Basic tutorial 8: Short-cutting the pipeline

Warning Please port this tutorial to python!

Warning Please port this tutorial to javascript!

Goal

Pipelines constructed with GStreamer do not need to be completely closed. Data can be injected into the pipeline and extracted from it at any time, in a variety of ways. This tutorial shows:

  • How to inject external data into a general GStreamer pipeline.

  • How to extract data from a general GStreamer pipeline.

  • How to access and manipulate this data.

Playback tutorial 3: Short-cutting the pipeline explains how to achieve the same goals in a playbin-based pipeline.

Introduction

Applications can interact with the data flowing through a GStreamer pipeline in several ways. This tutorial describes the easiest one, since it uses elements that have been created for this sole purpose.

The element used to inject application data into a GStreamer pipeline is appsrc, and its counterpart, used to extract GStreamer data back to the application is appsink. To avoid confusing the names, think of it from GStreamer's point of view: appsrc is just a regular source, that provides data magically fallen from the sky (provided by the application, actually). appsink is a regular sink, where the data flowing through a GStreamer pipeline goes to die (it is recovered by the application, actually).

appsrc and appsink are so versatile that they offer their own API (see their documentation), which can be accessed by linking against the gstreamer-app library. In this tutorial, however, we will use a simpler approach and control them through signals.

appsrc can work in a variety of modes: in pull mode, it requests data from the application every time it needs it. In push mode, the application pushes data at its own pace. Furthermore, in push mode, the application can choose to be blocked in the push function when enough data has already been provided, or it can listen to the enough-data and need-data signals to control flow. This example implements the latter approach. Information regarding the other methods can be found in the appsrc documentation.

Buffers

Data travels through a GStreamer pipeline in chunks called buffers. Since this example produces and consumes data, we need to know about GstBuffers.

Source Pads produce buffers, that are consumed by Sink Pads; GStreamer takes these buffers and passes them from element to element.

A buffer simply represents a unit of data, do not assume that all buffers will have the same size, or represent the same amount of time. Neither should you assume that if a single buffer enters an element, a single buffer will come out. Elements are free to do with the received buffers as they please. GstBuffers may also contain more than one actual memory buffer. Actual memory buffers are abstracted away using GstMemory objects, and a GstBuffer can contain multiple GstMemory objects.

Every buffer has attached time-stamps and duration, that describe in which moment the content of the buffer should be decoded, rendered or displayed. Time stamping is a very complex and delicate subject, but this simplified vision should suffice for now.

As an example, a filesrc (a GStreamer element that reads files) produces buffers with the “ANY” caps and no time-stamping information. After demuxing (see Basic tutorial 3: Dynamic pipelines) buffers can have some specific caps, for example “video/x-h264”. After decoding, each buffer will contain a single video frame with raw caps (“video/x-raw,format=...”) and very precise time stamps indicating when should that frame be displayed.

This tutorial

This tutorial expands Basic tutorial 7: Multithreading and Pad Availability in two ways: firstly, the audiotestsrc is replaced by an appsrc that will generate the audio data. Secondly, a new branch is added to the tee so data going into the audio sink and the wave display is also replicated into an appsink. The appsink uploads the information back into the application, which then just notifies the user that data has been received, but it could obviously perform more complex tasks.

A crude waveform generator

Copy this code into a text file named basic-tutorial-8.c (or find it in your GStreamer installation).

