GStreamer: Difference between revisions

From LinuxTVWiki
Jump to navigation Jump to search
m (Minor fixes)
 
(42 intermediate revisions by 2 users not shown)
Line 1: Line 1:
GStreamer is a toolkit for building audio- and video-processing pipelines. A pipeline might stream video from a file to a network, or add an echo to a recording, or (most interesting to us) capture the output of a Video4Linux device. Gstreamer is most often used to power graphical applications such as [https://wiki.gnome.org/Apps/Videos Totem], but this page will explain how to build an encoder using its command-line interface.
GStreamer is a toolkit for building audio- and video-processing pipelines. A pipeline might stream video from a file to a network, or add an echo to a recording, or (most interesting to us) capture the output of a Video4Linux device. Gstreamer is most often used to power graphical applications such as [https://wiki.gnome.org/Apps/Videos Totem], but can also be used directly from the command-line. This page will explain how GStreamer is better than the alternatives, and how to build an encoder using its command-line interface.


'''Before reading this page''', see [[V4L_capturing|V4L capturing]] to set your system up and create an initial recording. This page assumes you have already implemented the simple pipeline described there.
== Getting Started with GStreamer ==

== Introduction to GStreamer ==

No two use cases for encoding are quite alike. What's your preferred workflow? Is your processor fast enough to encode high quality video in real-time? Do you have enough disk space to store the raw video then process it after the fact? Do you want to play your video in DVD players, or is it enough that it works in your version of [http://www.videolan.org/vlc/index.en_GB.html VLC]? How will you work around your system's obscure quirks?


'''Use GStreamer if''' you want the best video quality possible with your hardware, and don't mind spending a weekend browsing the Internet for information.
'''Use GStreamer if''' you want the best video quality possible with your hardware, and don't mind spending a weekend browsing the Internet for information.
Line 7: Line 11:
'''Avoid GStreamer if''' you just want something quick-and-dirty, or can't stand programs with bad documentation and unhelpful error messages.
'''Avoid GStreamer if''' you just want something quick-and-dirty, or can't stand programs with bad documentation and unhelpful error messages.


=== Why is GStreamer better at encoding? ===
GStreamer, its most common plugins and tools are available through your distribution's package manager. But <code>entrans</code> and some of the plugins used in the examples below are not. You can find their sources bundled by the [http://sourceforge.net/projects/gentrans/files/gst-entrans/ GEntrans project] at sourceforge. Google may help you to find precompiled packages for your distro.
Two series of GStreamer are available - ''0.10'' and ''1.0''. Most Linux distributions include both, but this page discusses the older ''0.10'' series because I was unable to get the ''1.0'' series to work with my TV card. Converting the commands below to work with ''1.0'' is mostly just search-and-replace work (e.g. changing instances of <code>ff</code> to <code>av</code> because of the switch from <code>ffmpeg</code> to <code>libavcodec</code>). See [http://gstreamer.freedesktop.org/data/doc/gstreamer/head/manual/html/chapter-porting-1.0.html the porting guide] for more.


GStreamer isn't as easy to use as <code>mplayer</code>, and doesn't have as advanced editing functionality as <code>ffmpeg</code>. But it has superior support for synchronising audio and video in disturbed sources such as VHS tapes. If you specify your input is (say) 25 frames per second video and 48,000Hz audio, most tools will synchronise audio and video simply by writing 1 video frame, 1,920 audio frames, 1 video frame and so on. There are at least three ways this can cause errors:
=== Using GStreamer with gst-launch ===

* '''initialisation timing''': audio and video desynchronised by a certain amount from the first frame, usually caused by audio and video devices taking different amounts of time to initialise. For example, the first audio frame might be delivered to GStreamer 0.01 seconds after it was requested, but the first video frame might not be delivered until 0.7 seconds after it was requested, causing all video to be 0.6 seconds behind the audio
** <code>mencoder</code>'s ''-delay'' option solves this by delaying the audio
* '''failure to encode''': frames that desynchronise gradually over time, usually caused by audio and video shifting relative to each other when frames are dropped. For example if your CPU is not fast enough and sometimes drops a video frame, after 25 dropped frames the video will be one second ahead of the audio
** <code>mencoder</code>'s ''-harddup'' option solves this by duplicating other frames to fill in the gaps
* '''source frame rate''': frames that aren't delivered at the advertised rate, usually caused by inaccurate clocks in the source hardware. For example, a low-cost webcam that advertises 25 FPS video and 48kHz audio might actually deliver 25.01 video frames and 47,999 audio frames per second, causing your audio and video to drift apart by a second or so per hour
** video tapes are especially problematic here - if you've ever seen a VCR struggle during those few seconds between two recordings on a tape, you've seen them adjusting the tape speed to accurately track the source. Frame counts can vary enough during these periods to instantly desynchronise audio and video
** <code>mencoder</code> has no solution for this problem

GStreamer solves these problems by attaching a timestamp to each incoming frame based on the time GStreamer receives the frame. It can then mux the sources back together accurately using these timestamps, either by using a format that supports variable framerates or by duplicating frames to fill in the blanks:
# If you choose a container format that supports timestamps (e.g. Matroska), timestamps are automatically written to the file and used to vary the playback speed
# If you choose a container format that does not support timestamps (e.g. AVI), you must duplicate other frames to fill in the gaps by adding the <code>videorate</code> and <code>audiorate</code> plugins to the end of the relevant pipelines

=== Getting GStreamer ===

GStreamer, its most common plugins and tools are available through your distribution's package manager. Most Linux distributions include both the legacy ''0.10'' and modern ''1.0'' release series - each has bugs that stop them from working on some hardware, and this page focuses mostly on the modern ''1.0'' series. Converting between ''0.10'' and ''1.0'' is mostly just search-and-replace work (e.g. changing instances of <code>av</code> to <code>ff</code> because of the switch from <code>ffmpeg</code> to <code>libavcodec</code>). See [http://gstreamer.freedesktop.org/data/doc/gstreamer/head/manual/html/chapter-porting-1.0.html the porting guide] for more.

Other plugins are also available, such as <code>[http://gentrans.sourceforge.net/ GEntrans]</code> (used in some examples below). Google might help you find packages for your distribution, otherwise you'll need to download and compile them yourself.

=== Using GStreamer with gst-launch-1.0 ===


<code>gst-launch</code> is the standard command-line interface to GStreamer. Here's the simplest pipline you can build:
<code>gst-launch</code> is the standard command-line interface to GStreamer. Here's the simplest pipline you can build:


gst-launch-0.10 fakesrc ! fakesink
gst-launch-1.0 fakesrc ! fakesink


This connects a single (fake) source to a single (fake) sink using the 0.10 series of GStreamer:
This connects a single (fake) source to a single (fake) sink using the 1.0 series of GStreamer:


[[File:GStreamer-simple-pipeline.png]]
[[File:GStreamer-simple-pipeline.png|center|Very simple pipeline]]


GStreamer can build all kinds of pipelines, but you probably want to build one that looks something like this:
To learn more about the source and sink elements, do:


[[File:Example-pipeline.png|center|Idealised pipeline example]]
gst-inspect-0.10 fakesrc
gst-inspect-0.10 sink


To get a list of elements that can go in a GStreamer pipeline, do:
If you have installed [http://www.graphviz.org Graphviz], you can build a graph like the above yourself:


gst-inspect-1.0 | less
mkdir gst-visualisations
GST_DEBUG_DUMP_DOT_DIR=gst-visualisations gst-launch-0.10 fakesrc ! fakesink
dot -Tpng gst-visualisations/*-gst-launch.PLAYING_READY.dot > my-pipeline.png


Pass an element name to <code>gst-inspect-1.0</code> for detailed information. For example:
To get graphs of the example pipelines below, prepend <code>GST_DEBUG_DUMP_DOT_DIR=gst-visualisations </code> to the <code>gst-launch</code> command. Run this command to generate a PNG version of GStreamer's most interesting stage:


gst-inspect-1.0 fakesrc
dot -Tpng gst-visualisations/*-gst-launch.PLAYING_READY.dot > my-pipeline.png
gst-inspect-1.0 fakesink


The images above are based on graphs created by GStreamer itself. Install [http://www.graphviz.org Graphviz] to build graphs of your pipelines. For faster viewing of those graphs, you may install xdot from [http://www.semicomplete.com/projects/xdotool/]:
Remember to empty the <code>gst-visualisations</code> directory between runs.


mkdir gst-visualisations
=== Using GStreamer with entrans ===
GST_DEBUG_DUMP_DOT_DIR=gst-visualisations gst-launch-1.0 fakesrc ! fakesink
xdot gst-visualisations/*-gst-launch.*_READY.dot


You may also compiles those graph to PNG, SVG or other supported formats:
<code>gst-launch</code> is the main command-line interface to GStreamer, available by default. But <code>entrans</code> is a bit smarter:


dot -Tpng gst-visualisations/*-gst-launch.*_READY.dot > my-pipeline.png
* it provides partly-automatically composing of GStreamer pipelines
* it allows cutting of streams; e.g. to capture for a predefined duration. That ensures headers are written correctly, which is not always the case if you close <code>gst-launch</code> by pressing Ctrl+C. To use this feature one has to insert a ''dam'' element after the first ''queue'' of each part of the pipeline


To get graphs of the example pipelines below, prepend <code>GST_DEBUG_DUMP_DOT_DIR=gst-visualisations </code> to the <code>gst-launch-1.0</code> command. Run this command to generate a graph of GStreamer's most interesting stage:
==Using GStreamer for V4L TV capture==


xdot gst-visualisations/*-gst-launch.PLAYING_READY.dot
No two use cases for encoding are quite alike. Is your processor fast enough to encode high quality video? Do you want to play your video in DVD players, or is it enough that it works in your version of [http://www.videolan.org/vlc/index.en_GB.html VLC]? Which obscure quirks does your system have?


