Add -artp option for audio RTP output

Add a new -artp command-line option that routes decoded audio to an RTP
stream instead of the local audio sink, following the existing -vrtp
pattern for video.

Usage: uxplay -artp "pt=96 ! udpsink host=127.0.0.1 port=5002"

The implementation:
- Decodes audio (AAC-ELD/ALAC) to PCM
- Converts to S16BE format required by rtpL16pay
- Preserves volume control for iOS volume adjustment
- Sends L16 RTP packets (16-bit signed big-endian, 44100Hz, stereo)

README.md

@@ -20,7 +20,9 @@
 -   option `-vrtp <rest-of-pipeline>`  bypasses rendering by UxPlay, and instead
     transmits rtp packets of decrypted h264 or h265 video to
     an external renderer (e.g. OBS Studio) at an address specified in `rest-of-pipeline`.
-    (Note: this is video only, an option "-rtp" which muxes audio and video into a mpeg4 container still needs to be created:
+    Similarly, `-artp <rest-of-pipeline>` forwards decoded audio as L16 RTP packets.
+    Both options can be used together to forward video and audio (as separate concurrent streams) to external applications.
+    (Note: an option "-rtp" which muxes audio and video into a mpeg4 container still needs to be created:
     Pull Requests welcomed).
     
 -   (for Linux/*BSD Desktop Environments using D-Bus). New option `-scrsv <n>` provides screensaver inhibition (e.g., to
@@ -1313,6 +1315,11 @@ Uses rtph264pay or rtph265pay as appropriate: *pipeline* should start with any
 rtph26xpay options (such as config_interval= or aggregate-mode =), followed by
 a sending method:  *e.g.*, `"config-interval=1 ! udpsink host=127.0.0.1 port=5000`".
 
+**-artp *pipeline***:  forward decoded audio as L16 RTP packets to somewhere else, without local playback.
+Uses rtpL16pay (16-bit signed big-endian PCM, 44100Hz stereo): *pipeline* should start with any
+rtpL16pay options (such as pt=), followed by a sending method:
+*e.g.*, `"pt=96 ! udpsink host=127.0.0.1 port=5002"`.  iOS volume control still works over RTP.
+
 **-v4l2** Video settings for hardware h264 video decoding in the GPU by
 Video4Linux2. Equivalent to `-vd v4l2h264dec -vc v4l2convert`.
 

renderers/audio_renderer.c

@@ -40,6 +40,7 @@ static gboolean render_audio = FALSE;
 static gboolean async = FALSE;
 static gboolean vsync = FALSE;
 static gboolean sync = FALSE;
+static gboolean audio_rtp = FALSE;
 
 typedef struct audio_renderer_s {
     GstElement *appsrc; 
@@ -124,12 +125,17 @@ bool gstreamer_init(){
     return (bool) check_plugins ();
 }
 
-void audio_renderer_init(logger_t *render_logger, const char* audiosink, const bool* audio_sync, const bool* video_sync) {
+void audio_renderer_init(logger_t *render_logger, const char* audiosink, const bool* audio_sync, const bool* video_sync, const char *artp_pipeline) {
     GError *error = NULL;
     GstCaps *caps = NULL;
     GstClock *clock = gst_system_clock_obtain();
     g_object_set(clock, "clock-type", GST_CLOCK_TYPE_REALTIME, NULL);
 
+    audio_rtp = (bool) strlen(artp_pipeline);
+    if (audio_rtp) {
+        g_print("*** Audio RTP mode enabled: sending to %s\n", artp_pipeline);
+    }
+
     logger = render_logger;
     
     aac = check_plugin_feature (avdec_aac);
@@ -155,27 +161,38 @@ void audio_renderer_init(logger_t *render_logger, const char* audiosink, const b
         }
         g_string_append (launch, "audioconvert ! ");
         g_string_append (launch, "audioresample ! ");    /* wasapisink must resample from 44.1 kHz to 48 kHz */
-        g_string_append (launch, "volume name=volume ! level ! ");
+        g_string_append (launch, "volume name=volume ! ");
-        g_string_append (launch, audiosink);
+
-        switch(i) {
+        if (!audio_rtp) {
-        case 1:  /*ALAC*/
+            /* Normal path: local audio output */
-            if (*audio_sync) {
+            g_string_append (launch, "level ! ");
-                g_string_append (launch, " sync=true");
+            g_string_append (launch, audiosink);
-                async = TRUE;
+            switch(i) {
-            } else {
+            case 1:  /*ALAC*/
-                g_string_append (launch, " sync=false");
+                if (*audio_sync) {
-                async = FALSE;
+                    g_string_append (launch, " sync=true");
+                    async = TRUE;
+                } else {
+                    g_string_append (launch, " sync=false");
+                    async = FALSE;
+                }
+                break;
+            default:
+                if (*video_sync) {
+                    g_string_append (launch, " sync=true");
+                    vsync = TRUE;
+                } else {
+                    g_string_append (launch, " sync=false");
+                    vsync = FALSE;
+                }
+                break;
             }
-            break;
+        } else {
-        default:
+            /* RTP path: send decoded PCM over RTP */
-	    if (*video_sync) {
+            /* rtpL16pay requires S16BE (big-endian) format */
-                g_string_append (launch, " sync=true");
+            g_string_append (launch, "audioconvert ! audio/x-raw,format=S16BE,rate=44100,channels=2 ! ");
-                vsync = TRUE;
+            g_string_append (launch, "rtpL16pay ");
-            } else {
+            g_string_append (launch, artp_pipeline);
-                g_string_append (launch, " sync=false");
-                vsync = FALSE;
-            }
-            break;
         }
         renderer_type[i]->pipeline  = gst_parse_launch(launch->str, &error);
 	if (error) {

