Audio::Ecasound::Multitrack - Perl extensions for multitrack audio processing
Nama - Lightweight recorder, mixer and mastering system
nama [options] [project_name]
Nama is a lightweight recorder/mixer application using Ecasound in the back end to provide effects processing, cut-and-paste, mastering, and other functions typically found in digital audio workstations.
By default, Nama starts up a GUI interface with a command line interface running in the terminal window. The -t option provides a text-only interface for console users.
Use project_root as Nama's top-level directory.
Output debugging information
Use config_file instead of default .namarc
GUI mode (default)
Text-only mode
Create the specified project if necessary
Save and reload ALSA mixer state using alsactl
Don't load saved state
Ecasound is configured through use of chain setups. Chain setups are central to controlling Ecasound. Nama generates appropriate chain setups for recording, playback, and mixing covering a large portion of Ecasound's functionality.
Commands for audio processing with Nama/Ecasound fall into two categories: static commands that influence the chain setup and dynamic commands that influence the realtime behavior of the audio processing engine.
The rec, mon and B <off> commands are examples of static commands.
They influence whether the current track will get
its audio stream from a live source (soundcard or JACK
client) and store its as a WAV file, for whether
an existing file will be played back.
Other static commands include loop_enable
and the stereo/mono commands which specify
the width of a track's signal source. Like a spreadsheet
recalculating, Nama automatically detects when static
commands require that Ecasound be reconfigured with a new
chain setup.
Once a chain setup is loaded and the engine launched,
another subset of commands controls the audio processing
engine. Commonly used dynamic commands include start
and stop; forward, rewind and setpos commands
for repositioning the playback head; and vol and pan
for adjusting effect parameters. Effect parameters may be
adjusted at any time. Effects may be added during audio
processing.
The current chain setup may be inspected with the chains
command. The showio command displays the data structure
used to generate the chain setup. dump displays data for
the current track. dumpall shows the state of most
program objects and variables (identical to the State.yml
file created by the save command.)
Invoked by default, the Tk interface provides all functionality on two panels, one for general control, the second for effects.
Many plugins include hinting data, which is used to set parameter range and to provide logarithmic sliders for such parameters as frequency.
The GUI provides text-entry widgets to specify parameters for plugins without hinting.
The GUI project name bar and time display change color to indicate whether the upcoming operation will include live recording (red), mixdown only (yellow) or playback only (green). Live recording and mixdown can take place simultaneously.
The text command prompt appears in the terminal window during GUI operation, and text commands may be issued at any time.
Press the Enter key if necessary to get the following command prompt.
nama ('h' for help)>
You can enter Nama and Ecasound commands directly, Perl code
preceded by eval or shell code preceded by !.
Multiple commands on a single line are allowed if delimited
by semicolons. Usually the lines are split on semicolons and
the parts are executed sequentially, however if the line
begins with eval or ! the entire line will be given to
the corresponding interpreter.
You can access command history using up-arrow/down-arrow.
Type help for general help, help command for help with
command, help foo for help with commands containing
the string foo. help_effect foo bar lists all
plugins/presets/controller containing both foo and
bar. Tab-completion is provided for Nama commands, Ecasound-iam
commands, plugin/preset/controller names, and project names.
Ecasound deals with audio processing at the level of devices, files, and signal-processing chains. Nama implements tracks to provide a level of control and convenience comparable to many digital audio workstations.
Each track has a descriptive name (i.e. vocal) and an integer track-number assigned when the track is created.
Multiple WAV files can be recorded for each track. These are identified by a version number that increments with each recording run, i.e. sax_1.wav, sax_2.wav, etc. All files recorded at the same time have the same version numbers.
Version numbers for playback can be selected at the group or track level. By setting the group version number to 5, you can play back the fifth take of a song, or perhaps the fifth song of a live recording session.
The track's version setting, if present, overrides the group setting. Setting the track version to zero restores control of the version number to the default group setting.
REC/MON/OFF status is used to generate the chain setup for an audio processing run.
Each track, including Master and Mixdown, has its own REC/MON/OFF setting and displays its own REC/MON/OFF status. The Tracker group, which includes all user tracks, also has REC, MON and OFF settings. These provides a convenient way to control the behavior of all user tracks.
As the name suggests, REC status indicates that a track is ready to record a WAV file. You need to set both track and group to REC to source an audio stream from JACK or the soundcard.
MON status indicates an audio stream available from disk. It requires a MON setting for the track or group as well as the presence of file with the selected version number. A track set to REC with no audio stream available with default to MON status.
OFF status means that no audio is available for the track from any source. A track with no recorded WAV files will show OFF status, even if set to MON.
An OFF setting for the track or group always results in OFF
status. A track with OFF status will be excluded from the
chain setup. (This setting is distinct from the action of
the mute command, which sets the volume of the track to
zero.)
All user tracks belong to the Tracker group, which has a group REC/MON/OFF setting and a default version setting for the entire group.