#include <gst/gst.h>
#include <gst/audio/audio.h>
#include <string.h>

#define CHUNK_SIZE 1024   /* Amount of bytes we are sending in each buffer */
#define SAMPLE_RATE 44100 /* Samples per second we are sending */

/* Structure to contain all our information, so we can pass it to callbacks */
typedef struct _CustomData {
  GstElement *pipeline, *app_source, *tee, *audio_queue, *audio_convert1, *audio_resample, *audio_sink;
  GstElement *video_queue, *audio_convert2, *visual, *video_convert, *video_sink;
  GstElement *app_queue, *app_sink;

  guint64 num_samples;   /* Number of samples generated so far (for timestamp generation) */
  gfloat a, b, c, d;     /* For waveform generation */

  guint sourceid;        /* To control the GSource */

  GMainLoop *main_loop;  /* GLib's Main Loop */
} CustomData;

/* This method is called by the idle GSource in the mainloop, to feed CHUNK_SIZE bytes into appsrc.
 * The idle handler is added to the mainloop when appsrc requests us to start sending data (need-data signal)
 * and is removed when appsrc has enough data (enough-data signal).
 */
static gboolean push_data (CustomData *data) {
  GstBuffer *buffer;
  GstFlowReturn ret;
  int i;
  GstMapInfo map;
  gint16 *raw;
  gint num_samples = CHUNK_SIZE / 2; /* Because each sample is 16 bits */
  gfloat freq;

  /* Create a new empty buffer */
  buffer = gst_buffer_new_and_alloc (CHUNK_SIZE);

  /* Set its timestamp and duration */
  GST_BUFFER_TIMESTAMP (buffer) = gst_util_uint64_scale (data->num_samples, GST_SECOND, SAMPLE_RATE);
  GST_BUFFER_DURATION (buffer) = gst_util_uint64_scale (num_samples, GST_SECOND, SAMPLE_RATE);

  /* Generate some psychodelic waveforms */
  gst_buffer_map (buffer, &map, GST_MAP_WRITE);
  raw = (gint16 *)map.data;
  data->c += data->d;
  data->d -= data->c / 1000;
  freq = 1100 + 1000 * data->d;
  for (i = 0; i < num_samples; i++) {
    data->a += data->b;
    data->b -= data->a / freq;
    raw[i] = (gint16)(500 * data->a);
  }
  gst_buffer_unmap (buffer, &map);
  data->num_samples += num_samples;

  /* Push the buffer into the appsrc */
  g_signal_emit_by_name (data->app_source, "push-buffer", buffer, &ret);

  /* Free the buffer now that we are done with it */
  gst_buffer_unref (buffer);

  if (ret != GST_FLOW_OK) {
    /* We got some error, stop sending data */
    return FALSE;
  }

  return TRUE;
}

/* This signal callback triggers when appsrc needs data. Here, we add an idle handler
 * to the mainloop to start pushing data into the appsrc */
static void start_feed (GstElement *source, guint size, CustomData *data) {
  if (data->sourceid == 0) {
    g_print ("Start feeding\n");
    data->sourceid = g_idle_add ((GSourceFunc) push_data, data);
  }
}

/* This callback triggers when appsrc has enough data and we can stop sending.
 * We remove the idle handler from the mainloop */
static void stop_feed (GstElement *source, CustomData *data) {
  if (data->sourceid != 0) {
    g_print ("Stop feeding\n");
    g_source_remove (data->sourceid);
    data->sourceid = 0;
  }
}

/* The appsink has received a buffer */
static GstFlowReturn new_sample (GstElement *sink, CustomData *data) {
  GstSample *sample;

  /* Retrieve the buffer */
  g_signal_emit_by_name (sink, "pull-sample", &sample);
  if (sample) {
    /* The only thing we do in this example is print a * to indicate a received buffer */
    g_print ("*");
    gst_sample_unref (sample);
    return GST_FLOW_OK;
  }

  return GST_FLOW_ERROR;
}

/* This function is called when an error message is posted on the bus */
static void error_cb (GstBus *bus, GstMessage *msg, CustomData *data) {
  GError *err;
  gchar *debug_info;