Remember to empty the <code>gst-visualisations</code> directory between runs.
=== Why prefer GStreamer? ===


=== Using GStreamer with entrans ===
GStreamer is better than most tools at synchronising audio with video in disturbed sources such as VHS tapes. If you specify your input is (say) 25 frames per second video and 48,000kHz audio, most tools will synchronise audio and video simply by writing 1 video frame, 1,920 audio frames, 1 video frame and so on. There are at least three ways this can cause errors:


<code>gst-launch-1.0</code> is the main command-line interface to GStreamer, available by default. But <code>entrans</code> is a bit smarter:
* '''initialisation timing''': audio and video desynchronised by a certain amount from the first frame, usually caused by audio and video devices taking different amounts of time to initialise. For example, the first audio frame might be delivered to GStreamer 0.01 seconds after it was requested, but the first video frame might not be delivered until 0.7 seconds after it was requested, causing all video to be 0.6 seconds behind the audio
** <code>mencoder</code>'s ''-delay'' option solves this by delaying the audio
* '''failure to encode''': frames that desynchronise gradually over time, usually caused by audio and video shifting relative each other when frames are dropped. For example if your CPU is not fast enough and sometimes drops a video frame, after 25 dropped frames the video will be one second ahead of the audio
** <code>mencoder</code>'s ''-harddup'' option solves this by duplicating other frames to fill in the gaps
* '''source frame rate''': frames that aren't delivered at the advertised rate, usually caused by inaccurate clocks in the source hardware. For example, a low-cost webcam might deliver 25.01 video frames per second and 47,999kHz, causing your audio and video drifting apart by a second or so per hour
** video tapes are especially problematic here - if you've ever seen a VCR struggle during those few seconds between two recordings on a tape, you've seen them adjusting the tape speed to accurately track the source. Frame counts can vary enough during these periods to instantly desynchronise audio and video
** <code>mencoder</code> has no solution for this problem


* it provides partly-automated composition of GStreamer pipelines
GStreamer solves these problems by attaching a timestamp to each incoming frame based on the time GStreamer receives the frame. It can then mux the sources back together accurately using these timestamps, either by using a format that supports variable framerates or by duplicating frames to fill in the blanks:
* it allows you to cut streams, for example to capture for a predefined duration. That ensures headers are written correctly, which is not always the case if you close <code>gst-launch-1.0</code> by pressing Ctrl+C. To use this feature one has to insert a ''dam'' element after the first ''queue'' of each part of the pipeline
# If you choose a container format that supports timestamps (e.g. Matroska), timestamps are automatically written to the file and used to vary the playback speed
# If you choose a container format that does not support timestamps (e.g. AVI), you must duplicate other frames to fill in the gaps by adding the <code>videorate</code> and <code>audiorate</code> plugins to the end of the relevant pipelines


== Building pipelines ==
To get accurate timestamps, specify the <code>do-timestamp=true</code> option for all your sources. This will ensure accurate timestamps are retrieved from the driver where possible. Sadly, many v4l2 drivers don't support timestamps - GStreamer will add timestamps for these drivers to stop audio and video drifting apart, but you will need to fix the initialisation timing yourself (discussed below).


You will probably need to build your own GStreamer pipeline for your particular use case. This section contains examples to give you the basic idea.
== Common caputuring issues and their solutions ==


Note: for consistency and ease of copy/pasting, all filenames in this section are of the form <code>test-$( date --iso-8601=seconds )</code> - your shell should automatically convert this to e.g. <code>test-2010-11-12T13:14:15+1600.avi</code>
=== Determining your video source ===


See all your video sources by doing:
=== Record raw video only ===


A simple pipeline that initialises one video ''source'', sets the video format, ''muxes'' it into a file format, then saves it to a file:
ls /dev/video*


gst-launch-1.0 \
One of these is the card you want. Most people only have one, or can figure it out by disconnecting devices and rerunning the above command. Otherwise, check the capabilites of each device:
v4l2src device=$VIDEO_DEVICE \
! $VIDEO_CAPABILITIES \
! avimux \
! filesink location=test-$( date --iso-8601=seconds ).avi


This will create an AVI file with raw video and no audio. It should play in most software, but the file will be huge.
for VIDEO_DEVICE in /dev/video* ; do echo ; echo ; echo $VIDEO_DEVICE ; echo ; v4l2-ctl --device=$VIDEO_DEVICE --list-inputs ; done


=== Record raw audio only ===
Usually you will see e.g. a webcam with a single input and a TV card with multiple inputs. If you're still not sure which one is yours, try each one in turn:


A simple pipeline that initialises one audio ''source'', sets the audio format, ''muxes'' it into a file format, then saves it to a file:
v4l2-ctl --device=<device> --set-input=<whichever-input-you-want-to-use>
gst-launch-0.10 v4l2src do-timestamp=true device=<device> ! autovideosink


gst-launch-1.0 \
(if your source is a VCR, remember to play a video so you know the right one when you see it)
alsasrc device=$AUDIO_DEVICE \
! $AUDIO_CAPABILITIES \
! avimux \
! filesink location=test-$( date --iso-8601=seconds ).avi


This will create an AVI file with raw audio and no video.
If you like, you can store your device in an environment variable:


=== Record video and audio ===
VIDEO_DEVICE=<device>


gst-launch-1.0 \
All further examples will use <CODE>$VIDEO_DEVICE</CODE> in place of an actual video device
v4l2src device=$VIDEO_DEVICE \
! $VIDEO_CAPABILITIES \
! mux. \
alsasrc device=$AUDIO_DEVICE \
! $AUDIO_CAPABILITIES \
! mux. \
avimux name=mux \
! filesink location=test-$( date --iso-8601=seconds ).avi


Instead of a straightforward pipe with a single source leading into a muxer, this pipe has three parts:
=== Determining your audio source ===


# a video source leading to a named element (<code>! ''name''.</code> with a full stop means "pipe to the ''name'' element")
See all of our audio sources by doing:
# an audio source leading to the same element
# a named muxer element leading to a file sink


Muxers combine data from many inputs into a single output, allowing you to build quite flexible pipes.
arecord -l


=== Create multiple sinks ===
Again, it should be fairly obvious which of these is the right one. Get the device names by doing:


The <code>tee</code> element splits a single source into multiple outputs:
arecord -L | grep ^hw:


gst-launch-1.0 \
If you're not sure which one you want, try each in turn:
v4l2src device=$VIDEO_DEVICE \
! $VIDEO_CAPABILITIES \
! avimux \
! tee name=network \
! filesink location=test-$( date --iso-8601=seconds ).avi \
tcpclientsink host=127.0.0.1 port=5678


This sends your stream to a file (<code>filesink</code>) and out over the network (<code>tcpclientsink</code>). To make this work, you'll need another program listening on the specified port (e.g. <code>nc -l 127.0.0.1 -p 5678</code>).
gst-launch-0.10 alsasrc do-timestamp=true device=hw:<device> ! autoaudiosink


=== Encode audio and video ===
Again, you should hear your tape playing when you get the right one. Note: always use an ALSA ''hw'' device, as they are closest to the hardware. Pulse audio devices and ALSA's ''plughw'' devices add extra layers that, while more convenient for most uses, only cause headaches for us.


As well as piping streams around, GStreamer can manipulate their contents. The most common manipulation is to encode a stream:
Optionally set your device in an environment variable:


gst-launch-1.0 \
AUDIO_DEVICE=<device>
v4l2src device=$VIDEO_DEVICE \
! $VIDEO_CAPABILITIES \
! videoconvert \
! theoraenc \
! queue \
! mux. \
alsasrc device=$AUDIO_DEVICE \
! $AUDIO_CAPABILITIES \
! audioconvert \
! vorbisenc \
! mux. \
oggmux name=mux \
! filesink location=test-$( date --iso-8601=seconds ).ogg


The <code>theoraenc</code> and <code>vorbisenc</code> elements encode the video and audio using [https://en.wikipedia.org/wiki/Theora Ogg Theora] and [https://en.wikipedia.org/wiki/Vorbis Ogg Vorbis] encoders. The pipes are then muxed together into an [https://en.wikipedia.org/wiki/Ogg Ogg] container before being saved.
All further examples will use <CODE>$AUDIO_DEVICE</CODE> in place of an actual audio device


=== Getting your device capabilities ===
=== Add buffers ===


Different elements work at different speeds. For example, a CPU-intensive encoder might fall behind when another process uses too much processor time, or a duplicate frame detector might hold frames back while it examines them. This can cause streams to fall out of sync, or frames to be dropped altogether. You can add queues to smooth these problems out:
To find the capabilities of your audio device, do:
gst-launch-0.10 --gst-debug=alsa:5 alsasrc device=$AUDIO_DEVICE ! fakesink 2>&1 | grep 'returning caps'


gst-launch-1.0 -q -e \
To find the capabilities of your video source, do:
gst-launch-0.10 --gst-debug=v4l2src:3 v4l2src device=$VIDEO_DEVICE ! fakesink 2>&1 | grep 'probed caps:'
v4l2src device=$VIDEO_DEVICE \
! queue max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
! $VIDEO_CAPABILITIES \
! videoconvert \
! x264enc interlaced=true pass=quant quantizer=0 speed-preset=ultrafast byte-stream=true \
! progressreport update-freq=1 \
! mux. \
alsasrc device=$AUDIO_DEVICE \
! queue max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
! $AUDIO_CAPABILITIES \
! audioconvert \
! flacenc \
! mux. \
matroskamux name=mux min-index-interval=1000000000 \
! queue max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
! filesink location=test-$( date --iso-8601=seconds ).mkv


This creates a file using FLAC audio and x264 video in lossless mode, muxed into in a Matroska container. Because we used <code>speed-preset=ultrafast</code>, the buffers should just smooth out the flow of frames through the pipelines. Even though the buffers are set to the maximum possible size, <code>speed-preset=veryslow</code> would eventually fill the video buffer and start dropping frames.
These will display a list of capability ranges. You might find them more readable if you paste them into a text editor and replace all the semicolons with newlines. When you build a pipeline, you will need to specify capabilities based on these ranges. For example, given a video capability range <code>video/x-raw-yuv, format=(fourcc)I420, framerate=(fraction)25/1, width=(int)[ 48, 720 ], height=(int)[ 32, 578 ], interlaced=(boolean)true, pixel-aspect-ratio=(fraction)1/1</code>, you might choose a capability <code>video/x-raw-yuv, width=720, height=578</code> and allow GStreamer to choose sensible defaults for the other values.


Some other things to note about this pipeline:
=== Measuring your video framerate ===


* [https://trac.ffmpeg.org/wiki/Encode/H.264 FFmpeg's H.264 page] includes a useful discussion of speed presets (both programs use the same underlying library)
As mentioned above, some video cards produce slightly too many (or too few) frames per second. To check your system's actual frames per second, start your video source (e.g. a VCR or webcam) then run this command:
* <code>quantizer=0</code> sets the video codec to lossless mode (~30GB/hour). Anything up to <code>quantizer=18</code> should not lose information visible to the human eye, and will produce much smaller files (~10GB/hour)
* <code>min-index-interval=1000000000</code> improves seek times by telling the Matroska muxer to create one ''cue data'' entry per second of playback. Cue data is a few kilobytes per hour, added to the end of the file when encoding completes. If you try to watch your Matroska video while it's being recorded, it will take a long time to skip forward/back because the cue data hasn't been written yet


== Common caputuring issues and their solutions ==
gst-launch-0.10 v4l2src ! fpsdisplaysink fps-update-interval=100000

# Let it run for 100 seconds to get a large enough sample. It should print some statistics in the bottom of the window - write down the number of frames dropped
# Let it run for another 100 seconds, then write down the new number of frames dropped
# Calculate <code>(second number) - (first number) - 1</code> (e.g. 5007 - 2504 - 1 == 2502)
#* You need to subtract one because <code>fpsdisplaysink</code> drops one frame every time it displays the counter
# That number is exactly one hundred times your framerate, so you should tell GStreamer e.g. <code>framerate=2502/100</code>


