3 A source element is an element that provides data to the pipeline. It
4 does typically not have any sink (input) pads.
6 Typical source elements include:
10 - network elements (live or not)
12 - capture elements (video/audio/…)
14 - generators (signals/video/audio/…)
18 A source is said to be a live source when it has the following property:
20 - temporarily stopping reading from the source causes data to be lost.
22 In general when this property holds, the source also produces data at a
23 fixed rate. Most sources have a limit on the rate at which they can
24 deliver data, which might be faster or slower than the consumption rate.
25 This property however does not make them a live source.
27 Let’s look at some example sources.
29 - file readers: you can PAUSE without losing data. There is however a
30 limit to how fast you can read from this source. This limit is
31 usually much higher than the consumption rate. In some cases it
32 might be slower (an NFS share, for example) in which case you might
33 need to use some buffering (see [buffering](design/buffering.md)).
35 - HTTP network element: you can PAUSE without data loss. Depending on
36 the available network bandwidth, consumption rate might be higher
37 than production rate in which case buffering should be used (see
38 [buffering](design/buffering.md)).
40 - audio source: pausing the audio capture will lead to lost data. this
41 source is therefore definitely live. In addition, an audio source
42 will produce data at a fixed rate (the samplerate). Also depending
43 on the buffersize, this source will introduce a latency (see
44 [latency](design/latency.md)).
46 - udp network source: Pausing the receiving part will lead to lost
47 data. This source is therefore a live source. Also in a typical case
48 the udp packets will be received at a certain rate, which might be
49 difficult to guess because of network jitter. This source does not
50 necessarily introduce latency on its own.
52 - dvb source: PAUSING this element will lead to data loss, it’s a live
53 source similar to a UDP source.
57 A source element can operate in three ways:
59 - it is fully seekable, this means that random access can be performed
60 on it in an efficient way. (a file reader,…). This also typically
61 means that the source is not live.
63 - data can be obtained from it with a variable size. This means that
64 the source can give N bytes of data. An example is an audio source.
65 A video source always provides the same amount of data (one video
66 frame). Note that this is not a fully seekable source.
68 - it is a live source, see above.
70 When writing a source, one has to look at how the source can operate to
71 decide on the scheduling methods to implement on the source.
73 - fully seekable sources implement a getrange function on the source
76 - sources that can give N bytes but cannot do seeking also implement a
77 getrange function but state that they cannot do random access.
79 - sources that are purely live sources implement a task to push out
82 Any source that has a getrange function must also implement a push based
83 scheduling mode. In this mode the source starts a task that gets N bytes
84 and pushes them out. Whenever possible, the peer element will select the
85 getrange based scheduling method of the source, though.
87 A source with a getrange function must activate itself in the pad
88 activate function. This is needed because the downstream peer element
89 will decide and activate the source element in its state change function
90 before the source’s state change function is called.
92 ## Source base classes
96 This base class provides an implementation of a random access source and
97 is very well suited for file reader like sources.
101 Base class for block-based sources. This class is mostly useful for
102 elements that cannot do random access, or at least very slowly. The
103 source usually prefers to push out a fixed size buffer.
105 Classes extending this base class will usually be scheduled in a push
106 based mode. If the peer accepts to operate without offsets and within
107 the limits of the allowed block size, this class can operate in getrange
108 based mode automatically.
110 The subclass should extend the methods from the baseclass in addition to
111 the create method. If the source is seekable, it needs to override
112 `GstBaseSrc::event()` in addition to `GstBaseSrc::is_seekable()` in order
113 to retrieve the seek offset, which is the offset of the next buffer to
116 Flushing, scheduling and sync is all handled by this base class.
120 A non-live source should timestamp the buffers it produces starting from
121 0. If it is not possible to timestamp every buffer (filesrc), the source
122 is allowed to only timestamp the first buffer (as 0).
124 Live sources only produce data in the `PLAYING` state, when the clock is
125 running. They should timestamp each buffer they produce with the current
126 `running_time` of the pipeline, which is expressed as:
129 absolute_time - base_time
132 With `absolute_time` being the time obtained from the global pipeline with
133 `gst_clock_get_time()` and `base_time` being the time of that clock when
134 the pipeline was last set to `PLAYING`.