  /* Print error details on the screen */
  gst_message_parse_error (msg, &err, &debug_info);
  g_printerr ("Error received from element %s: %s\n", GST_OBJECT_NAME (msg->src), err->message);
  g_printerr ("Debugging information: %s\n", debug_info ? debug_info : "none");
  g_clear_error (&err);
  g_free (debug_info);

  g_main_loop_quit (data->main_loop);
}

int main(int argc, char *argv[]) {
  CustomData data;
  GstPad *tee_audio_pad, *tee_video_pad, *tee_app_pad;
  GstPad *queue_audio_pad, *queue_video_pad, *queue_app_pad;
  GstAudioInfo info;
  GstCaps *audio_caps;
  GstBus *bus;

  /* Initialize custom data structure */
  memset (&data, 0, sizeof (data));
  data.b = 1; /* For waveform generation */
  data.d = 1;

  /* Initialize GStreamer */
  gst_init (&argc, &argv);

  /* Create the elements */
  data.app_source = gst_element_factory_make ("appsrc", "audio_source");
  data.tee = gst_element_factory_make ("tee", "tee");
  data.audio_queue = gst_element_factory_make ("queue", "audio_queue");
  data.audio_convert1 = gst_element_factory_make ("audioconvert", "audio_convert1");
  data.audio_resample = gst_element_factory_make ("audioresample", "audio_resample");
  data.audio_sink = gst_element_factory_make ("autoaudiosink", "audio_sink");
  data.video_queue = gst_element_factory_make ("queue", "video_queue");
  data.audio_convert2 = gst_element_factory_make ("audioconvert", "audio_convert2");
  data.visual = gst_element_factory_make ("wavescope", "visual");
  data.video_convert = gst_element_factory_make ("videoconvert", "video_convert");
  data.video_sink = gst_element_factory_make ("autovideosink", "video_sink");
  data.app_queue = gst_element_factory_make ("queue", "app_queue");
  data.app_sink = gst_element_factory_make ("appsink", "app_sink");

  /* Create the empty pipeline */
  data.pipeline = gst_pipeline_new ("test-pipeline");

  if (!data.pipeline || !data.app_source || !data.tee || !data.audio_queue || !data.audio_convert1 ||
      !data.audio_resample || !data.audio_sink || !data.video_queue || !data.audio_convert2 || !data.visual ||
      !data.video_convert || !data.video_sink || !data.app_queue || !data.app_sink) {
    g_printerr ("Not all elements could be created.\n");
    return -1;
  }

  /* Configure wavescope */
  g_object_set (data.visual, "shader", 0, "style", 0, NULL);

  /* Configure appsrc */
  gst_audio_info_set_format (&info, GST_AUDIO_FORMAT_S16, SAMPLE_RATE, 1, NULL);
  audio_caps = gst_audio_info_to_caps (&info);
  g_object_set (data.app_source, "caps", audio_caps, "format", GST_FORMAT_TIME, NULL);
  g_signal_connect (data.app_source, "need-data", G_CALLBACK (start_feed), &data);
  g_signal_connect (data.app_source, "enough-data", G_CALLBACK (stop_feed), &data);

  /* Configure appsink */
  g_object_set (data.app_sink, "emit-signals", TRUE, "caps", audio_caps, NULL);
  g_signal_connect (data.app_sink, "new-sample", G_CALLBACK (new_sample), &data);
  gst_caps_unref (audio_caps);

  /* Link all elements that can be automatically linked because they have "Always" pads */
  gst_bin_add_many (GST_BIN (data.pipeline), data.app_source, data.tee, data.audio_queue, data.audio_convert1, data.audio_resample,
      data.audio_sink, data.video_queue, data.audio_convert2, data.visual, data.video_convert, data.video_sink, data.app_queue,
      data.app_sink, NULL);
  if (gst_element_link_many (data.app_source, data.tee, NULL) != TRUE ||
      gst_element_link_many (data.audio_queue, data.audio_convert1, data.audio_resample, data.audio_sink, NULL) != TRUE ||
      gst_element_link_many (data.video_queue, data.audio_convert2, data.visual, data.video_convert, data.video_sink, NULL) != TRUE ||
      gst_element_link_many (data.app_queue, data.app_sink, NULL) != TRUE) {
    g_printerr ("Elements could not be linked.\n");
    gst_object_unref (data.pipeline);
    return -1;
  }