=== Reducing Jerkiness ===
=== Reducing Jerkiness ===
Line 145: Line 206:
'''Check your CPU load'''. When GStreamer uses 100% CPU, it may need to drop frames to keep up.
'''Check your CPU load'''. When GStreamer uses 100% CPU, it may need to drop frames to keep up.
* If frames are dropped occasionally when CPU usage spikes to 100%, add a (larger) buffer to help smooth things out.
* If frames are dropped occasionally when CPU usage spikes to 100%, add a (larger) buffer to help smooth things out.
** this can be a source's internal buffer (e.g. ''v4l2src queue-size=16'' or ''alsasrc buffer-time=2000000''), or it can be an extra buffering step in your pipeline (''! queue max-size-buffers=0 max-size-time=0 max-size-bytes=0'')
** this can be a source's internal buffer (e.g. ''alsasrc buffer-time=2000000''), or it can be an extra buffering step in your pipeline (''! queue max-size-buffers=0 max-size-time=0 max-size-bytes=0'')
* If frames are dropped when other processes have high CPU load, consider using [https://en.wikipedia.org/wiki/Nice_(Unix) nice] to make sure encoding gets CPU priority
* If frames are dropped when other processes have high CPU load, consider using [https://en.wikipedia.org/wiki/Nice_(Unix) nice] to make sure encoding gets CPU priority
* If frames are dropped regularly, use a different codec, change the parameters, lower the resolution, or otherwise choose a less resource-intensive solution
* If frames are dropped regularly, use a different codec, change the parameters, lower the resolution, or otherwise choose a less resource-intensive solution
Line 154: Line 215:
* <code>videorate</code> on its own will actually make this problem worse by picking one frame and replacing all the others with it. Instead install <code>entrans</code> and add its ''stamp'' element between ''v4l2src'' and ''queue'' (e.g. ''v4l2src do-timestamp=true ! stamp sync-margin=2 sync-interval=5 ! videorate ! queue'')
* <code>videorate</code> on its own will actually make this problem worse by picking one frame and replacing all the others with it. Instead install <code>entrans</code> and add its ''stamp'' element between ''v4l2src'' and ''queue'' (e.g. ''v4l2src do-timestamp=true ! stamp sync-margin=2 sync-interval=5 ! videorate ! queue'')
** ''stamp'' intelligently guesses timestamps if drivers don't support timestamping. Its ''sync-'' options drop or copy frames to get a nearly-constant framerate. Using <code>videorate</code> as well does no harm and can solve some remaining problems
** ''stamp'' intelligently guesses timestamps if drivers don't support timestamping. Its ''sync-'' options drop or copy frames to get a nearly-constant framerate. Using <code>videorate</code> as well does no harm and can solve some remaining problems

=== Fixing initialisation timing errors ===

If your hardware doesn't support timestamps, your encoded audio and video might be desynchronised by a fixed amount throughout the video. This offset is based on too many factors to isolate (e.g. a new driver version might increase or decrease the value), so fixing this is a manual process that probably needs to be done every time you encode a file.

'''Calculate your desired offset:'''

# Record a video using one of the techniques below
# Open the video in your favourite video player
# Adjust the A/V sync until it looks right to you - different players put this in different places, for example it's ''Tools > Track Synchronisation'' in VLC
# write down your desired offset

If possible, look for (or create) [https://en.wikipedia.org/wiki/Clapperboard clapperboard]-like events - moments where an obvious visual element occurred at the same moment as an obvious audio moment. A hand clapping or a cup being placed on a table are good examples.

Extract your audio:

gst-launch-0.10 \
uridecodebin uri="file:///path/to/my.file" \
! progressreport \
! audioconvert \
! audiorate \
! wavenc \
! filesink location="/path/to/my.file.wav"

If you have a clapperboard event, you might want to examine the extracted file in an audio editor like [http://audacityteam.org/ Audacity]. You should be able to see the exact time of the clap sound in the audio stream, watch the video to isolate the exact frame, and use that information to calculate the precise audio delay.

Use <code>sox</code> to prepend some silence:

sox -S -t wav <( sox -V1 -n -r <bitrate> -c <audio-channels> -t wav - trim 0.0 <delay-in-seconds> ) "/path/to/my.file.wav" "/path/to/my.file.flac"

Mix the new audio and the old video into a new file:

gst-launch-0.10 \
uridecodebin uri="file:///path/to/my.file" \
! video/your-video-settings \
! mux. \
uridecodebin uri="file:///path/to/my.file.flac" \
! audioconvert \
! audiorate \
! your_preferred_audio_encoder \
! mux. \
avimux name=mux \
! filesink location="/path/to/my.file.new"

Note: you can apply any <code>sox</code> filter this way, like normalising the volume or removing background noise.

==== A specific solution for measuring your offset ====

Measuring your offset will probably be the most unique part of your recording solution. Here is one solution you could use when digitising old VHS tapes:

# Connect a camcorder to your VCR
# Tune the VCR so it shows the camcorder output when it's not playing
# Start your GStreamer pipeline
# Clap your hands in front of the camcorder so you can later measure A/V synchronisation
# Press play on the VCR
# When the video has finished recording, split the audio and video tracks as described above
# Examine the audio with [http://audacityteam.org/ Audacity] and identify the precise time of the clap sound
# Examine the video with [http://avidemux.sourceforge.net/ avidemux] and identify the frame of the clap image

You'll probably need to change every step of the above to match your situation, but hopefully it will provide some inspiration.


=== Avoiding pitfalls with video noise ===
=== Avoiding pitfalls with video noise ===
Line 219: Line 220:
If your video contains periods of [https://en.wikipedia.org/wiki/Noise_(video) video noise] (snow), you may need to deal with some extra issues:
If your video contains periods of [https://en.wikipedia.org/wiki/Noise_(video) video noise] (snow), you may need to deal with some extra issues:


* Most video capturing devices send EndOfStream signals if the input signal quality drops too low, causing GStreamer to finish capturing. To prevent the device from sending EOS, set ''num-buffers=-1'' on the ''v4l2src'' element.
* Most devices send an EndOfStream signal if the input signal quality drops too low, causing GStreamer to finish capturing. To prevent the device from sending EOS, set ''num-buffers=-1'' on the ''v4l2src'' element.
* The ''stamp'' plugin gets confused by periods of snow, causing it to generate faulty timestamps and framedropping. ''stamp'' will recover normal behaviour when the break is over, but will probably leave the buffer full of weirdly-stamped frames. ''stamp'' only drops one weirdly-stamped frame each sync-interval, so it can take several minutes until everything works fine again. To solve this problem, set ''leaky=2'' on each ''queue'' element to allow dropping old frames
* The ''stamp'' plugin gets confused by periods of snow, causing it to generate faulty timestamps and framedropping. ''stamp'' will recover normal behaviour when the break is over, but will probably leave the buffer full of weirdly-stamped frames. ''stamp'' only drops one weirdly-stamped frame each sync-interval, so it can take several minutes until everything works fine again. To solve this problem, set ''leaky=2'' on each ''queue'' element to allow dropping old frames
* Periods of noise (snow, bad signal etc.) are hard to encode. Variable bitrate encoders will often drive up the bitrate during the noise then down afterwards to maintain the average bitrate. To minimise these issues, specify a minimum and maximum bitrate in your encoder
* Periods of noise (snow, bad signal etc.) are hard to encode. Variable bitrate encoders will often drive up the bitrate during the noise then down afterwards to maintain the average bitrate. To minimise the issues, specify a minimum and maximum bitrate in your encoder
* Snow at the start of a recording is just plain ugly. To get black input instead, use the remote control to change the input source before you start recording
* Snow at the start of a recording is just plain ugly. To get black input instead from a VCR, use the remote control to change the input source before you start recording


=== Choosing formats ===
=== Investigating bugs in GStreamer ===


GStreamer comes with a extensive tracing system that let you figure-out the problems. Yet, you often need to understand the internals of GStreamer to be able to read those traces. You should read this [https://gstreamer.freedesktop.org/data/doc/gstreamer/head/gstreamer/html/gst-running.html documentation page] for the basic of how the tracing system works. When something goes wrong you should:
When you create a video, you need to choose your ''audio format'' (e.g. WAV or MP3), ''video format'' (e.g. XviD or MPEG-4) and ''container format'' (e.g. AVI or MP4). There's constant work to improve the ''codecs'' that create audio/video and ''muxers'' that create containers, and whole new formats are invented fairly regularly, so this page can't recommend any specific formats.
Wikipedia's comparisons of [https://en.wikipedia.org/wiki/Comparison_of_audio_coding_formats audio], [https://en.wikipedia.org/wiki/Comparison_of_video_codecs video] and [https://en.wikipedia.org/wiki/Comparison_of_container_formats container] formats are a good place to start your research - here are some important things to look for:


# try and see if there is a useful error message by enabling the ERROR debug level, <code>GST_DEBUG=2 gst-launch-1.0</code>
* '''encoding speed''' - encoders that generate too much CPU load will cause GStreamer to drop frames
# try similar pipelines - reducing to its most minimal form, and add more elements until you can reproduce the issue.
* '''accuracy''' - some formats are ''lossless'', others throw away information to improve speed and/or reduce file size
# as you are most likely having issue with V4L2 element, you may enable full v4l2 traces using <code>GST_DEBUG="v4l2*:7,2" gst-launch-1.0</code>.
* '''file size''' - different formats use different amounts of disk space, even with the same accuracy
# find an error message that looks relevant, search the Internet for information about it
* '''compatibility''' - newer formats usually produce better results but can't be played by older software
# try more variations based on what you learnt, until you eventually find something that works

# ask on Freenode #gstreamer or through [mailto:gstreamer-devel@lists.freedesktop.org GStreamer Mailing List]
Speed and accuracy are usually the most important when encoding, but size and compatibility the most important for playback. So it can make sense to encode with modern, fast, lossless formats then ''transcode'' to a format that produces a smaller or more compatible file. For example, as of 2015 it might make sense to encode FLAC audio and x264 video into a Matroska file, then transcode MP3 audio and MPEG-4 video into an AVI file. The transcoded file might be larger or lower quality, but it should play on most software, and you can just delete it and try again if your grandmother's DVD player doesn't like it.
# if you think you found a bug, you should report it through [https://bugzilla.gnome.org/enter_bug.cgi?product=GStreamer Gnome Bugzilla]

If you use separate ''encode'' and ''transcode'' pipelines, make sure to do any video processing in the transcode pipeline. Reducing noise, masking out overscan areas etc. usually increases CPU usage and sometimes introduces errors, which can only be fixed by re-recording if they're present in the original.

=== Cleaning audio ===

Any analogue recording will contain a certain amount of background noise. Cleaning noise is optional, and you'll always be able to produce a slightly better result if you spend a little longer on it. GStreamer doesn't have built-in noise reduction, so this section will just introduce enough theory to get you started. [https://wiki.audacityteam.org/wiki/Noise_Reduction Audacity's noise reduction effect] is a good place to start experimenting.

