- Generated by
1.16.1
|
PipeWire 1.6.0
|
The filter-chain allows you to create an arbitrary processing graph from LADSPA, LV2, sofa, ffmpeg and builtin filters.
This filter can be made into a virtual sink/source or between any 2 nodes in the graph.
The filter chain is built with 2 streams, a capture stream providing the input to the filter chain and a playback stream sending out the filtered stream to the next nodes in the graph.
Because both ends of the filter-chain are built with streams, the session manager can manage the configuration and connection with the sinks and sources automatically.
libpipewire-module-filter-chain
The general structure of the graph description is as follows:
Nodes describe the processing filters in the graph. Use a tool like lv2ls or listplugins to get a list of available plugins, labels and the port names.
Some examples ladspa and lv2 plugins:
Links can be made between ports of nodes. The portname is given as <node_name>:<port_name>.
You can tee the output of filters to multiple other filters. You need to use a mixer if you want the output of multiple filters to go into one filter input port.
links can be omitted when the graph has just 1 filter.
These are the entry and exit ports into the graph definition. Their number defines the number of channels used by the filter-chain.
The <portname> can be null when a channel is to be ignored.
Each input/output in the graph can only be linked to one filter input/output. You need to use the copy builtin filter if the stream signal needs to be routed to multiple filters. You need to use the mixer builtin plugin if multiple graph outputs need to go to one output stream.
inputs and outputs can be omitted, in which case the filter-chain will use all inputs from the first filter and all outputs from the last filter node. The graph will then be duplicated as many times to match the number of input/output channels of the streams.
If the graph has no inputs and the capture channels is set as 0, only the playback stream will be created. Likewise, if there are no outputs and the playback channels is 0, there will be no capture stream created.
Normally the volume of the sink/source is handled by the stream software volume. With the capture.volumes and playback.volumes properties this can be handled by a control port in the graph instead. Use capture.volumes for the volume of the input of the filter (when for example used as a sink). Use playback,volumes for the volume of the output of the filter (when for example used as a source).
The min and max values (defaults 0.0 and 1.0) respectively can be used to scale and translate the volume min and max values.
Normally the control values are linear and it is assumed that the plugin does not perform any scaling to the values. This can be changed with the scale property. By default this is linear but it can be set to cubic when the control applies a cubic transformation.
There are some useful builtin filters available. The type should be builtin and you select the specific builtin filter with the label of the filter node.
Use the mixer plugin if you have multiple input signals that need to be mixed together.
The mixer plugin has up to 8 input ports labeled "In 1" to "In 8" and each with a gain control labeled "Gain 1" to "Gain 8". There is an output port labeled "Out". Unused input ports will be ignored and not cause overhead.
Use the copy plugin if you need to copy a stream input signal to multiple filters.
It has one input port "In" and one output port "Out".
Biquads can be used to do all kinds of filtering. They are also used when creating equalizers.
All biquad filters have an input port "In" and an output port "Out". They have a "Freq", "Q" and "Gain" control. Their meaning depends on the particular biquad that is used. The biquads also have "b0", "b1", "b2", "a0", "a1" and "a2" ports that are read-only except for the bq_raw biquad, which can configure default values depending on the graph rate and change those at runtime.
We refer to https://arachnoid.com/BiQuadDesigner/index.html for an explanation of the controls.
The following labels can be used:
The parametric EQ chains a number of biquads together. It is more efficient than specifying a number of chained biquads and it can also load configuration from a file.
The parametric EQ supports multichannel processing and has 8 input and 8 output ports that don't all need to be connected. The ports are named In 1 to In 8 and Out 1 to Out 8.
Either a filename or a filters array can be specified. The configuration will be used for all channels. Alternatively filenameX or filtersX where X is the channel number (between 1 and 8) can be used to load a channel specific configuration.
The filename must point to a parametric equalizer configuration generated from the AutoEQ project or Squiglink. Both the projects allow equalizing headphones or an in-ear monitor to a target curve.
A popular example of the above being EQ'ing to the Harman target curve or EQ'ing one headphone/IEM to another.
For AutoEQ, see https://github.com/jaakkopasanen/AutoEq. For SquigLink, see https://squig.link/.
Parametric equalizer configuration generated from AutoEQ or Squiglink looks like below.
Fc, Gain and Q specify the frequency, gain and Q factor respectively. The fourth column can be one of PK, LSC or HSC specifying peaking, low shelf and high shelf filter respectively. More often than not only peaking filters are involved.
The filters (or channel specific filtersX where X is the channel between 1 and 8) can contain an array of filter specification object with the following keys:
type specifies the filter type, choose one from the available biquad labels. freq is the frequency passed to the biquad. gain is the gain passed to the biquad. q is the Q passed to the biquad.
This makes it possible to also use the param eq without a file and with all the available biquads.
The convolver can be used to apply an impulse response to a signal. It is usually used for reverbs or virtual surround. The convolver is implemented with a fast FFT implementation.
