Moreesc

Download

silva — 2010-02-27T10:36:51Z

Download

Moreesc is freely available under the terms of the Cecill license agreement

Release 2.2 - Py3 ready (October 20th, 2016)
This release fixes some bugs concerning the simulation for acoustic resonators with a single mode, and the estimation of the fundamental frequency (due to changes in Aubio). It also updates the embedded h5pickle module. Last but not least, Moreesc now supports Python3!

Source files archive
Python3 debian package
Python2 debian package

Release 2.1 (February 3th, 2013)

This release fixes some bugs concerning the plot of the results (due to changes from Matplotlib 1.11 to 1.12), allows to write wav files even without the scikits.audiolab package.

Source files archive

Release 2.0 (January 7th, 2013)

This release brings a brand new rewrite of the code allowing simulation of transients as slurs, vibratos, and others effects through the ability to define time varying acoustic resonator and reeds (cane reed or lips). Moreover it eases the interpolation between measured input impedances, and the use of measured pressure and/or lip stress via the Profiles objects. Note also that the valve is now not stuck to single d.o.f but may involve several resonances or even none!

Finally new ODE integrators are available allowing to choose the right tool for each case.

Source files archive

Release 1.7 - Buzzing version (April 24th, 2010)

Lip damping is now controllable too.

Source files archive
Windows Installer for Python 2.6

Release 1.6 - Buzzing version (February 27th, 2010)

Special version for vibrating lips!
In this release, valve has an outwards behaviour intended to model lip valve for brass instruments. Moreover, in order to mimic real playing, lip natural frequency (only one by now) is a time varying function.

Source files archive
Windows Installer for Python 2.6

Release 1.5 Not distributed (internal to the lab).

Release 1.4 - SingleReed version (June 18th, 2009)

Acoustic Resonator
Added functions to convert pole/residue couple to/from resonance frequency/quality factor/peak value tuple.
Performing the computation of the supplementary pole linked to the reed motion induced terminal condition is now available.
Modal decomposition of an experimental input impedance is now done using the procedure suggested by LeRoux & Herzog.
Physical constants (as the density, the wave speed, the visco-thermal boundary layers thicknesses) are now possibly adjusted according to the temperature.
Solver
The initial conditions corresponding to the static regime (even instable) are checked using AUTO (thanks to Sami Karkar)
Computation of the instantaneous frequency of the pressure signal.
Once an auto-oscillation has been calculated, it is possible to store the configuration and the results to restore it later.
Alpha stage : a method to prolongate a simulation on an extended time interval
Installation:
We are know able to install from sources or from the binary installer on windows platform.
Bug fixing:
Docstrings are now readable (unicode fixed)
Fixed minor bugs in the SimulationTemporelle methods (condition_initiales, grandeurs_controle).

Source files archive
Windows Installer for Python 2.5

About us

silva — 2009-10-28T14:09:39Z

About us

This software has been developed during the Ph.D of F. Silva in collaboration with Ch. Vergez, J. Kergomard and Ph. Guillemain. We all are working in the Physics of Musical Instruments research group in the Laboratory of Mechanics and Acoustics (UPR CNRS 7051), Marseille France.

The studies lead within this group are aimed at providing a better understanding of the musical instruments behaviour, particularly in theirs nonlinear aspects. Physical phenomena are investigated and modelled in order to be usable in sound synthesis, to develop original strategies of the control of the models (by, for example, the analyze of bifurcations diagrams), and/or to provide a help to instrument makers. The line of work used mixes both theoretical, numerical and experimental aspects.

The development of Moreesc has been possible with the support of the French National Research Agency ANR within the Consonnes and Cagima projects.

Documentation

silva — 2009-10-28T11:32:27Z

Documentation

Moreesc's documentation is available in PDF format:

An obsolete HTML documentation is still available at online.

An article has been recently submitted and can be view here. It contains a detailed description of the internals of the code and various examples allowing to have a deeper glance into the new features.

Some useful links
Numpy and Scipy docs
Matplotlib docs

Tutorial

silva — 2009-06-25T22:36:57Z

Getting Started

A minimal full example that should work:

We first need to import some modules

import sys
# Tell python where to find moreesc (if non-standard path)
sys.path.append('/your/path/to/the/folder/containing/the/moreesc/folder')
# Importing moreesc
import moreesc as mo
mp = mo.Profiles
mac = mo.AcousticResonator
mv = mo.Valve
ms = mo.Simulation
import numpy as np

Download

Then we build the acoustic resonator and the reed objects

# Bore and reed defition
D = mv.ReedDynamics(wr=2*np.pi*1500., qr=0.4, kr=8e8)
Ze = mac.Cylinder(L=.5, r=7e-3, radiates=False, nbmodes=10)

Download

Time-varying mouth pressure (with silence, attack, sustain and decay) and constant lip stress on the reed are defined with the Profil class

# Control parameters
tsim = 1.
S0 = mp.Constant(3e-6)
Pm = mp.SmoothStep([.001, .003, tsim-.2, tsim-.15], 1e3, 1e2)
# 2ms attack starting at 1ms, almost .8s sustain, 50ms decay

Download

Finally, the TimeDomainSimulation instance is created, and it can compute the auto-oscillations of the combination of the resonator and of the reed for such control parameters

# Creating the simulation object
sim = ms.TimeDomainSimulation(D, Ze, pm=Pm, h0=S0,
fs=44100, piecewise_constant=False)
# changing integrator and solving
sim.set_integrator('vode', nsteps=20000)
sim.integrate(t=tsim)

Download

When calculations were achieved, we can trace the results and save the simulation (in order to continue later).

sim.trace(trace_all=True)
sim.save('test.h5')

Download

Recording the synthetic sound and viewing the signals with:

sim.save_wav('test.wav', where='in')

For more specific help about the features, look for class and functions documentation, for example:

help(sim)

Presentation

silva — 2009-06-12T16:31:32Z

What is Moreesc ?