renderers/audio_renderer.h

@@ -33,7 +33,7 @@ extern "C" {
 #include "../lib/logger.h"
 
 bool gstreamer_init();
-void audio_renderer_init(logger_t *logger, const char* audiosink, const bool *audio_sync, const bool *video_sync);
+void audio_renderer_init(logger_t *logger, const char* audiosink, const bool *audio_sync, const bool *video_sync, const char *artp_pipeline);
 void audio_renderer_start(unsigned char* compression_type);
 void audio_renderer_stop();
 void audio_renderer_render_buffer(unsigned char* data, int *data_len, unsigned short *seqnum, uint64_t *ntp_time);

uxplay.1

@@ -107,6 +107,12 @@ UxPlay 1.73: An open\-source AirPlay mirroring (+ audio streaming) server:
    is the remaining pipeline, starting with rtph26*pay options:
 .IP
    e.g. "config-interval=1 ! udpsink host=127.0.0.1 port=5000"
+.TP
+\fB\-artp\fI pl\fR  Use rtpL16pay to send decoded audio elsewhere: "pl"
+.IP
+   is the remaining pipeline, starting with rtpL16pay options:
+.IP
+   e.g. "pt=96 ! udpsink host=127.0.0.1 port=5002"
 .PP
 .TP
 \fB\-v4l2\fR     Use Video4Linux2 for GPU hardware h264 video decoding.

uxplay.cpp

@@ -195,6 +195,7 @@ static std::string artist;
 static std::string coverart_artist;
 static std::string ble_filename = "";
 static std::string rtp_pipeline = "";
+static std::string audio_rtp_pipeline = "";
 static GMainLoop *gmainloop = NULL;
 
 //Support for D-Bus-based screensaver inhibition (org.freedesktop.ScreenSaver) 
@@ -959,6 +960,9 @@ static void print_info (char *name) {
     printf("          some choices:pulsesink,alsasink,pipewiresink,jackaudiosink,\n");
     printf("          osssink,oss4sink,osxaudiosink,wasapisink,directsoundsink.\n");
     printf("-as 0     (or -a)  Turn audio off, streamed video only\n");
+    printf("-artp pl  Use rtpL16pay to send decoded audio elsewhere: \"pl\"\n");
+    printf("          is the remaining pipeline, starting with rtpL16pay options:\n");
+    printf("          e.g. \"pt=96 ! udpsink host=127.0.0.1 port=5002\"\n");
     printf("-al x     Audio latency in seconds (default 0.25) reported to client.\n");
     printf("-ca [<fn>]In Audio (ALAC) mode, render cover-art [or write to file <fn>]\n");
     printf("-md <fn>  In Airplay Audio (ALAC) mode, write metadata text to file <fn>\n");
@@ -1444,7 +1448,15 @@ static void parse_arguments (int argc, char *argv[]) {
           }
 	  rtp_pipeline.erase();
 	  rtp_pipeline.append(argv[++i]);
-        } else if (arg == "-vdmp") {
+	} else if (arg == "-artp") {
+	  if (!option_has_value(i, argc, arg, argv[i+1])) {
+	    fprintf(stderr,"option \"-artp\" must be followed by a pipeline for sending the audio stream:\n"
+		    "e.g., \"<rtpL16pay options> ! udpsink host=127.0.0.1 port=5002\"\n");
+	    exit(1);
+          }
+	  audio_rtp_pipeline.erase();
+	  audio_rtp_pipeline.append(argv[++i]);
+	} else if (arg == "-vdmp") {
             dump_video = true;
             if (i < argc - 1 && *argv[i+1] != '-') {
                 unsigned int n = 0;
@@ -3036,7 +3048,7 @@ int main (int argc, char *argv[]) {
     logger_set_level(render_logger, log_level);
 
     if (use_audio) {
-        audio_renderer_init(render_logger, audiosink.c_str(), &audio_sync, &video_sync);
+        audio_renderer_init(render_logger, audiosink.c_str(), &audio_sync, &video_sync, audio_rtp_pipeline.c_str());
     } else {
         LOGI("audio_disabled");
     }