The convolver has an input port "In" and an output port "Out". It requires a config section in the node declaration in this format:
The delay can be used to delay a signal in time. With the Feedback and Feedforward controls it can also be used as a comb and an allpass filter.
The delay has an input port "In" and an output port "Out". It also has a "Delay (s)" control port and a "Feedback" and "Feedforward" port. It requires a config section in the node declaration in this format:
With the "Feedback" port one can create a comb filter. With the "Feedback" port and "Feedforward" port set to A and -A respectively, one can create an allpass filter. These settings can be used to create custom reverb units.
The invert plugin can be used to invert the phase of the signal.
It has an input port "In" and an output port "Out".
The clamp plugin can be used to clamp samples between min and max values.
It has an input port "In" and an output port "Out". It also has a "Control" and "Notify" port for the control values.
The final result is clamped to the "Min" and "Max" control values.
The linear plugin can be used to apply a linear transformation on samples or control values.
It has an input port "In" and an output port "Out". It also has a "Control" and "Notify" port for the control values.
The control value "Mult" and "Add" are used to configure the linear transform. Each sample or control value will be calculated as: new = old * Mult + Add.
The recip plugin can be used to calculate the reciprocal (1/x) of samples or control values.
It has an input port "In" and an output port "Out". It also has a "Control" and "Notify" port for the control values.
The abs plugin can be used to calculate the absolute value of samples.
It has an input port "In" and an output port "Out".
The sqrt plugin can be used to calculate the square root of samples.
It has an input port "In" and an output port "Out".
The exp plugin can be used to calculate the exponential (base^x) of samples or control values.
It has an input port "In" and an output port "Out". It also has a "Control" and "Notify" port for the control values.
The control value "Base" is used to calculate base ^ x for each sample.
The log plugin can be used to calculate the logarithm of samples or control values.
It has an input port "In" and an output port "Out". It also has a "Control" and "Notify" port for the control values.
The control value "Base", "M1" and "M2" are used to calculate out = M2 * log2f(fabsf(in * M1)) / log2f(Base) for each sample.
The mult plugin can be used to multiply samples together.
It has 8 input ports named "In 1" to "In 8" and an output port "Out".
All input ports samples are multiplied together into the output. Unused input ports will be ignored and not cause overhead.
The sine plugin generates a sine wave.
It has an output port "Out" and also a control output port "notify".
"Freq", "Ampl", "Offset" and "Phase" can be used to control the sine wave frequency, amplitude, offset and phase.
Use the max plugin if you need to select the max value of a number of input ports.
It has 8 input ports named "In 1" to "In 8" and one output port "Out".
All input ports samples are checked to find the maximum value per sample. Unused input ports will be ignored and not cause overhead.
Use the dcblock plugin implements a DC blocker.
It has 8 input ports "In 1" to "In 8" and corresponding output ports "Out 1" to "Out 8". Not all ports need to be connected.
It also has 1 control input port "R" that controls the DC block R factor.
Use the ramp plugin creates a linear ramp from Start to Stop.
It has 3 input control ports "Start", "Stop" and "Duration (s)". It also has one output port "Out". A linear ramp will be created from "Start" to "Stop" for a duration given by the "Duration (s)" control in (fractional) seconds. The current value will be stored in the output notify port "Current".
The ramp output can, for example, be used as input for the mult plugin to create a volume ramp up or down. For more a more coarse volume ramp, the "Current" value can be used in the linear plugin.
The debug plugin can be used to debug the audio and control data of other plugins.
It has an "In" input port and an "Out" output data ports. The data from "In" will be copied to "Out" and the data will be dumped into the INFO log.
There is also a "Control" input port and an "Notify" output control ports. The control from "Control" will be copied to "Notify" and the control value will be dumped into the INFO log.
The pipe plugin can be used to filter the audio with another application using pipes for sending and receiving the raw audio.
The application needs to consume raw float32 samples from stdin and produce filtered float32 samples on stdout.
It has an "In" input port and an "Out" output data ports.
The node requires a config section with extra configuration:
The zeroramp plugin can be used to detect unnatural silence parts in the audio stream and ramp the volume down or up when entering or leaving the silent area respectively. This can be used to avoid loud pops and clicks that occur when the sample values suddenly drop to zero or jump from zero to a large value caused by a pause, resume or an error of the stream. It only detect areas where the sample values are absolute zero values, such as those inserted when pausing a stream.
It has an "In" input port and an "Out" output data ports.
There are also "Gap (s)" and an "Duration (s)" input control ports. "Gap (s)" determines how long the silence gap is in seconds (default 0.000666) and "Duration (s)" determines how long the fade-in and fade-out should last (default 0.000666).
The noisegate plugin can be used to remove low volume noise.
It has an "In" input port and an "Out" output data ports. Normally the input data is passed directly to the output.
The "Level" control port can be used to control the measured volume of the "In" port. When not connected, a simple volume algorithm on the "In" port will be used.