The major noise sources are:

* '''your audio codec''' might throw away sound it thinks you won't hear in order to reduce file size
* '''your recording system''' will produce a small, consistent amount of noise based on its various electrical and mechanical components
* '''VHS format limitations''' cause static at high and low frequencies, depending on the VCR's settings
* '''imperfections in tape recording and playback''' produce noise that differs between recordings and even between scenes

Always use a lossless audio format while cleaning (e.g. WAV or FLAC). Even if you plan to eventually use a format like MP3 that throws information away, the extra noise makes it harder to isolate the noise sources you're actually trying to measure.

The primary means of reducing noise is the frequency-based [http://en.wikipedia.org/wiki/Noise_gate noise gate], which ''blocks'' some frequencies and ''passes'' others. ''High-pass'' and ''low-pass'' filters pass noise above or below a certain frequency, and can be combined into ''band-pass'' or even ''multi-band'' filters. The rest of this section discusses how to build a series of noise gates for your audio.

Identify noise from your recording system by recording the sound of a paused tape or silent television channel for a few seconds:

gst-launch-0.10 alsasrc device=$AUDIO_DEVICE ! wavenc ! filesink location=baseline.wav

You can use this baseline recording as a ''noise profile'', which your software uses to build a multi-band noise gate. You can then apply that noise gate to any future recordings with the same hardware.

Identify VHS format limitations by searching online for information based on your TV norm (NTSC, PAL or SECAM), your recording quality (normal or Hi-Fi) and your VHS play mode (short- or long-play). [https://en.wikipedia.org/wiki/VHS#Audio_recording Wikipedia's discussion of VHS audio recording] is a good place to start. If you're able to find the information, gate your recordings with high-pass and low-pass filters that only allow frequencies within the range your tape actually records. For example, a long-play recording of a PAL tape will produce static below 100Hz and above 4kHz so you should gate your recording with a high-pass filter at 100Hz and a low-pass filter at 4000Hz. If you can't find the information, you can determine it experimentally by trying out different filters to see what sounds right - your system probably produces static below about 10Hz or 100Hz and above about 4kHz or 12kHz, so try high- and low-pass filters in those ranges. If you don't remove this noise source, the next step will do a reasonable job of guessing it for you anyway.

Identify imperfections in recording and playback by watching the video and looking for periods of silence. You only need half a second of background noise to generate a profile, but the number of profiles is up to you. Some people grab one profile for a whole recording, others combine clips into averaged noise profiles, others cut audio into scenes and de-noise each in turn. At a minimum, tapes with multiple recordings should be split up and each one de-noised separately - a tape containing a TV program recorded in LP mode in one VCR followed by a home video recorded in SP in another VCR will produce two very different noise profiles, even if played back all in one go.

It's good to apply filters in the right order (system profile, then VHS limits, then recording profiles), but beyond that noise reduction is very subjective. You can run your audio through as many gates as you like, and even repeat the same filter several times. If you use a noise reduction profile, you can even get different results from different programs (see for example [http://sourceforge.net/p/sox/mailman/message/30019023/ this comparison of sox and Audacity's algorithms]). You'll always be able to get a better result if you spend more time on the problem, so you'll need to decide for yourself when the result is good enough.


== Sample pipelines ==
== Sample pipelines ==


=== record from a bad analog signal to MJPEG video and RAW mono audio ===
At some point, you will probably need to build your own GStreamer pipeline. Here are some examples to give you the basic idea:

=== Record raw video only ===

A simple pipeline that initialises one video ''source'', sets the video format, ''muxes'' it into a file format, then saves it to a file:

gst-launch-0.10 \
v4l2src do-timestamp=true device=$VIDEO_DEVICE \
! video/x-raw-yuv,width=640,height=480 \
! avimux
! filesink location=test0.avi

<code>tcprobe</code> says this video-only file uses the I420 codec and gives the framerate as correct NTSC:

$ tcprobe -i test1.avi
[tcprobe] RIFF data, AVI video
[avilib] V: 29.970 fps, codec=I420, frames=315, width=640, height=480
[tcprobe] summary for test1.avi, (*) = not default, 0 = not detected
import frame size: -g 640x480 [720x576] (*)
frame rate: -f 29.970 [25.000] frc=4 (*)
no audio track: use "null" import module for audio
length: 315 frames, frame_time=33 msec, duration=0:00:10.510
The files will play in mplayer, using the codec [raw] RAW Uncompressed Video.

=== Record to ogg theora ===

Here is a more complex example that initialises two sources - one video and audio:

gst-launch-0.10 \
v4l2src do-timestamp=true device=$VIDEO_DEVICE \
! video/x-raw-yuv,width=640,height=480,framerate=\(fraction\)30000/1001 \
! ffmpegcolorspace \
! theoraenc \
! queue \
! mux. \
alsasrc do-timestamp=true device=$AUDIO_DEVICE \
! audio/x-raw-int,channels=2,rate=32000,depth=16 \
! audioconvert \
! vorbisenc \
! mux. \
oggmux name=mux \
! filesink location=test0.ogg

Each source is encoded and piped into a <i>muxer</i> that builds an ogg-formatted data stream. The stream is then saved to <code>test0.ogg</code>. Note the required workaround to get sound on a saa7134 card, which is set at 32000Hz (cf. [http://pecisk.blogspot.com/2006/04/alsa-worries-countinues.html bug]). However, I was still unable to get sound output, though mplayer claimed there was sound -- the video is good quality:

VIDEO: [theo] 640x480 24bpp 29.970 fps 0.0 kbps ( 0.0 kbyte/s)
Selected video codec: [theora] vfm: theora (Theora (free, reworked VP3))
AUDIO: 32000 Hz, 2 ch, s16le, 112.0 kbit/10.94% (ratio: 14000->128000)
Selected audio codec: [ffvorbis] afm: ffmpeg (FFmpeg Vorbis decoder)

=== Record to mpeg4 ===

This is similar to the above, but generates an AVI file with streams encoded using AVI-compatible encoders:

gst-launch-0.10 \
v4l2src do-timestamp=true device=$VIDEO_DEVICE \
! video/x-raw-yuv,width=640,height=480,framerate=\(fraction\)30000/1001 \
! ffmpegcolorspace \
! ffenc_mpeg4 \
! queue \
! mux. \
alsasrc do-timestamp=true device=$AUDIO_DEVICE \
! audio/x-raw-int,channels=2,rate=32000,depth=16 \
! audioconvert \
! lame \
! mux. \
avimux name=mux \
! filesink location=test0.avi

I get a file out of this that plays in mplayer, with blocky video and no sound. Avidemux cannot open the file.

=== GStreamer 1.0: record from a bad analog signal to MJPEG video and RAW mono audio ===

''stamp'' is not available in GStreamer 1.0, ''cogcolorspace'' and ''ffmpegcolorspace'' have been replaced by ''videoconvert'':


gst-launch-1.0 \
gst-launch-1.0 \
v4l2src do-timestamp=true device=$VIDEO_DEVICE do-timestamp=true \
v4l2src device=$VIDEO_DEVICE do-timestamp=true \
! $VIDEO_CAPABILITIES \
! 'video/x-raw,format=(string)YV12,width=(int)720,height=(int)576' \
! videorate \
! videorate \
! $VIDEO_CAPABILITIES \
! 'video/x-raw,format=(string)YV12,framerate=25/1' \
! videoconvert \
! videoconvert \
! $VIDEO_CAPABILITIES \
! 'video/x-raw,format=(string)YV12,width=(int)720,height=(int)576'
! jpegenc \
! jpegenc \
! queue \
! queue \
! mux. \
! mux. \
alsasrc do-timestamp=true device=$AUDIO_DEVICE \
alsasrc device=$AUDIO_DEVICE \
! $AUDIO_CAPABILITIES \
! 'audio/x-raw,format=(string)S16LE,rate=(int)48000,channels=(int)2' \
! audiorate \
! audiorate \
! audioresample \
! audioresample \
! 'audio/x-raw,rate=(int)44100' \
! $AUDIO_CAPABILITIES, rate=44100 \
! audioconvert \
! audioconvert \
! 'audio/x-raw,channels=(int)1' \
! $AUDIO_CAPABILITIES, rate=44100, channels=1 \
! queue \
! queue \
! mux. \
! mux. \
avimux name=mux ! filesink location=test.avi
avimux name=mux ! filesink location=test-$( date --iso-8601=seconds ).avi


The chip that captures audio and video might not deliver the exact framerates specified, which the AVI format can't handle. The <code>audiorate</code> and <code>videorate</code> elements remove or duplicate frames to maintain a constant rate.
As stated above, it is best to use both audiorate and videorate: you problably use the same chip to capture both audio stream and video stream so the audio part is subject to disturbance as well.


=== View pictures from a webcam ===
=== View pictures from a webcam (GStreamer 0.10) ===

Here are some miscellaneous examples for viewing webcam video:


gst-launch-0.10 \
gst-launch-0.10 \
v4l2src do-timestamp=true use-fixed-fps=false \
v4l2src do-timestamp=true device=$VIDEO_DEVICE \
! video/x-raw-yuv,format=\(fourcc\)UYVY,width=320,height=240 \
! video/x-raw-yuv,format=\(fourcc\)UYVY,width=320,height=240 \
! ffmpegcolorspace \
! ffmpegcolorspace \
! autovideosink
! autovideosink


In GStreamer 0.10, ''videoconvert'' was called ''ffmpegcolorspace''.
gst-launch-0.10 \
v4lsrc do-timestamp=true autoprobe-fps=false device=$VIDEO_DEVICE \
! "video/x-raw-yuv,format=(fourcc)I420,width=160,height=120,framerate=10" \
! autovideosink


=== Entrans: Record to DVD-compliant MPEG2 ===
=== Entrans: Record to DVD-compliant MPEG2 (GStreamer 0.10) ===


entrans -s cut-time -c 0-180 -v -x '.*caps' --dam -- --raw \
entrans -s cut-time -c 0-180 -v -x '.*caps' --dam -- --raw \
v4l2src queue-size=16 do-timestamp=true device=$VIDEO_DEVICE norm=PAL-BG num-buffers=-1 ! stamp silent=false progress=0 sync-margin=2 sync-interval=5 ! \
v4l2src queue-size=16 do-timestamp=true device=$VIDEO_DEVICE norm=PAL-BG num-buffers=-1 \
queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! dam ! \
! stamp silent=false progress=0 sync-margin=2 sync-interval=5 \
cogcolorspace ! videorate silent=false ! \
! queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
! dam \
'video/x-raw-yuv,width=720,height=576,framerate=25/1,interlaced=true,aspect-ratio=4/3' ! \
! cogcolorspace \
queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! \
! videorate silent=false \
ffenc_mpeg2video rc-buffer-size=1500000 rc-max-rate=7000000 rc-min-rate=3500000 bitrate=4000000 max-key-interval=15 pass=pass1 ! \
! 'video/x-raw-yuv,width=720,height=576,framerate=25/1,interlaced=true,aspect-ratio=4/3' \
queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! mux. \
! queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
pulsesrc buffer-time=2000000 do-timestamp=true ! \
! ffenc_mpeg2video rc-buffer-size=1500000 rc-max-rate=7000000 rc-min-rate=3500000 bitrate=4000000 max-key-interval=15 pass=pass1 \
queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! dam ! \
audioconvert ! audiorate silent=false ! \
! queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
! mux. \
audio/x-raw-int,rate=48000,channels=2,depth=16 ! \
pulsesrc buffer-time=2000000 do-timestamp=true \
queue silent=false max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! \
! queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
ffenc_mp2 bitrate=192000 ! \
! dam \
queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! mux. \
! audioconvert \
ffmux_mpeg name=mux ! filesink location=my_recording.mpg
! audiorate silent=false \
! audio/x-raw-int,rate=48000,channels=2,depth=16 \
! queue silent=false max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
! ffenc_mp2 bitrate=192000 \
! queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
! mux. \
ffmux_mpeg name=mux \
! filesink location=test-$( date --iso-8601=seconds ).mpg