  /* Manually link the Tee, which has "Request" pads */
  tee_audio_pad = gst_element_request_pad_simple (data.tee, "src_%u");
  g_print ("Obtained request pad %s for audio branch.\n", gst_pad_get_name (tee_audio_pad));
  queue_audio_pad = gst_element_get_static_pad (data.audio_queue, "sink");
  tee_video_pad = gst_element_request_pad_simple (data.tee, "src_%u");
  g_print ("Obtained request pad %s for video branch.\n", gst_pad_get_name (tee_video_pad));
  queue_video_pad = gst_element_get_static_pad (data.video_queue, "sink");
  tee_app_pad = gst_element_request_pad_simple (data.tee, "src_%u");
  g_print ("Obtained request pad %s for app branch.\n", gst_pad_get_name (tee_app_pad));
  queue_app_pad = gst_element_get_static_pad (data.app_queue, "sink");
  if (gst_pad_link (tee_audio_pad, queue_audio_pad) != GST_PAD_LINK_OK ||
      gst_pad_link (tee_video_pad, queue_video_pad) != GST_PAD_LINK_OK ||
      gst_pad_link (tee_app_pad, queue_app_pad) != GST_PAD_LINK_OK) {
    g_printerr ("Tee could not be linked\n");
    gst_object_unref (data.pipeline);
    return -1;
  }
  gst_object_unref (queue_audio_pad);
  gst_object_unref (queue_video_pad);
  gst_object_unref (queue_app_pad);

  /* Instruct the bus to emit signals for each received message, and connect to the interesting signals */
  bus = gst_element_get_bus (data.pipeline);
  gst_bus_add_signal_watch (bus);
  g_signal_connect (G_OBJECT (bus), "message::error", (GCallback)error_cb, &data);
  gst_object_unref (bus);

  /* Start playing the pipeline */
  gst_element_set_state (data.pipeline, GST_STATE_PLAYING);

  /* Create a GLib Main Loop and set it to run */
  data.main_loop = g_main_loop_new (NULL, FALSE);
  g_main_loop_run (data.main_loop);

  /* Release the request pads from the Tee, and unref them */
  gst_element_release_request_pad (data.tee, tee_audio_pad);
  gst_element_release_request_pad (data.tee, tee_video_pad);
  gst_element_release_request_pad (data.tee, tee_app_pad);
  gst_object_unref (tee_audio_pad);
  gst_object_unref (tee_video_pad);
  gst_object_unref (tee_app_pad);

  /* Free resources */
  gst_element_set_state (data.pipeline, GST_STATE_NULL);
  gst_object_unref (data.pipeline);
  return 0;
}

Information Need help?

If you need help to compile this code, refer to the Building the tutorials section for your platform: Linux, Mac OS X or Windows, or use this specific command on Linux:

gcc basic-tutorial-8.c -o basic-tutorial-8 `pkg-config --cflags --libs gstreamer-1.0 gstreamer-audio-1.0`

If you need help to run this code, refer to the Running the tutorials section for your platform: Linux, Mac OS X or Windows.

This tutorial plays an audible tone for varying frequency through the audio card and opens a window with a waveform representation of the tone. The waveform should be a sinusoid, but due to the refreshing of the window might not appear so.

Required libraries: gstreamer-1.0

Walkthrough

The code to create the pipeline (Lines 131 to 205) is an enlarged version of Basic tutorial 7: Multithreading and Pad Availability. It involves instantiating all the elements, link the elements with Always Pads, and manually link the Request Pads of the tee element.