If the volume drops below "Close threshold", the noisegate will ramp down the volume to zero for a duration of "Release (s)" seconds. When the volume is above "Open threshold", the noisegate will ramp up the volume to 1 for a duration of "Attack (s)" seconds. The noise gate stays open for at least "Hold (s)" seconds before it can close again.
The busy plugin has no input or output ports and it can be used to keep the CPU or graph busy for the given percent of time.
The node requires a config section with extra configuration:
The null plugin has one data input "In" and one control input "Control" that simply discards the data.
There is an optional sofa type available (when compiled with libmysofa).
The spatializer can be used to place the sound in a 3D space.
The spatializer has an input port "In" and a stereo pair of output ports called "Out L" and "Out R". It requires a config section in the node declaration in this format:
The control can be changed at runtime to move the sounds around in the 3D space.
There is an optional EBU R128 plugin available (when compiled with libebur128) selected with the ebur128 type. Filters in the plugin can be selected with the label field.
The ebur128 filter can be used to measure the loudness of a signal.
It has 7 input ports "In FL", "In FR", "In FC", "In UNUSED", "In SL", "In SR" and "In DUAL MONO", corresponding to the different input channels for EBUR128. Not all ports need to be connected for this filter.
The input signal is passed unmodified on the "Out FL", "Out FR", "Out FC", "Out UNUSED", "Out SL", "Out SR" and "Out DUAL MONO" output ports.
There are 7 output control ports that contain the measured loudness information and that can be used to control the processing of the audio. Some of these ports contain values in LUFS, or "Loudness Units relative to Full Scale". These are negative values, closer to 0 is louder. You can use the lufs2gain plugin to convert this value to a gain to adjust a volume (See below).
"Momentary LUFS" contains the momentary loudness measurement with a 400ms window and 75% overlap. It works mostly like an R.M.S. meter.
"Shortterm LUFS" contains the shortterm loudness in LUFS over a 3 second window.
"Global LUFS" contains the global integrated loudness in LUFS over the max-history window. "Window LUFS" contains the global integrated loudness in LUFS over the max-window window.
"Range LU" contains the loudness range (LRA) in LU units.
"Peak" contains the peak loudness.
"True Peak" contains the true peak loudness oversampling the signal. This can more accurately reflect the peak compared to "Peak".
The node also has an optional config section with extra configuration:
The lufs2gain filter can be used to convert LUFS control values to gain. It needs a target LUFS control input to drive the conversion.
It has 2 input control ports "LUFS" and "Target LUFS" and will produce 1 output control value "Gain". This gain can be used as input for the builtin linear filter, for example, to adust the gain.
There is an optional FFmpeg filter available (when compiled with libavfilter) that can be selected with the ffmpeg type. Use the plugin field to select the plugin to use.
The filtergraph FFmpeg plugin is selected with the filtergraph plugin field in the node.
The filtergraph filter allows you to specify an set of audio filters using the FFmpeg filtergraph syntax (https://ffmpeg.org/ffmpeg-filters.html).
The label field should be used to describe the filtergraph in use.
FFmpeg filtergraph input and output ports can have multiple channels. The filter-chain can split those into individual ports to use as input and output ports. For this, the ports in the filtergraph need to have a specific name convention, either <port-name>_<channel-name> or <port-name>_<channel-layout>.
When a single channel is specified, the port can be referenced in inputs and outputs sections with <name>:<port-name>_<channel-name>. When a channel-layout is specified, each port name gets a _<number> appended, starting from 0 and counting up for each channel in the layout.
The filtergraph plugin will automatically add format converters when the input port channel-layout, format or graph sample-rates don't match.
Note that the FFmpeg filtergraph is not Real-time safe because it might do allocations from the processing thread. It is advised to run the filter-chain streams in async mode (node.async = true) to avoid interrupting the other RT threads.
Some examples:
The stereo ports are split into their channels with the _0 and _1 suffixes.
It is possible to have multiple input and output ports for the filtergraphs. In the next example, the ports have a single channel name and so don't have the _0 suffix to identify them. This can be simplified by removing the amerge and channelsplit filters and using the _stereo suffix on port names to let PipeWire do the splitting and merging more efficiently.
Here is a last example of a surround sound upmixer:
There is an optional ONNX filter available (when compiled with libonnxruntime) that can be selected with the onnx type. Use the label field to select the model to use and how to map the tensors to ports.
The label must contain an object with the configuration of the plugin.
The input-tensors and output-tensors configuration must contain an object with keys named after the tensors in the model and the value must be an object with the the following keys:
Here is an example of the silero VAD model:
Options with well-known behavior. Most options can be added to the global configuration or the individual streams:
Stream only properties:
This example uses the rnnoise LADSPA plugin to create a new virtual source.
Run with pipewire -c filter-chain.conf. The configuration can also be put under pipewire.conf.d/ to run it inside the PipeWire server.
This example uses the ladpsa surround encoder to encode a 5.1 signal to a stereo Dolby Surround signal.
1.16.1