This captures 3 minutes (180 seconds, see first line of the command) to ''my_recording.mpg'' and even works for bad input signals.
This captures 3 minutes (180 seconds, see first line of the command) to ''test-$( date --iso-8601=seconds ).mpg'' and even works for bad input signals.
* I wasn't able to figure out how to produce a mpeg with ac3-sound as neither ffmux_mpeg nor mpegpsmux support ac3 streams at the moment. mplex does but I wasn't able to get it working as one needs very big buffers to prevent the pipeline from stalling and at least my GStreamer build didn't allow for such big buffers.
* I wasn't able to figure out how to produce a mpeg with ac3-sound as neither ffmux_mpeg nor mpegpsmux support ac3 streams at the moment. mplex does but I wasn't able to get it working as one needs very big buffers to prevent the pipeline from stalling and at least my GStreamer build didn't allow for such big buffers.
* The limited buffer size on my system is again the reason why I had to add a third queue element to the middle of the audio as well as of the video part of the pipeline to prevent jerking.
* The limited buffer size on my system is again the reason why I had to add a third queue element to the middle of the audio as well as of the video part of the pipeline to prevent jerking.
Line 406: Line 307:
* It seems to be important that the ''video/x-raw-yuv,width=720,height=576,framerate=25/1,interlaced=true,aspect-ratio=4/3''-statement is after ''videorate'' as videorate seems to drop the aspect-ratio-metadata otherwise resulting in files with aspect-ratio 1 in theis headers. Those files are probably played back warped and programs like dvdauthor complain.
* It seems to be important that the ''video/x-raw-yuv,width=720,height=576,framerate=25/1,interlaced=true,aspect-ratio=4/3''-statement is after ''videorate'' as videorate seems to drop the aspect-ratio-metadata otherwise resulting in files with aspect-ratio 1 in theis headers. Those files are probably played back warped and programs like dvdauthor complain.


=== Bash script to record video tapes with entrans ===
== Ready-made scripts ==


For most use cases, you'll want to wrap GStreamer in a larger shell script. This script protects against several common mistakes during encoding.
Although no two use cases are the same, it can be useful to see scripts used by other people. These can fill in blanks and provide inspiration for your own work.


See also [[V4L_capturing/script|the V4L capturing script]] for a a wrapper that represents a whole workflow.
=== Bash script to record video tapes with GStreamer (work-in-progress) ===


Note: as of August 2015, this script is still being fine-tuned. Come back in a month or two to see the final version.

<nowiki>#!/bin/bash
#
# GStreamer lets you build a pipeline (a DAG of elements) to process audio and video.
#
# At the time of writing, both the 0.1 and 1.0 series were installed by default.
# So far as I can tell, the 1.0 series has some kind of bug that breaks TV-recording utterly
# (possibly a bug in selecting the output formats from v4l2src)
# If the 1.0 series gets fixed, you should only need to change a few commands here and there
# (see http://gstreamer.freedesktop.org/data/doc/gstreamer/head/manual/html/chapter-porting-1.0.html)
#
# We also use `v4l2-ctl` from the v4l-utils package to set the input source,
# and `sox` (from the `sox` package) to edit the audio
#
# Approximate system requirements for maximum quality settings (smaller images and lower bitrates need less):
# * about 10GB free for every hour of recording (6-7GB for temporary files, 3-4GB for the output file)
# * 3GHz processor (preferably at least dual core, so other processes don't steal the encoder's CPU time)
# * about 2GB of memory for every hour of recording (the second encoding pass needs to see the whole file)

HELP_MESSAGE="Usage: $0 --init
$0 --record <directory>
$0 --kill <directory> <timeout>
$0 --process <directory>
$0 --clean <directory>

Record a video into a directory (one directory per video).

--init create an initial ~/.gstreamer-record-scriptrc
please edit this file before your first recording

--record create a first-pass recording in the specified directory

--kill stop the recording in the specified directory after a specific amount time
see \`man sleep\` for details about allowed time formats

--process build the final recording in the specified directory
make sure to edit \`.gstreamer-record-scriptrc\` in that directory first

--clean delete temporary files
"

CONFIGURATION='#
# CONFIGURATION FOR GSTREAMER RECORD SCRIPT
# For more information, see http://www.linuxtv.org/wiki/index.php/GStreamer
#

#
# VARIABLES YOU NEED TO EDIT
# Every system and every use case is slightly different.
# Here are the things you will probably need to change:
#

# Set these based on your hardware/location:
VIDEO_DEVICE=${VIDEO_DEVICE:-/dev/video0} # `ls /dev/video*` for a list
AUDIO_DEVICE=${AUDIO_DEVICE:-hw:CARD=SAA7134,DEV=0} # `arecord -L` for a list
NORM=${NORM:-PAL} # (search Wikipedia for the exact norm in your country)
VIDEO_KBITRATE="${VIDEO_KBITRATE:-8000}" # test for yourself, but 8000 seems to produce a high quality result (we use bitrate/1000 for readability and to help calculate the values below)
AUDIO_BITRATE="${AUDIO_BITRATE:-32000}" # only bitrate supported by SAA7134 drivers - do `arecord -D $AUDIO_DEVICE --dump-hw-params -d 1 /dev/null` to see what your device supports
VIDEO_INPUT="${VIDEO_INPUT:-1}" # composite input - `v4l2-ctl --device=$VIDEO_DEVICE --list-inputs` for a list

# PAL video is approximately 720x576 resolution. VHS tapes have about half the horizontal quality, but this post convinced me to encode at 720x576 anyway:
# http://forum.videohelp.com/threads/215570-Sensible-resolution-for-VHS-captures?p=1244415#post1244415
ASPECT_W="${ASPECT_W:-5}"
ASPECT_H="${ASPECT_H:-4}"
SIZE_MULTIPLIER="${SIZE_MULTIPLIER:-144}" # common multipliers include 144 (720x576 - PAL), 128 (640x480 - VGA) and 72 (360x288 - half PAL). Set this lower to reduce CPU usage

# GStreamer automatically keeps audio and video in sync, but most systems start recording audio shortly before video video.
# If your system has this problem...
#
# 1. run the first pass of AVI recording
# 2. watch the video in your favourite video player
# 3. adjust the audio delay until the video looks right
# 4. pass the relevant number to the second pass
# 5. if you plan to do several recordings in one session, you can set the following default value
#
# Note: you will have an opportunity to set the audio delay for a specific file later
AUDIO_DELAY="${AUDIO_DELAY:-0.3}"

# Some VCRs consistently run slightly fast or slow. If you suspect your VCR has this problem...
#
# Do a quick test:
# 1. Run this command: gst-launch-0.10 v4l2src ! fpsdisplaysink fps-update-interval=1000
# * this will measure your average frame rate every second. After a few seconds, it should say "drop rate 25.00"
# 2. Change "FRAMERATE" below to your actual frame rate (e.g. 2502/100 if your frame rate is 25.02 FPS)
#
# Or if you want to be precise:
# 1. Run this command: gst-launch-0.10 v4l2src ! fpsdisplaysink fps-update-interval=100000
# * this will measure your average frame rate every 100 seconds (you can try different intervals if you like)
# 2. wait 100 seconds, then record the number of frames dropped
# 3. wait another 100 seconds, then record the number of frames dropped again
# 4. calculate (result of step 4) - (result of step 3) - 1
# * e.g. 5007 - 2504 - 1 == 2502
# * you need to subtract one because fpsdisplaysink drops one frame every time it displays the counter
# 5. Change "FRAMERATE" below to (result of step 4)/100 (e.g. 2502/100 if 2502 frames were dropped)
FRAMERATE="${FRAMERATE:-2500/100}"

#
# VARIABLES YOU MIGHT NEED TO EDIT
# These are defined in the script, but you can override them here if you need non-default values:
#

# set this to 1 to get lots of debugging data (including DOT graphs of your pipelines):
#DEBUG_MODE=

# Set these to alter the recording quality:
#GST_MPEG4_OPTS="..."
#GST_MPEG4_OPTS_PASS1="..."
#GST_MPEG4_OPTS_PASS2="..."
#GST_LAME_OPTS="..."

# Set these to control the audio/video pipelines:
#GST_QUEUE="..."
#GST_VIDEO_SRC="..."
#GST_AUDIO_SRC="..."
'

#
# CONFIGURATION SECTION
#

CONFIG_SCRIPT="$HOME/.gstreamer-record-scriptrc"
[ -e "$CONFIG_SCRIPT" ] && source "$CONFIG_SCRIPT"
source <( echo "$CONFIGURATION" )

# set this to 1 to get lots of debugging data (including DOT graphs of your pipelines):
DEBUG_MODE="${DEBUG_MODE:-}"

# `gst-inspect` has more information here too:
GST_MPEG4_OPTS="${GST_MPEG4_OPTS:-interlaced=true bitrate=$(( VIDEO_KBITRATE * 1000 )) max-key-interval=15}"
GST_MPEG4_OPTS_PASS1="${GST_MPEG4_OPTS_PASS1:-rc-buffer-size=$(( VIDEO_KBITRATE * 4000 )) rc-max-rate=$(( VIDEO_KBITRATE * 2000 )) rc-min-rate=$(( VIDEO_KBITRATE * 875 )) pass=pass1 $GST_MPEG4_OPTS}" # pictures of white noise will max out your bitrate - setting min/max bitrates ensures the video after a period of snow will be reasonable quality
GST_MPEG4_OPTS_PASS2="${GST_MPEG4_OPTS_PASS2:-pass=pass2 $GST_MPEG4_OPTS}" # pictures of white noise will max out your bitrate - setting min/max bitrates ensures the video after a period of snow will be reasonable quality
GST_LAME_OPTS="${GST_LAME_OPTS:-quality=0}"