Regarding the configuration of the appsrc and appsink elements:

/* Configure appsrc */
gst_audio_info_set_format (&info, GST_AUDIO_FORMAT_S16, SAMPLE_RATE, 1, NULL);
audio_caps = gst_audio_info_to_caps (&info);
g_object_set (data.app_source, "caps", audio_caps, NULL);
g_signal_connect (data.app_source, "need-data", G_CALLBACK (start_feed), &data);
g_signal_connect (data.app_source, "enough-data", G_CALLBACK (stop_feed), &data);

The first property that needs to be set on the appsrc is caps. It specifies the kind of data that the element is going to produce, so GStreamer can check if linking with downstream elements is possible (this is, if the downstream elements will understand this kind of data). This property must be a GstCaps object, which is easily built from a string with gst_caps_from_string().

We then connect to the need-data and enough-data signals. These are fired by appsrc when its internal queue of data is running low or almost full, respectively. We will use these signals to start and stop (respectively) our signal generation process.

/* Configure appsink */
g_object_set (data.app_sink, "emit-signals", TRUE, "caps", audio_caps, NULL);
g_signal_connect (data.app_sink, "new-sample", G_CALLBACK (new_sample), &data);
gst_caps_unref (audio_caps);

Regarding the appsink configuration, we connect to the new-sample signal, which is emitted every time the sink receives a buffer. Also, the signal emission needs to be enabled through the emit-signals property, because, by default, it is disabled.

Starting the pipeline, waiting for messages and final cleanup is done as usual. Let's review the callbacks we have just registered:

/* This signal callback triggers when appsrc needs data. Here, we add an idle handler
 * to the mainloop to start pushing data into the appsrc */
static void start_feed (GstElement *source, guint size, CustomData *data) {
  if (data->sourceid == 0) {
    g_print ("Start feeding\n");
    data->sourceid = g_idle_add ((GSourceFunc) push_data, data);
  }
}

This function is called when the internal queue of appsrc is about to starve (run out of data). The only thing we do here is register a GLib idle function with g_idle_add() that feeds data to appsrc until it is full again. A GLib idle function is a method that GLib will call from its main loop whenever it is “idle”, this is, when it has no higher-priority tasks to perform. It requires a GLib GMainLoop to be instantiated and running, obviously.

This is only one of the multiple approaches that appsrc allows. In particular, buffers do not need to be fed into appsrc from the main thread using GLib, and you do not need to use the need-data and enough-data signals to synchronize with appsrc (although this is allegedly the most convenient).

We take note of the sourceid that g_idle_add() returns, so we can disable it later.

/* This callback triggers when appsrc has enough data and we can stop sending.
 * We remove the idle handler from the mainloop */
static void stop_feed (GstElement *source, CustomData *data) {
  if (data->sourceid != 0) {
    g_print ("Stop feeding\n");
    g_source_remove (data->sourceid);
    data->sourceid = 0;
  }
}

This function is called when the internal queue of appsrc is full enough so we stop pushing data. Here we simply remove the idle function by using g_source_remove() (The idle function is implemented as a GSource).

/* This method is called by the idle GSource in the mainloop, to feed CHUNK_SIZE bytes into appsrc.
 * The ide handler is added to the mainloop when appsrc requests us to start sending data (need-data signal)
 * and is removed when appsrc has enough data (enough-data signal).
 */
static gboolean push_data (CustomData *data) {
  GstBuffer *buffer;
  GstFlowReturn ret;
  GstMapInfo map;
  int i;
  gint num_samples = CHUNK_SIZE / 2; /* Because each sample is 16 bits */
  gfloat freq;

  /* Create a new empty buffer */
  buffer = gst_buffer_new_and_alloc (CHUNK_SIZE);

  /* Set its timestamp and duration */
  GST_BUFFER_TIMESTAMP (buffer) = gst_util_uint64_scale (data->num_samples, GST_SECOND, SAMPLE_RATE);
  GST_BUFFER_DURATION (buffer) = gst_util_uint64_scale (num_samples, GST_SECOND, SAMPLE_RATE);

  /* Generate some psychodelic waveforms */
  if (gst_buffer_map (buf, &map, GST_MAP_READ)) {
    gint16 *raw = (gint16 *) map.data;

    /* create samples here */

    /* unmap buffer when done */
    gst_buffer_unmap (buf, &map);
  }

This is the function that feeds appsrc. It will be called by GLib at times and rates which are out of our control, but we know that we will disable it when its job is done (when the queue in appsrc is full).