Setting the group to MON (C<group_monitor> or C<gmon>) forces user tracks with a REC setting to MON status if a WAV file is available to play, or OFF status if no audio stream is available.
The group MON mode triggers automatically after a recording has created new WAV files.
The group OFF setting (text command group_off) excludes all user tracks from the chain setup, and is typically used when playing back mixdown tracks. The mixplay command sets the Mixdown track to MON and the Tracker group to OFF.
The Master bus has only MON/OFF status. Setting REC status
for the Mixdown bus has the same effect as issuing the
mixdown command. (A start command must be issued for
mixdown to commence.)
The region command allows you to define endpoints
for a portion of an audio file. Use the shift command
to specify a delay for starting playback.
Only one region may be specified per track. Use the
link_track command to clone a track in order to make use
of multiple regions or versions of a single track.
link_track can clone tracks from other projects.
Thus you could create the sections of a song in
separate projects, then assemble them using
link_track to pull the Mixdown tracks
into a single project for mastering.
Nama identifies tracks by both a name and a number. The track number is used to identify corresponding Ecasound signal-processing chains.
Nama's signal flow is organized at three levels: raw, cooked and mixed.
"Raw" signals are the inputs to user tracks. Raw signals can come from the soundcard, a WAV file, or a JACK client.
"Cooked" signals are the output of user tracks after volume, pan and effects processing.
The "mixed" signal is the combined outputs of all user tracks. It is delivered to the Master fader, and to the Mixdown WAV file during mixdown.
Nama uses Ecasound loop devices to be able to deliver each of these signals classes to multiple "customers", i.e. to other chains using that signal as input.
An optimizing pass eliminates loop devices that have only one customer for the signal they provide.
Let's examine the signal flow from track 3, the first available user track. Assume track 3 is named "sax". All effects for track 3 are applied to chain 3.
We will divide the signal flow into track and mixer sections. Parentheses indicate chain identifiers or the corresponding track name.
All "cooked" signals (i.e. the outputs of each user track) terminate at loop,111.
Sound device --+---(3)----> loop,3 ---(J3)----> loop,111
/JACK client |
+---(R3)---> sax_1.wav
REC status indicates that the source of the signal is the soundcard or JACK client. The input signal will be written directly to a file except in the preview and doodle modes.
sax_1.wav ------(3)----> loop,3 ----(J3)----> loop,111
In the second part of the flow graph, the mixed signal is delivered to an output device through the Master chain, which can host additional effects. The Mixdown track can also host effects, however these should be used during playback only.
loop,111 --(MixLink)---> loop,222 --(1/Master)---> Sound device
|
+------(2/Mixdown)--> Mixdown_1.wav
loop,111 --(1/Master)-> loop,222 -> Sound device
|
+->(2/Mixdown)--> Mixdown_1.wav
loop,111 --(1/Master)-> loop,221---(mastering)-->loop,222 -> Sound device
|
+->(2/Mixdown)--> Mixdown_1.wa
In Mastering mode, Master outputs to $loop_mastering
In normal mode, Master outputs to $loop_output, optimized to soundcard
In mixdown mode, Master outputs to $loop_output optimized to file
In mixdown mode, no effects will be applied on the mixdown track, just route to file
In mixplay mode, effects may be applied as normal to Mixdown track.
TODO
Toggle needed on sound device output
revise normalizing_mixdown automix nmix
In mastering mode (invoked by master_on and released
master_off) the MixLink chain is replaced by several
tracks with mastering-related effects.
+---(Low)---+
| |
loop,111 --(Eq)--> loop,120 -+---(Mid)---+ loop,130 --(Boost)--> loop,222
| |
+---(High)--+
The Eq track hosts an equalizer.
The Low, Mid and High tracks each apply a bandpass filter, a compressor and a spatialiser.
The Boost track applies gain and a limiter.
These effects (and optional default parameters) are defined in the configuration file .namarc.
These non-recording modes, invoked by preview and doodle commands
tweak the routing rules for special purposes. Preview
mode simply disables recording of WAV files to disk. Doodle
mode disables MON inputs while enabling only one REC track per
signal source. The arm command releases both preview
and doodle modes.
[% qx(./emit_command_headers pod) %]
Text commands are required to invoke preview and mastering modes, and to define and shift track regions.
Compensation for LADSPA plugin latencies and ADC and buffering delays during multitrack recording is not fully implemented.
The default GUI colors aren't great.
No waveform display.
The main module, Multitrack.pm is a concatenation of several source files. Patches should be made against the source files (see below.)
None by default.
CPAN, for the distribution.
cpan Audio::Ecasound::Multitrack
You will need to install Tk to use the GUI.
cpan Tk
You can pull the source code as follows:
git clone git://github.com/bolangi/nama.git
Build instructions are contained in the README file.
Joel Roth, <joelz@pobox.com>