MOREESC is an application for the calculations of the auto-oscillations in wind instruments, designed originally for single reed instruments but it can be used for brass too. It takes advantage of a modal decomposition of the acoustic pressure field in the bore, or of its input impedance, to let you compute the sound of any strangely-shaped instrument.

Languages

MOREESC is written in the Python language. Compiled code is used for the numerical computation core, wheras Python is the high-level language allowing simple postprocessing and glue actions.

Availability

The package is distributed under the Cecill-C license and freely available for scientific use. MOREESC is a typical research program wich is provided "as is", with no guarantee whatsoever.

Contact

The code is developed and maintained by Fabrice Silva (silva[at]lma.cnrs-mrs.fr).

Installing: Python, compiler and MOREESC...

silva — 2009-06-11T12:25:17Z

The first requirement to use MOREESC is to have a Python installation with some useful packages (numpy, scipy, matplotlib...) and also C compilers. Once you have all the required tools, MOREESC installs itself in the classic Python directory, using basic machinery.

Installation

Please read the whole page before starting the installation...

Python

MOREESC is a open-source software distributed under the CECILL-C License and written in Python. This is an easy to learn, powerful programming language. It has efficient high-level data structures and a simple but effective approach to object-oriented programming. Python's elegant syntax and dynamic typing make it an ideal language for scripting and rapid application development in many areas on most platforms. The Python interpreter and the extensive standard library are freely available in source or binary form for all major platforms from the Python Web site, and may be freely distributed. The same site also contains distributions of and pointers to many free third party Python modules, programs and tools, and additional documentation.

Mac Users
Please pay attention that Mac OSX ships an old-released Python installation (usually in /System/Library/Frameworks/Python.framework with the interpreter in /usr/bin/python). Modifying these files is unsafe and may damage your system. You would rather install the updated binary from python.org (which goes to /Library/Frameworks/Python.framework — not in /System/... — with the interpreter in /usr/local/bin/python).

Mac Users

Please pay attention that Mac OSX ships an old-released Python installation (usually in /System/Library/Frameworks/Python.framework with the interpreter in /usr/bin/python). Modifying these files is unsafe and may damage your system. You would rather install the updated binary from python.org (which goes to /Library/Frameworks/Python.framework — not in /System/... — with the interpreter in /usr/local/bin/python).

Python scientific computing

MOREESC makes also an extensive use of NumPy and SciPy, which are the two major open-source packages for scientific computing that use the Python programming language. The instructions for installations are given on Scipy Web Site. Another useful tool is Matplotlib, a python 2D plotting library which produces publication quality figures in a variety of hardcopy formats and interactive environments.

The physics of the musical instruments is modelled through a set of Ordinary Differential Equations that are then written in typed Python (cython), and it thus requires a C compiler. Freely available compiler collections are GNU GCC on Unixes and MinGW on Windows. In comparison with pure python calculations, the use of the compiled ODE system fastens the computation by a huge factor.

Windows Users
PythonXY provides a scientific-oriented Python Distribution. Its purpose is to help scientific programmers used to interpreted languages (such as Matlab or IDL) or compiled languages (C/C++ or Fortran) to switch to Python. C/C++ or Fortran programmers should appreciate to reuse their code "as is" by wrapping it so it can be called directly from Python scripts. Checking NumPy, SciPy, Matplotlib and MinGW during installation process will insure a functional basis for MOREESC. Additionally, Spyder (offered within PythonXY if PyQtShell is checked) offers a Matlab-like integrated development environment.

Windows Users

PythonXY provides a scientific-oriented Python Distribution. Its purpose is to help scientific programmers used to interpreted languages (such as Matlab or IDL) or compiled languages (C/C++ or Fortran) to switch to Python. C/C++ or Fortran programmers should appreciate to reuse their code "as is" by wrapping it so it can be called directly from Python scripts. Checking NumPy, SciPy, Matplotlib and MinGW during installation process will insure a functional basis for MOREESC. Additionally, Spyder (offered within PythonXY if PyQtShell is checked) offers a Matlab-like integrated development environment.

Binaries

BInaries for Windows, Mac OSX, Red hat (.rpm) and Ubuntu (.deb) are provided on the Download Page. The installer detects your Python installation, and puts MOREESC in the default tree. If you are not familiar with building from sources, you are strongly advised to use those. If you intent to modify the source (not just run simulation), pay attention to the permissions you will have on the files (root or user installation), you may also prefer the source distribution with its development mode.

Source files

Extract the files from the archive to a temporary directory, and enter the newly created directory

The Python distutils mechanisms are then invoked to install the files
as root, in the usual site-packages directory

or as a normal user, in an arbitrary directory you will need to add to your path

or in the development mode which compiles the Cython extension in the local directory. Each time you make changes in the Cython files (.pyx), you only need to build them with:

It may be all that simple!