Its first task is to create a new buffer with a given size (in this example, it is arbitrarily set to 1024 bytes) with gst_buffer_new_and_alloc().

We count the number of samples that we have generated so far with the CustomData.num_samples variable, so we can time-stamp this buffer using the GST_BUFFER_TIMESTAMP macro in GstBuffer.

Since we are producing buffers of the same size, their duration is the same and is set using the GST_BUFFER_DURATION in GstBuffer.

gst_util_uint64_scale() is a utility function that scales (multiply and divide) numbers which can be large, without fear of overflows.

In order access the memory of the buffer you first have to map it with gst_buffer_map(), which will give you a pointer and a size inside the GstMapInfo structure which gst_buffer_map() will populate on success. Be careful not to write past the end of the buffer: you allocated it, so you know its size in bytes and samples.

We will skip over the waveform generation, since it is outside the scope of this tutorial (it is simply a funny way of generating a pretty psychedelic wave).

/* Push the buffer into the appsrc */
g_signal_emit_by_name (data->app_source, "push-buffer", buffer, &ret);

/* Free the buffer now that we are done with it */
gst_buffer_unref (buffer);

Note that there is also gst_app_src_push_buffer() as part of the gstreamer-app-1.0 library, which is perhaps a better function to use to push a buffer into appsrc than the signal emission above, because it has a proper type signature so it's harder to get wrong. However, be aware that if you use gst_app_src_push_buffer() it will take ownership of the buffer passed instead, so in that case you won't have to unref it after pushing.

Once we have the buffer ready, we pass it to appsrc with the push-buffer action signal (see information box at the end of Playback tutorial 1: Playbin usage), and then gst_buffer_unref() it since we no longer need it.

/* The appsink has received a buffer */
static GstFlowReturn new_sample (GstElement *sink, CustomData *data) {
  GstSample *sample;
  /* Retrieve the buffer */
  g_signal_emit_by_name (sink, "pull-sample", &sample);
  if (sample) {
    /* The only thing we do in this example is print a * to indicate a received buffer */
    g_print ("*");
    gst_sample_unref (sample);
    return GST_FLOW_OK;
  }
  return GST_FLOW_ERROR;
}

Finally, this is the function that gets called when the appsink receives a buffer. We use the pull-sample action signal to retrieve the buffer and then just print some indicator on the screen.

Note that there is also gst_app_src_pull_sample() as part of the gstreamer-app-1.0 library, which is perhaps a better function to use to pull a sample/buffer out of an appsink than the signal emission above, because it has a proper type signature so it's harder to get wrong.

In order to get to the data pointer we need to use gst_buffer_map() just like above, which will populate a GstMapInfo helper struct with a pointer to the data and the size of the data in bytes. Don't forget to gst_buffer_unmap() the buffer again when done with the data.

Remember that this buffer does not have to match the buffer that we produced in the push_data function, any element in the path could have altered the buffers in any way (Not in this example: there is only a tee in the path between appsrc and appsink, and the tee does not change the content of the buffers).

We then gst_sample_unref() the retrieved sample, and this tutorial is done.

Conclusion

This tutorial has shown how applications can:

  • Inject data into a pipeline using the appsrcelement.
  • Retrieve data from a pipeline using the appsink element.
  • Manipulate this data by accessing the GstBuffer.

In a playbin-based pipeline, the same goals are achieved in a slightly different way. Playback tutorial 3: Short-cutting the pipeline shows how to do it.

It has been a pleasure having you here, and see you soon!

The results of the search are