# `gst-inspect-0.10 <element> | less -i` for a list of properties (e.g. `gst-inspect-0.10 v4l2src | less -i`):
GST_QUEUE="${GST_QUEUE:-queue max-size-buffers=0 max-size-time=0 max-size-bytes=0}"
GST_VIDEO_FORMAT="${GST_VIDEO_FORMAT:-video/x-raw-yuv,width=$(( ASPECT_W * SIZE_MULTIPLIER )),height=$(( ASPECT_H * SIZE_MULTIPLIER )),framerate=$FRAMERATE,interlaced=true,aspect-ratio=$ASPECT_W/$ASPECT_H}"
GST_AUDIO_FORMAT="${GST_AUDIO_FORMAT:-audio/x-raw-int,channels=2,rate=$AUDIO_BITRATE,depth=16}"
GST_VIDEO_SRC="${GST_VIDEO_SRC:-v4l2src device=$VIDEO_DEVICE do-timestamp=true norm=$NORM ! $GST_QUEUE ! videorate silent=false ! $GST_VIDEO_FORMAT}"
GST_AUDIO_SRC="${GST_AUDIO_SRC:-alsasrc device=$AUDIO_DEVICE do-timestamp=true ! $GST_QUEUE ! audioconvert ! audiorate silent=false ! $GST_AUDIO_FORMAT}"

#
# MAIN LOOP
# You should only need to edit this if you're making significant changes to the way the script works
#

echo_bold() {
echo -e "\e[1m$@\e[0m"
}

show_progress() {
START_TIME="$( date +%s )"
PROGRESS_NEWLINE=
while read HEAD TAIL
do
if [ "$HEAD" = "progressreport0" ]
then
NOW_TIME="$( date +%s )"
echo "$TAIL" | {
read TIME PROCESSED SLASH TOTAL REPLY
if [ "$PROCESSED" -eq 0 ]
then
echo -n $'\r'"$TAIL"
else
echo -n $'\r'"$TAIL ETA:" $( date -d "$(( ( $(date +%s) - $START_TIME ) * ( $TOTAL - $PROCESSED ) / $PROCESSED )) seconds" )
fi
}
PROGRESS_NEWLINE=$'\n'
else
echo "$PROGRESS_NEWLINE$HEAD $TAIL"
PROGRESS_NEWLINE=
fi
done
}

set_directory() {
if [ -z "$1" ]
then

echo "$HELP_MESSAGE"
exit 1

else

DIRECTORY="$( readlink -f "$1" )"
FILE="$DIRECTORY/gstreamer-recording"
URI="file://$( echo "$FILE" | sed -e 's/ /%20/g' )"
mkdir -p -- "$DIRECTORY" || exit

GST_CMD="gst-launch-0.10"
if [ -n "$DEBUG_MODE" ]
then
export GST_DEBUG_DUMP_DOT_DIR="$DIRECTORY/graphs"
if [ -d "$GST_DEBUG_DUMP_DOT_DIR" ]
then
rm -f "$GST_DEBUG_DUMP_DOT_DIR"/*.dot
else
mkdir "$GST_DEBUG_DUMP_DOT_DIR"
fi
GST_CMD="$GST_CMD -v --gst-debug=2"
fi

fi
}

case "$1" in

-i|--i|--in|--ini|--init)
if [ -e "$CONFIG_SCRIPT" ]
then
echo "Please delete $CONFIG_SCRIPT if you want to recreate it"
else
echo "$CONFIGURATION" > "$CONFIG_SCRIPT"
echo "Please edit $CONFIG_SCRIPT to match your system"
fi
;;

-r|--r|--re|--rec|--reco|--recor|--record)

# Build a pipeline with sources being encoded as MPEG4 video and FLAC audio, then being muxed into a Matroska container.
# FLAC and Matroska are used during encoding to ensure we don't lose much data between passes

set_directory "$2"

if [ -e "$FILE-temp.mkv" ]
then
echo "Please delete the old $FILE-temp.mkv before making a new recording"
exit 1
fi

v4l2-ctl --device="$VIDEO_DEVICE" --set-input $VIDEO_INPUT

echo_bold "Press ctrl+c to finish recording"
sudo nice -20 sh -c "echo \$\$ > '$FILE-temp.pid' && exec $GST_CMD -e \
$GST_VIDEO_SRC ! ffenc_mpeg4 $GST_MPEG4_OPTS_PASS1 'multipass-cache-file=$FILE-temp.log' ! $GST_QUEUE ! mux. \
$GST_AUDIO_SRC ! flacenc ! $GST_QUEUE ! mux. \
matroskamux name=mux ! filesink location='$FILE-temp.mkv'"

echo_bold extracting audio...
$GST_CMD -q \
uridecodebin uri="$URI-temp.mkv" \
! progressreport \
! audioconvert \
! audiorate \
! wavenc \
! filesink location="$FILE-temp.wav" \
| show_progress
echo

cat <<EOF > "$FILE.conf"
#
# NOISE REDUCTION (optional)
#
# To reduce noise in the final stream, identify a period in the recording
# which only has background noise (a second or two should be enough)
#
# If you want to reduce noise, set these two variables to the start and
# duration of the noise period (in seconds):
#
NOISE_START=
NOISE_DURATION=

#
# AUDIO DELAY
#
# To add a period of silence at the beginning of the video, watch the .mkv
# file and decide how much silence you want.
#
# If you want to add a delay, set this variable to the duration in seconds
# (can be fractional):
#
AUDIO_DELAY=$AUDIO_DELAY

# Uncomment the following when you've finished editing
# (it's just here to prevent silly mistakes)
# Then re-run the script with --process
#CONFIG_FILE_DONE=done
EOF

cat <<EOF

Now please check $FILE-temp.mkv
If you've recorded the wrong thing, delete it and start again
Otherwise edit $FILE.conf and set any variables you want
EOF
;;

-k|--k|--ki|--kil|--kill)
set_directory "$2"
sudo sh -c "sleep '$3' && kill -s 2 '$(< "$FILE-temp.pid" )'"
;;

-p|--p|--pr|--pro|--proc|--proce|--proces|--process)

set_directory "$2"

[ -e "$FILE.conf" ] && source "$FILE.conf"

if [ -z "$CONFIG_FILE_DONE" ]
then
echo "Please edit $FILE.conf before processing the file"
exit 1
fi

if [ -e "$FILE.avi" ]
then
echo "Please delete the old $FILE.avi before making a new recording"
exit 1
fi

# Add $AUDIO_DELAY seconds of silence to the audio, calculate a noise profile, and reduce noise based on that profile
if [ -n "$NOISE_START$AUDIO_DELAY" ]
then
if [ -n "$NOISE_START" ]
then
exec 6< <( sox -V1 "$FILE-temp.wav" -t wav - trim "$NOISE_START" "$NOISE_DURATION" | sox -t wav - -n noiseprof - )
NOISE_CMD="noisered /dev/fd/6 0.21"
else
NOISE_CMD=
fi
case "${AUDIO_DELAY:0:1}X" in
X)
# NO AUDIO DELAY
exec 3< <( sox -V1 "$FILE-temp.wav" -t wav - $NOISE_CMD )
;;
-)
# NEGATIVE AUDIO DELAY - APPEND SILENCE
exec 4< <( sox -V1 "$FILE-temp.wav" -t wav - trim 0.0 "$AUDIO_DELAY" )
exec 5< <( sox -V1 -n -r "$AUDIO_BITRATE" -c 2 -t wav - trim 0.0 "${AUDIO_DELAY:1}" )
exec 3< <( sox -V1 -t wav /dev/fd/4 -t wav /dev/fd/5 $NOISE_CMD )
;;
*)
# POSITIVE AUDIO DELAY - PREPEND SILENCE
exec 4< <( sox -V1 -n -r "$AUDIO_BITRATE" -c 2 -t wav - trim 0.0 "$AUDIO_DELAY" )
exec 5< <( sox -V1 "$FILE-temp.wav" -t wav - trim 0.0 -"$AUDIO_DELAY" )
exec 3< <( sox -V1 -t wav /dev/fd/4 -t wav /dev/fd/5 $NOISE_CMD )
;;
esac
else
exec 3< "$FILE-temp.wav"
fi

# Do a second pass over the video (shrinking the file size), and replace the audio with the improved version:
echo_bold "building final file..."
nice -n +20 $GST_CMD -q \
uridecodebin uri="$URI-temp.mkv" ! progressreport ! ffenc_mpeg4 $GST_MPEG4_OPTS_PASS2 "multipass-cache-file=$FILE-temp.log" ! $GST_QUEUE ! mux.video_0 \
fdsrc fd=3 ! decodebin ! audioconvert ! audiorate ! lamemp3enc $GST_LAME_OPTS ! $GST_QUEUE ! mux.audio_0 \
avimux name=mux ! filesink location="$FILE.avi" \
| show_progress

echo_bold "Saved to $FILE.avi"

;;

-c|--c|--cl|--clea|--clean)
rm -f "$2"-temp.*
;;

*)
echo "$HELP_MESSAGE"

esac
</nowiki>

This script generates a video in two passes: first it records and builds statistics, then lets you analyse the output, then builds an optimised final version.

=== Bash script to record video tapes with entrans ===
<nowiki>#!/bin/bash
<nowiki>#!/bin/bash
Line 895: Line 428:
* As setting of the inputs and settings of the capture device is only partly possible via GStreamer other tools are used.
* As setting of the inputs and settings of the capture device is only partly possible via GStreamer other tools are used.
* Adjust the settings to match your input sources, the recording volume, capturing saturation and so on.
* Adjust the settings to match your input sources, the recording volume, capturing saturation and so on.



==Further documentation resources==
==Further documentation resources==


* [[V4L_capturing|V4L Capturing]]
* [http://gstreamer.freedesktop.org/ Gstreamer project]
* [http://gstreamer.freedesktop.org/ Gstreamer project]
* [http://gstreamer.freedesktop.org/data/doc/gstreamer/head/faq/html/ FAQ]
* [http://gstreamer.freedesktop.org/data/doc/gstreamer/head/faq/html/ FAQ]

Latest revision as of 14:01, 11 March 2018

GStreamer is a toolkit for building audio- and video-processing pipelines. A pipeline might stream video from a file to a network, or add an echo to a recording, or (most interesting to us) capture the output of a Video4Linux device. Gstreamer is most often used to power graphical applications such as Totem, but can also be used directly from the command-line. This page will explain how GStreamer is better than the alternatives, and how to build an encoder using its command-line interface.

Before reading this page, see V4L capturing to set your system up and create an initial recording. This page assumes you have already implemented the simple pipeline described there.

Introduction to GStreamer

No two use cases for encoding are quite alike. What's your preferred workflow? Is your processor fast enough to encode high quality video in real-time? Do you have enough disk space to store the raw video then process it after the fact? Do you want to play your video in DVD players, or is it enough that it works in your version of VLC? How will you work around your system's obscure quirks?

Use GStreamer if you want the best video quality possible with your hardware, and don't mind spending a weekend browsing the Internet for information.

Avoid GStreamer if you just want something quick-and-dirty, or can't stand programs with bad documentation and unhelpful error messages.

Why is GStreamer better at encoding?

