• mcfx is a suite of multichannel vst plug-ins or standalone applications (standalone currently meter and convolver only)

• channel count is configurable with compile time flag

• cross plattform VST for MacOSX, Windows and Linux

• uses the JUCE framework (www.juce.com, GPLv3), libsoxr (LGPL, http://soxr.sourceforge.net)

• ready to use binaries for MacOSX (> 10.5, 32/64 bit) and Windows (32/64 bit) can be found at http://www.matthiaskronlachner.com/?p=1910

• these plug-ins have been developed to be used with the ambiX Ambisonic plug-ins: http://www.matthiaskronlachner.com/?p=2015

mcfx is free software and licensed under the GNU General Public License version 3 (GPLv3).

CMake, working build environment

• Linux:
  Install LINUX Libraries (Debian, Ubuntu):

  sudo apt install libasound2-dev libjack-jackd2-dev \
      ladspa-sdk \
      libcurl4-openssl-dev  \
      libfreetype6-dev \
      libx11-dev libxcomposite-dev libxcursor-dev libxcursor-dev libxext-dev libxinerama-dev libxrandr-dev libxrender-dev \
      libwebkit2gtk-4.0-dev \
      libglu1-mesa-dev mesa-common-dev \
      libfftw3-dev \
      libzita-convolver3 \
      libzita-convolver-dev
  

• Windows x64:
  Download FFTW for Windows 64bit and run the command lib /machine:x64 /def:libfftw3f-3.def from the Visual Studio Developer Command Prompt to generate libfftw3f-3.lib.
  Alternatively, run the command lib /def:libfftw3f-3.def from the x64 Native Tools Command Prompt for VS. Note that this simpler command, instructed by the FFTW documentation, will generate a 32bit library when called from the default Developer Command Prompt or Powershell (and therefore fail to link)

Make sure of using fftw-3.3.6-pl2 or later, as earlier versions do not contain the fftwf_make_planner_thread_safe() function.

Clone the repository with the submodules:

git clone --recurse-submodules https://github.com/kronihias/mcfx/

Use cmake gui or cmake/ccmake from terminal to configure the build.

1. Create a folder in the mcfx folder eg. BUILD

   mkdir BUILD # On Linux, MacOS or Windows Powershell    
   

2. Run cmake or ccmake from the BUILD folder, or use the cmake gui and press Configure.
   In the terminal:

   cd BUILD
   ccmake ..
   

   Press the C key.

   If using the GUI instead, select the target build system (e.g., in Windows use Visual Studio or MinGW).

3. Set the Build Parameters, Configure again. Then Generate.

4. Build the project with the selected build system (e.g., Visual Studio, Xcode, or make).
   E.g., for Linux:

   make -j$(nproc) config=Release
   

   -j$(nproc) will instruct make to use all available cores for the build.

5. After a successful build, find the binaries in BUILD/_bin/ or BUILD/vst/ and copy to system VST folder

   VST installation folders:

   • MacOSX: /Library/Audio/Plug-Ins/VST
   • Windows: C:\Program Files\Steinberg\VstPlugins, C:\Program Files\Common Files\VST2, or C:\Program Files\VSTPlugins
   • Linux: ~/.vst/, /usr/lib/lxvst or /usr/local/lib/lxvst

• NUM_CHANNELS - Number of input/output channels for each plugin. Default is 36.
• MAX_DELAYTIME_S - Maximum delay time for mcfx_delay in seconds. Default is 0.5s.
• BUILD_STANDALONE - Build standalone applications. Default is OFF.
• BUILD_VST - Build VST2 plugins. Default is ON.
• BUILD_LV2 - Build LV2 plugins. Default is ON.
• VST2SDKPATH - Path to the VST2 SDK. Default is "~/SDKs/vstsdk2.4"
• WITH_ZITA_CONVOLVER - Build with zita-convolver (better performance under Linux). Default is OFF.

• FFTW3F_LIBRARY - Path to the FFTW3F library file (Note the final f!).
  • On Linux this should point to the .so shared library file, e.g., /usr/lib/x86_64-linux-gnu/libfftw3f-3.so.
  • On Windows with Visual Studio, this should point to the .lib file generated in a previous step, e.g., C:/Program Files/fftw-3.3.6-pl2-dll64/libfftw3f-3.lib.
  • On Windows with MinGW, this should point to the .dll dynamic library file, e.g., C:/Program Files/fftw-3.3.6-pl2-dll64/libfftw3f-3.dll.
• FFTW3F_INCLUDE_DIR - Path to the FFTW3F include directory, e.g., /usr/include.
• [FFTW3F_THREADS_LIBRARY] - If present (e.g., on Linux), this should point to the .so file for the FFTW3F threads library, e.g., /usr/lib/x86_64-linux-gnu/libfftw3f_threads.so.

multichannel convolution matrix

loads configuration files (compatible to jconvolver .conf files)

supports loading .wav files directly (optionally reading input channel metadata)

just drag/drop a .conf or .wav file into the GUI to load it

have a look at CONVOLVER_CONFIG_HOWTO.txt for details about the configuration files

check the MATLAB folder for simple export scripts

searches for configuration file in following folders:

• Windows 7,8: C:\Users\username\AppData\Roaming\mcfx\convolver_presets\
• MacOS: ~/Library/mcfx/convolver_presets/
• Linux: ~/mcfx/convolver_presets/

delay each channel about the same time (maximum delay time in seconds is a compile time flag (default 0.5s): MAX_DELAYTIME_S)

filter each channel with the same low/high cut, peak filter and high/low shelf filter settings, frequency analyzer that displays a sum of all channels

low and high pass: 2nd order butterworth filter or 2x 2nd order butterworth cascaded (resulting in linkwitz riley characteristic) for use as x-over network

plus 2x parametric filter +- 18dB

plus low and high shelf filters +- 18dB

filter parameters can be adjusted during playback without introducing audible glitches (except switching lp/hp bypass and order)

set different delay time and gain setting for each channel (good for multispeaker calibration), includes a signal generator for testing individual channels

the GUI allows to paste a list of float gain and delay values from the clipboard with semicolon, comma, newline, tab or space separated.

maximum delay time in seconds is a compile time flag (MAX_DELAYTIME_S)

multichannel level meter with RMS, peak and peak hold

• 0.6.3 (2023-12-21) - mcfx_filter: fix parameter smoothing to avoid instabilities and glitches while changing filter parameters, performance optimizations.

• 0.6.2 (2023-12-11) - add 128 channel version of all plugins - since REAPER does allow for 128 channels per track, adjust mcfx_meter and mcfx_gain_delay GUI to display 128 channels properly

• 0.6.1 (2023-12-08) - mcfx_convolver: support loading .wav files (a .conf file will be written to disk and loaded in the background!), support drag/drop .wav or .conf files onto the GUI for loading

• 0.6.0 (2023-04-16) - new builds optimized for Apple Silicon and 64 bit Intel Mac; Win 64 bit; Update to JUCE 7; removed soxr dependency to simplify build; mcfx_gain_delay: allow sine generator to start at f=10 Hz

• 0.5.11 (2020-05-20) - mcfx_convolver - Mac OS version added a +6dB gain to the filtered output, this is fixed now (Windows version was correct) -> this might influence old projects under OSX since mcfx_convolver will output 6dB less than older versions!

• 0.5.10 (2020-05-19) - mcfx_filter - High-Shelf Q was not stored in the plugin state, this is fixed now

• 0.5.9 (2020-02-05) - mcfx_convolver - fix dropouts/artifacts for hosts that send incomplete block sizes (eg. Adobe, Steinberg), fix reloading of stored presets, add filter length and latency debug messages, fix gui crash in Adobe hosts

• 0.5.8 (2020-01-31) - mcfx_convolver - option to store preset within the project -> allows to exchange a DAW (eg. Reaper) project without need to provide the preset files extra, allow to export stored preset as .zip file for recovering it from the project

• 0.5.7 (2019-04-28) - mcfx_convolver - osc receive support: /reload, /load <preset.conf> -> allows remote control of reloading/loading presets, port can be set in GUI

• 0.5.6 (2019-03-20) - mcfx_convolver - maintain fir filter gain if resampled, add plugin parameter to trigger reload of configuration

• 0.5.5 (2018-03-16) - filter, gain_delay, delay: slider behavior changed for more accurate control; gain_delay: ctrl+click for exclusive solo/phase/mute, add toneburst for signalgenerator, bugfix saving channel state of signalgenerator

• 0.5.4 (2017-05-20) - mcfx_convolver and mcfx_filter: fixed threadsafety to avoid startup crash if other plugins use fftw

• 0.5.3 (2017-05-02) - mcfx_delay and mcfx_gain_delay fixed glitch in delayline

• 0.5.2 (2017-03-20) - various bugfixes; mcfx_convolver: performance optimizations, adjustable maximum partition size

• 0.5.1 (2016-04-25) - mcfx_convolver: fixed bug in loading packed (dense) matrix; mcfx_gain_delay gui fix

• 0.5.0 (2016-04-08) - add signal generator to mcfx_gain_delay; convolver: support for packed wav file to load a dense FIR matrix from only one .wav file -> have a look at CONVOLVER_CONFIG_HOWTO.txt; filter: smooth iir filter to avoid clicks when parameters change

• 0.4.2 (2016-02-19) - fixed one more convolver bug

• 0.4.1 (2016-02-17) - fixed convolver bug which resulted in mixed up partitions

• 0.4.0 (2015-11-04) - gui for mcfx_delay, gui for mcfx_filter (with fft analyzer), added phase, solo and mute buttons to mcfx_gain_delay

• 0.3.3 (2015-07-19) - performance improvements for mcfx_convolver _

• 0.3.2 (2014-12-28) - audiomulch compatibility, gui for mcfx_gain_delay with paste from clipboard functionality, mcfx_meter added scale offset

• 0.3.1 (2014-06-16) - fixed vst id for bidule compatibility

• 0.3 (2014-03-15) - added mcfx_convolver

• 0.2 (2014-02-25) - removed some license incompatible code, juce update

• 0.1 (2014-01-10) - first release

(C) 2013-2017 Matthias Kronlachner m.kronlachner@gmail.com