		},
 		{
-			"missing metadata",
+			"missing metadata, video+audio",
 			&format.H264{
 				PayloadTyp:        96,
 				SPS:               sps,
@@ -556,6 +557,21 @@ func TestReadTracks(t *testing.T) {
 				IndexDeltaLength: 3,
 			},
 		},
+		{
+			"missing metadata, audio",
+			nil,
+			&format.MPEG4Audio{
+				PayloadTyp: 96,
+				Config: &mpeg4audio.Config{
+					Type:         2,
+					SampleRate:   44100,
+					ChannelCount: 2,
+				},
+				SizeLength:       13,
+				IndexLength:      3,
+				IndexDeltaLength: 3,
+			},
+		},
 		{
 			"obs studio h265",
 			&format.H265{
@@ -796,6 +812,7 @@ func TestReadTracks(t *testing.T) {
 					SPS: sps,
 					PPS: pps,
 				}.Marshal()
+
 				err = mrw.Write(&message.MsgVideo{
 					ChunkStreamID:   message.MsgVideoChunkStreamID,
 					MessageStreamID: 0x1000000,
@@ -811,6 +828,7 @@ func TestReadTracks(t *testing.T) {
 					ChannelCount: 2,
 				}.Marshal()
 				require.NoError(t, err)
+
 				err = mrw.Write(&message.MsgAudio{
 					ChunkStreamID:   message.MsgAudioChunkStreamID,
 					MessageStreamID: 0x1000000,
@@ -855,6 +873,7 @@ func TestReadTracks(t *testing.T) {
 					SPS: sps,
 					PPS: pps,
 				}.Marshal()
+
 				err = mrw.Write(&message.MsgVideo{
 					ChunkStreamID:   message.MsgVideoChunkStreamID,
 					MessageStreamID: 0x1000000,
@@ -893,6 +912,7 @@ func TestReadTracks(t *testing.T) {
 					SPS: sps,
 					PPS: pps,
 				}.Marshal()
+
 				err = mrw.Write(&message.MsgVideo{
 					ChunkStreamID:   message.MsgVideoChunkStreamID,
 					MessageStreamID: 0x1000000,
@@ -908,6 +928,7 @@ func TestReadTracks(t *testing.T) {
 					ChannelCount: 2,
 				}.Marshal()
 				require.NoError(t, err)
+
 				err = mrw.Write(&message.MsgAudio{
 					ChunkStreamID:   message.MsgAudioChunkStreamID,
 					MessageStreamID: 0x1000000,
@@ -919,11 +940,12 @@ func TestReadTracks(t *testing.T) {
 				})
 				require.NoError(t, err)
 
-			case "missing metadata":
+			case "missing metadata, video+audio":
 				buf, _ := h264conf.Conf{
 					SPS: sps,
 					PPS: pps,
 				}.Marshal()
+
 				err = mrw.Write(&message.MsgVideo{
 					ChunkStreamID:   message.MsgVideoChunkStreamID,
 					MessageStreamID: 0x1000000,
@@ -939,6 +961,7 @@ func TestReadTracks(t *testing.T) {
 					ChannelCount: 2,
 				}.Marshal()
 				require.NoError(t, err)
+
 				err = mrw.Write(&message.MsgAudio{
 					ChunkStreamID:   message.MsgAudioChunkStreamID,
 					MessageStreamID: 0x1000000,
@@ -950,6 +973,37 @@ func TestReadTracks(t *testing.T) {
 				})
 				require.NoError(t, err)
 
+			case "missing metadata, audio":
+				enc, err := mpeg4audio.Config{
+					Type:         2,
+					SampleRate:   44100,
+					ChannelCount: 2,
+				}.Marshal()
+				require.NoError(t, err)
+
+				err = mrw.Write(&message.MsgAudio{
+					ChunkStreamID:   message.MsgAudioChunkStreamID,
+					MessageStreamID: 0x1000000,
+					Rate:            flvio.SOUND_44Khz,
+					Depth:           flvio.SOUND_16BIT,
+					Channels:        flvio.SOUND_STEREO,
+					AACType:         flvio.AAC_SEQHDR,
+					Payload:         enc,
+				})
+				require.NoError(t, err)
+
+				err = mrw.Write(&message.MsgAudio{
+					ChunkStreamID:   message.MsgAudioChunkStreamID,
+					MessageStreamID: 0x1000000,
+					Rate:            flvio.SOUND_44Khz,
+					Depth:           flvio.SOUND_16BIT,
+					Channels:        flvio.SOUND_STEREO,
+					AACType:         flvio.AAC_SEQHDR,
+					Payload:         enc,
+					DTS:             1 * time.Second,
+				})
+				require.NoError(t, err)
+
 			case "obs studio h265":
 				err = mrw.Write(&message.MsgDataAMF0{
 					ChunkStreamID:   4,
@@ -1017,6 +1071,7 @@ func TestReadTracks(t *testing.T) {
 					ChannelCount: 2,
 				}.Marshal()
 				require.NoError(t, err)
+
 				err = mrw.Write(&message.MsgAudio{
 					ChunkStreamID:   message.MsgAudioChunkStreamID,
 					MessageStreamID: 0x1000000,