GStreamer isn't as easy to use as mplayer, and doesn't have as advanced editing functionality as ffmpeg. But it has superior support for synchronising audio and video in disturbed sources such as VHS tapes. If you specify your input is (say) 25 frames per second video and 48,000Hz audio, most tools will synchronise audio and video simply by writing 1 video frame, 1,920 audio frames, 1 video frame and so on. There are at least three ways this can cause errors:

  • initialisation timing: audio and video desynchronised by a certain amount from the first frame, usually caused by audio and video devices taking different amounts of time to initialise. For example, the first audio frame might be delivered to GStreamer 0.01 seconds after it was requested, but the first video frame might not be delivered until 0.7 seconds after it was requested, causing all video to be 0.6 seconds behind the audio
    • mencoder's -delay option solves this by delaying the audio
  • failure to encode: frames that desynchronise gradually over time, usually caused by audio and video shifting relative to each other when frames are dropped. For example if your CPU is not fast enough and sometimes drops a video frame, after 25 dropped frames the video will be one second ahead of the audio
    • mencoder's -harddup option solves this by duplicating other frames to fill in the gaps
  • source frame rate: frames that aren't delivered at the advertised rate, usually caused by inaccurate clocks in the source hardware. For example, a low-cost webcam that advertises 25 FPS video and 48kHz audio might actually deliver 25.01 video frames and 47,999 audio frames per second, causing your audio and video to drift apart by a second or so per hour
    • video tapes are especially problematic here - if you've ever seen a VCR struggle during those few seconds between two recordings on a tape, you've seen them adjusting the tape speed to accurately track the source. Frame counts can vary enough during these periods to instantly desynchronise audio and video
    • mencoder has no solution for this problem

GStreamer solves these problems by attaching a timestamp to each incoming frame based on the time GStreamer receives the frame. It can then mux the sources back together accurately using these timestamps, either by using a format that supports variable framerates or by duplicating frames to fill in the blanks:

  1. If you choose a container format that supports timestamps (e.g. Matroska), timestamps are automatically written to the file and used to vary the playback speed
  2. If you choose a container format that does not support timestamps (e.g. AVI), you must duplicate other frames to fill in the gaps by adding the videorate and audiorate plugins to the end of the relevant pipelines

Getting GStreamer

GStreamer, its most common plugins and tools are available through your distribution's package manager. Most Linux distributions include both the legacy 0.10 and modern 1.0 release series - each has bugs that stop them from working on some hardware, and this page focuses mostly on the modern 1.0 series. Converting between 0.10 and 1.0 is mostly just search-and-replace work (e.g. changing instances of av to ff because of the switch from ffmpeg to libavcodec). See the porting guide for more.

Other plugins are also available, such as GEntrans (used in some examples below). Google might help you find packages for your distribution, otherwise you'll need to download and compile them yourself.

Using GStreamer with gst-launch-1.0

gst-launch is the standard command-line interface to GStreamer. Here's the simplest pipline you can build:

gst-launch-1.0 fakesrc ! fakesink

This connects a single (fake) source to a single (fake) sink using the 1.0 series of GStreamer:

Very simple pipeline

GStreamer can build all kinds of pipelines, but you probably want to build one that looks something like this:

Idealised pipeline example

To get a list of elements that can go in a GStreamer pipeline, do:

gst-inspect-1.0 | less

Pass an element name to gst-inspect-1.0 for detailed information. For example:

gst-inspect-1.0 fakesrc
gst-inspect-1.0 fakesink

The images above are based on graphs created by GStreamer itself. Install Graphviz to build graphs of your pipelines. For faster viewing of those graphs, you may install xdot from [1]:

mkdir gst-visualisations
GST_DEBUG_DUMP_DOT_DIR=gst-visualisations gst-launch-1.0 fakesrc ! fakesink
xdot gst-visualisations/*-gst-launch.*_READY.dot

You may also compiles those graph to PNG, SVG or other supported formats:

 dot -Tpng gst-visualisations/*-gst-launch.*_READY.dot > my-pipeline.png

To get graphs of the example pipelines below, prepend GST_DEBUG_DUMP_DOT_DIR=gst-visualisations to the gst-launch-1.0 command. Run this command to generate a graph of GStreamer's most interesting stage:

xdot gst-visualisations/*-gst-launch.PLAYING_READY.dot

Remember to empty the gst-visualisations directory between runs.

Using GStreamer with entrans

gst-launch-1.0 is the main command-line interface to GStreamer, available by default. But entrans is a bit smarter:

  • it provides partly-automated composition of GStreamer pipelines
  • it allows you to cut streams, for example to capture for a predefined duration. That ensures headers are written correctly, which is not always the case if you close gst-launch-1.0 by pressing Ctrl+C. To use this feature one has to insert a dam element after the first queue of each part of the pipeline

Building pipelines

You will probably need to build your own GStreamer pipeline for your particular use case. This section contains examples to give you the basic idea.

Note: for consistency and ease of copy/pasting, all filenames in this section are of the form test-$( date --iso-8601=seconds ) - your shell should automatically convert this to e.g. test-2010-11-12T13:14:15+1600.avi

Record raw video only

A simple pipeline that initialises one video source, sets the video format, muxes it into a file format, then saves it to a file:

gst-launch-1.0 \
    v4l2src device=$VIDEO_DEVICE \
        ! $VIDEO_CAPABILITIES \
        ! avimux \
        ! filesink location=test-$( date --iso-8601=seconds ).avi

This will create an AVI file with raw video and no audio. It should play in most software, but the file will be huge.

Record raw audio only

A simple pipeline that initialises one audio source, sets the audio format, muxes it into a file format, then saves it to a file:

gst-launch-1.0 \
    alsasrc device=$AUDIO_DEVICE \
        ! $AUDIO_CAPABILITIES \
        ! avimux \
        ! filesink location=test-$( date --iso-8601=seconds ).avi

This will create an AVI file with raw audio and no video.

Record video and audio

gst-launch-1.0 \
    v4l2src device=$VIDEO_DEVICE \
        ! $VIDEO_CAPABILITIES \
        ! mux. \
    alsasrc device=$AUDIO_DEVICE \
        ! $AUDIO_CAPABILITIES \
        ! mux. \
    avimux name=mux \
        ! filesink location=test-$( date --iso-8601=seconds ).avi

Instead of a straightforward pipe with a single source leading into a muxer, this pipe has three parts:

  1. a video source leading to a named element (! name. with a full stop means "pipe to the name element")
  2. an audio source leading to the same element
  3. a named muxer element leading to a file sink

Muxers combine data from many inputs into a single output, allowing you to build quite flexible pipes.

Create multiple sinks

The tee element splits a single source into multiple outputs:

gst-launch-1.0 \
    v4l2src device=$VIDEO_DEVICE \
        ! $VIDEO_CAPABILITIES \
        ! avimux \
        ! tee name=network \
        ! filesink location=test-$( date --iso-8601=seconds ).avi \
    tcpclientsink host=127.0.0.1 port=5678 

This sends your stream to a file (filesink) and out over the network (tcpclientsink). To make this work, you'll need another program listening on the specified port (e.g. nc -l 127.0.0.1 -p 5678).

Encode audio and video

As well as piping streams around, GStreamer can manipulate their contents. The most common manipulation is to encode a stream:

gst-launch-1.0 \
    v4l2src device=$VIDEO_DEVICE \
        ! $VIDEO_CAPABILITIES \
        ! videoconvert \
        ! theoraenc \
        ! queue \
        ! mux. \
    alsasrc device=$AUDIO_DEVICE \
        ! $AUDIO_CAPABILITIES \
        ! audioconvert \
        ! vorbisenc \
        ! mux. \
    oggmux name=mux \
        ! filesink location=test-$( date --iso-8601=seconds ).ogg

The theoraenc and vorbisenc elements encode the video and audio using Ogg Theora and Ogg Vorbis encoders. The pipes are then muxed together into an Ogg container before being saved.

Add buffers

Different elements work at different speeds. For example, a CPU-intensive encoder might fall behind when another process uses too much processor time, or a duplicate frame detector might hold frames back while it examines them. This can cause streams to fall out of sync, or frames to be dropped altogether. You can add queues to smooth these problems out:

gst-launch-1.0 -q -e \
    v4l2src device=$VIDEO_DEVICE \
        ! queue max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
        ! $VIDEO_CAPABILITIES \
        ! videoconvert \
        ! x264enc interlaced=true pass=quant quantizer=0 speed-preset=ultrafast byte-stream=true \
        ! progressreport update-freq=1 \
        ! mux. \
    alsasrc device=$AUDIO_DEVICE \
        ! queue max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
        ! $AUDIO_CAPABILITIES \
        ! audioconvert \
        ! flacenc \
        ! mux. \
    matroskamux name=mux min-index-interval=1000000000 \
        ! queue max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
        ! filesink location=test-$( date --iso-8601=seconds ).mkv

This creates a file using FLAC audio and x264 video in lossless mode, muxed into in a Matroska container. Because we used speed-preset=ultrafast, the buffers should just smooth out the flow of frames through the pipelines. Even though the buffers are set to the maximum possible size, speed-preset=veryslow would eventually fill the video buffer and start dropping frames.

Some other things to note about this pipeline:

  • FFmpeg's H.264 page includes a useful discussion of speed presets (both programs use the same underlying library)
  • quantizer=0 sets the video codec to lossless mode (~30GB/hour). Anything up to quantizer=18 should not lose information visible to the human eye, and will produce much smaller files (~10GB/hour)
  • min-index-interval=1000000000 improves seek times by telling the Matroska muxer to create one cue data entry per second of playback. Cue data is a few kilobytes per hour, added to the end of the file when encoding completes. If you try to watch your Matroska video while it's being recorded, it will take a long time to skip forward/back because the cue data hasn't been written yet

Common caputuring issues and their solutions

Reducing Jerkiness

If motion that should appear smooth instead stops and starts, try the following:

Check for muxer issues. Some muxers need big chunks of data, which can cause one stream to pause while it waits for the other to fill up. Change your pipeline to pipe your audio and video directly to their own filesinks - if the separate files don't judder, the muxer is the problem.

  • If the muxer is at fault, add ! queue max-size-buffers=0 max-size-time=0 max-size-bytes=0 immediately before each stream goes to the muxer
    • queues have hard-coded maximum sizes - you can chain queues together if you need more buffering than one buffer can hold

Check your CPU load. When GStreamer uses 100% CPU, it may need to drop frames to keep up.

  • If frames are dropped occasionally when CPU usage spikes to 100%, add a (larger) buffer to help smooth things out.
    • this can be a source's internal buffer (e.g. alsasrc buffer-time=2000000), or it can be an extra buffering step in your pipeline (! queue max-size-buffers=0 max-size-time=0 max-size-bytes=0)
  • If frames are dropped when other processes have high CPU load, consider using nice to make sure encoding gets CPU priority
  • If frames are dropped regularly, use a different codec, change the parameters, lower the resolution, or otherwise choose a less resource-intensive solution

As a general rule, you should try increasing buffers first - if it doesn't work, it will just increase the pipeline's latency a bit. Be careful with nice, as it can slow down or even halt your computer.

Check for incorrect timestamps. If your video driver works by filling up an internal buffer then passing a cluster of frames without timestamps, GStreamer will think these should all have (nearly) the same timestamp. Make sure you have a videorate element in your pipeline, then add silent=false to it. If it reports many framedrops and framecopies even when the CPU load is low, the driver is probably at fault.

  • videorate on its own will actually make this problem worse by picking one frame and replacing all the others with it. Instead install entrans and add its stamp element between v4l2src and queue (e.g. v4l2src do-timestamp=true ! stamp sync-margin=2 sync-interval=5 ! videorate ! queue)
    • stamp intelligently guesses timestamps if drivers don't support timestamping. Its sync- options drop or copy frames to get a nearly-constant framerate. Using videorate as well does no harm and can solve some remaining problems

Avoiding pitfalls with video noise

If your video contains periods of video noise (snow), you may need to deal with some extra issues:

  • Most devices send an EndOfStream signal if the input signal quality drops too low, causing GStreamer to finish capturing. To prevent the device from sending EOS, set num-buffers=-1 on the v4l2src element.
  • The stamp plugin gets confused by periods of snow, causing it to generate faulty timestamps and framedropping. stamp will recover normal behaviour when the break is over, but will probably leave the buffer full of weirdly-stamped frames. stamp only drops one weirdly-stamped frame each sync-interval, so it can take several minutes until everything works fine again. To solve this problem, set leaky=2 on each queue element to allow dropping old frames
  • Periods of noise (snow, bad signal etc.) are hard to encode. Variable bitrate encoders will often drive up the bitrate during the noise then down afterwards to maintain the average bitrate. To minimise the issues, specify a minimum and maximum bitrate in your encoder
  • Snow at the start of a recording is just plain ugly. To get black input instead from a VCR, use the remote control to change the input source before you start recording

Investigating bugs in GStreamer

GStreamer comes with a extensive tracing system that let you figure-out the problems. Yet, you often need to understand the internals of GStreamer to be able to read those traces. You should read this documentation page for the basic of how the tracing system works. When something goes wrong you should:

  1. try and see if there is a useful error message by enabling the ERROR debug level, GST_DEBUG=2 gst-launch-1.0
  2. try similar pipelines - reducing to its most minimal form, and add more elements until you can reproduce the issue.
  3. as you are most likely having issue with V4L2 element, you may enable full v4l2 traces using GST_DEBUG="v4l2*:7,2" gst-launch-1.0.
  4. find an error message that looks relevant, search the Internet for information about it
  5. try more variations based on what you learnt, until you eventually find something that works
  6. ask on Freenode #gstreamer or through GStreamer Mailing List
  7. if you think you found a bug, you should report it through Gnome Bugzilla

Sample pipelines

record from a bad analog signal to MJPEG video and RAW mono audio

gst-launch-1.0 \
    v4l2src device=$VIDEO_DEVICE do-timestamp=true \
        ! $VIDEO_CAPABILITIES \
        ! videorate \
        ! $VIDEO_CAPABILITIES \
        ! videoconvert \
        ! $VIDEO_CAPABILITIES \
        ! jpegenc \
        ! queue \
        ! mux. \
    alsasrc device=$AUDIO_DEVICE \
        ! $AUDIO_CAPABILITIES \
        ! audiorate \
        ! audioresample \
        ! $AUDIO_CAPABILITIES, rate=44100 \
        ! audioconvert \
        ! $AUDIO_CAPABILITIES, rate=44100, channels=1 \
        ! queue \
        ! mux. \
    avimux name=mux ! filesink location=test-$( date --iso-8601=seconds ).avi

The chip that captures audio and video might not deliver the exact framerates specified, which the AVI format can't handle. The audiorate and videorate elements remove or duplicate frames to maintain a constant rate.

View pictures from a webcam (GStreamer 0.10)

gst-launch-0.10 \
    v4l2src do-timestamp=true device=$VIDEO_DEVICE \
        ! video/x-raw-yuv,format=\(fourcc\)UYVY,width=320,height=240 \
        ! ffmpegcolorspace \
        ! autovideosink

In GStreamer 0.10, videoconvert was called ffmpegcolorspace.

Entrans: Record to DVD-compliant MPEG2 (GStreamer 0.10)

entrans -s cut-time -c 0-180 -v -x '.*caps' --dam -- --raw \
    v4l2src queue-size=16 do-timestamp=true device=$VIDEO_DEVICE norm=PAL-BG num-buffers=-1 \
        ! stamp silent=false progress=0 sync-margin=2 sync-interval=5 \
        ! queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
        ! dam \
        ! cogcolorspace \
        ! videorate silent=false \
        ! 'video/x-raw-yuv,width=720,height=576,framerate=25/1,interlaced=true,aspect-ratio=4/3' \
        ! queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
        ! ffenc_mpeg2video rc-buffer-size=1500000 rc-max-rate=7000000 rc-min-rate=3500000 bitrate=4000000 max-key-interval=15 pass=pass1 \
        ! queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
        ! mux. \
    pulsesrc buffer-time=2000000 do-timestamp=true \
        ! queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
        ! dam \
        ! audioconvert \
        ! audiorate silent=false \
        ! audio/x-raw-int,rate=48000,channels=2,depth=16 \
        ! queue silent=false max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
        ! ffenc_mp2 bitrate=192000 \
        ! queue silent=false leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 \
        ! mux. \
    ffmux_mpeg name=mux \
        ! filesink location=test-$( date --iso-8601=seconds ).mpg

This captures 3 minutes (180 seconds, see first line of the command) to test-$( date --iso-8601=seconds ).mpg and even works for bad input signals.

  • I wasn't able to figure out how to produce a mpeg with ac3-sound as neither ffmux_mpeg nor mpegpsmux support ac3 streams at the moment. mplex does but I wasn't able to get it working as one needs very big buffers to prevent the pipeline from stalling and at least my GStreamer build didn't allow for such big buffers.
  • The limited buffer size on my system is again the reason why I had to add a third queue element to the middle of the audio as well as of the video part of the pipeline to prevent jerking.
  • In many HOWTOs you find ffmpegcolorspace instead of cogcolorspace. You can even use this but cogcolorspace is much faster.
  • It seems to be important that the video/x-raw-yuv,width=720,height=576,framerate=25/1,interlaced=true,aspect-ratio=4/3-statement is after videorate as videorate seems to drop the aspect-ratio-metadata otherwise resulting in files with aspect-ratio 1 in theis headers. Those files are probably played back warped and programs like dvdauthor complain.

Bash script to record video tapes with entrans

For most use cases, you'll want to wrap GStreamer in a larger shell script. This script protects against several common mistakes during encoding.

See also the V4L capturing script for a a wrapper that represents a whole workflow.

#!/bin/bash
 
 targetdirectory="~/videos"
 
 
 # Test ob doppelt geöffnet
 
 if [[ -e "~/.lock_shutdown.digitalisieren" ]]; then
     echo ""
     echo ""
     echo "Capturing already running. It is impossible to capture to tapes simultaneously. Hit a key to abort."
     read -n 1
     exit
 fi
 
 # trap keyboard interrupt (control-c)
 trap control_c 0 SIGHUP SIGINT SIGQUIT SIGABRT SIGKILL SIGALRM SIGSEGV SIGTERM
 
 control_c()
 # run if user hits control-c
 {
   cleanup
   exit $?
 }
 
 cleanup()
 {
   rm ~/.lock_shutdown.digitalisieren
   return $?
 }
 
 touch "~/.lock_shutdown.digitalisieren"
 
 echo ""
 echo ""
 echo "Please enter the length of the tape in minutes and press ENTER. (Press Ctrl+C to abort.)"
 echo ""
 while read -e laenge; do
     if [[ $laenge == [0-9]* ]]; then
         break 2
     else
         echo ""
         echo ""
         echo "That's not a number."
         echo "Please enter the length of the tape in minutes and press ENTER. (Press Ctrl+C to abort.)"
         echo ""
     fi
 done
 
 let laenge=laenge+10  # Sicherheitsaufschlag, falls Band doch länger
 let laenge=laenge*60
 
 echo ""
 echo ""
 echo "Please type in the description of the tape."
 echo "Don't forget to rewind the tape?"
 echo "Hit ENTER to start capturing. Press Ctrl+C to abort."
 echo ""
 read -e name;
 name=${name//\//_}
 name=${name//\"/_}
 name=${name//:/_}
 
 # Falls Name schon vorhanden
 if [[ -e "$targetdirectory/$name.mpg" ]]; then
     nummer=0
     while [[ -e "$targetdirectory/$name.$nummer.mpg" ]]; do
        let nummer=nummer+1
     done
     name=$name.$nummer
 fi
 
 # Audioeinstellungen setzen: unmuten, Regler
 amixer -D pulse cset name='Capture Switch' 1 >& /dev/null      # Aufnahme-Kanal einschalten
 amixer -D pulse cset name='Capture Volume' 20724 >& /dev/null  # Aufnahme-Pegel einstellen
 
 # Videoinput auswählen und Karte einstellen
 v4l2-ctl --set-input 3 >& /dev/null
 v4l2-ctl -c saturation=80 >& /dev/null
 v4l2-ctl -c brightness=130 >& /dev/null
 
 let ende=$(date +%s)+laenge
 
 echo ""
 echo "Working"
 echo "Capturing will be finished at "$(date -d @$ende +%H.%M)"."
 echo ""
 echo "Press Ctrl+C to finish capturing now."
 
 
 nice -n -10 entrans -s cut-time -c 0-$laenge -m --dam -- --raw \
 v4l2src queue-size=16 do-timestamp=true device=$VIDEO_DEVICE norm=PAL-BG num-buffers=-1 ! stamp sync-margin=2 sync-interval=5 silent=false progress=0 ! \
    queue leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! dam ! \
    cogcolorspace ! videorate ! \
    'video/x-raw-yuv,width=720,height=576,framerate=25/1,interlaced=true,aspect-ratio=4/3' ! \
    queue leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! \
    ffenc_mpeg2video rc-buffer-size=1500000 rc-max-rate=7000000 rc-min-rate=3500000 bitrate=4000000 max-key-interval=15 pass=pass1 ! \
    queue leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! mux. \
 pulsesrc buffer-time=2000000 do-timestamp=true ! \
    queue leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! dam ! \
    audioconvert ! audiorate ! \
    audio/x-raw-int,rate=48000,channels=2,depth=16 ! \
    queue max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! \
    ffenc_mp2 bitrate=192000 ! \
    queue leaky=2 max-size-buffers=0 max-size-time=0 max-size-bytes=0 ! mux. \
 ffmux_mpeg name=mux ! filesink location=\"$targetdirectory/$name.mpg\" >& /dev/null
 
 echo "Finished Capturing"
 rm ~/.lock_shutdown.digitalisieren

The script uses a command line similar to this to produce a DVD compliant MPEG2 file.

  • The script aborts if another instance is already running.
  • If not it asks for the length of the tape and its description
  • It records to description.mpg or if this file already exists to description.0.mpg and so on for the given time plus 10 minutes. The target-directory has to be specified in the beginning of the script.
  • As setting of the inputs and settings of the capture device is only partly possible via GStreamer other tools are used.
  • Adjust the settings to match your input sources, the recording volume, capturing saturation and so on.

Further documentation resources

External Links