Skip to main content
European Commission logo print header
Contenido archivado el 2022-12-23

Digital audio effects

Objetivo

A.Background

"Digital audio effects" is the name chosen for this COST Action and covers the use of digital processing, its application to sounds, and its musical use designed to put effects on a sound. It is sometimes called Audio-FX, a name based on the English term for effects.

Sound processing consists in taking one or more sound sources and producing another sound. Digital signal processing can be applied to sounds viewed as signals, and computer science has provided many algorithms for modifying signals.

It has been widely used in both pop music and contemporary music where the terms "effects processors" or "sound processing effects" have been used. Electronic engineers have developed devices for the transformation of sounds and computer music uses techniques to modify or enhance a sound.

Thus the project "digital audio effects" offers a wide palette for research. Many problems still remain to unify this field and industry is searching for appropriate algorithms and implementations.

B.Objectives and benefits

The main objective of this COST Action "digital audio effects" is to provide a synthesis of what can be done in the digital processing of sounds, and its application to music. It is intended to collect transformations of sounds in the form of computer algorithms and sound examples resulting from these transformations.

Objectives

The objectives can be described as twofold:

(1)to compare the different methods of the European teams in terms of algorithms, implementations and musical use. This comparison applies to domains such as:

-the mathematical aspect of audio effects: what kind of digital processing algorithms is used, (example techniques such as linear prediction coding, real or complex cepstrum, spectral line estimations can be used to change the "spectral content" of a sound);

-the implementation aspect: because of the cost of equipment, it is rarely possible to compare implementations such as programming languages and DSP cards. This is a value added by this COST Action;

-the musical and perceptual aspect: each team partner will use sound transformations which imply aesthetic choices. For example, pop music, contemporary music, concrete music will use the same effects in different configurations, and the sonic and musical result will be different. Thus this COST Action provides a good means of comparing different views which can be complementary or even opposite in terms of musical preference.

(2)to bring together the knowledge of different European teams in the domain of digital sound processing, here designated as "digital audio effects", in a form which can be made available inside and outside of the teams themselves.

The material which can be brought together are:

-the mathematical description and the historical documentation,

-the implementation of algorithms in terms of a general language such as Matlab, and

-the collection of sound examples in the form of digital sounds.

Benefits

The benefits are manifold:

-the teams involved in the COST Action can themselves benefit from this action: they can exchange information, allow mutual visits of researchers, and also learn how to provide a better framework for their research (for example by not duplicating work);

-education can be stimulated by the use of the knowledge resulting from this COST Action. In particular it could be possible to follow this Action by the writing of articles and a book. The use of a general mathematical computer language such as Matlab can be really unifying in this domain;

-European industry can benefit from this research. Many audio effects are still implemented with analog electronics and it would be a real improvement for them to have know-how in digital implementations. This action does not focus on any industrial application, but it can benefit current innovative fields such as multimedia, virtual reality or professionnal audio.

C.Scientific programme

The scientific approach can be described in two ways: "audio-effects-taxonomy" and "audio-effects-implementation"

Audio-effects-Taxonomy (classification of the effects):

The aim of this COST Action will be to provide a practical classification of the effects. Over one hundred effects are commercially or academically available but some of them belong to the same class of modification and have the same algorithm with just different control parameters. European research teams have devised special digital audio effects and some of the publications denote their competence in this field. Although some of these effects such as reverberation have been used for a long time, some interesting new developments make them a very new "effect". Some other domains are the consequence of active research; for example, transformations based upon the separation between signal and musical noise or the extraction of the constituent formant frequencies of a vocal sound.

An example of the divisions suggested might be, for instance:

-time domain transformations and micro-surgery (inverting, shuffling ...)

-modulations of amplitude, frequency and phase
(vibrato, tremolo, phasing, flanging, compression-expansion, waveshaping, ...)

-reverberations and 3D sound

-filtering and coloration (equalizing, timbre modification)

-time/frequency domain transformations, vocoders
(speed change, pitch shifting harmonisers)

-sound morphing (cross synthesis, hybridizations)

-pitch and amplitude tracking

Audio effects - Implementation

For each effect, the following can be provided:

-documentation on each effect which can be made available on an Internet web server. It will for example give the history of the effect with references in the literature, a formal description (physicist's point of view), a musical description, and an overview and explanation of the techniques with which this effect can be achieved.

-an off-line implementation in the Matlab language with graphical support which can eventually be gathered to form a book and the description of the control parameters of the method.

-comparisons between real-time implementations whenever possible, and sound and musical examples which can be the data base for a possible CD-Rom.

It must be noted that this domain can be seen from two different points of view: the strict scientific point of view, which can lead to conferences, articles, books and so on, and also can stimulate or be stimulated by national networks.

But it can also be viewed from the musical educational system whereby the learning and the application of these techniques to sounds is a real challenge of the end of this century.

D.Organization and timetable

D1Organization

Due to the small size and strong interdisciplinarity of this action, there will be only one working group. The coordination of the action will be held by a management Committee which will decide at each step if there is a further need for a subdivision. This Comittee will organize two seminars a year (see D2) and decide on the short term missions. It will also take care of the good diffusion of the results and of possible contacts with other groups.

The results from this COST Action can be gathered in three forms:

an information data base

This information can be ideally shared on an Internet site, for example a FTP site (file transfer protocol) and a Web server. It may include:

-the bibliography (theses and articles in all the native languages) and the references to all sites that provide information, especially on hardware and software implementations already available.

-the current work in progress. For example, the Matlab files that describe each effect can be exchanged in this way so that every member of the group can experiment.

-other information, for example a general description of each team's work for the outside world.

an algorithm data base

It will contain both the signal processing aspect and its "off line" implementation in the form of computer programs using preferably a general language such as Matlab. Matlab is used as a simulation tool by most teams. It is simple to learn to use and provides the fastest way of evaluating an algorithm.

a sound examples data base

At each step, sound examples will be gathered illustrating the transformations of sounds with these audio effects. It is also possible to have a unified base of sounds before their processing in order to compare different methods.

CDs and CD Rom are the ideal support to keep track and make available digital sounds. Also depending upon the results obtained in this Action, it is reasonable to think of a distribution of the algorithms in the form of a book including a diskette.

D2Timetable

This action will last 4 years. It is planned to have two seminars a year. One of them is smaller and will be more oriented on the sharing of the work of the teams while the second one is more open to a wider audience. Each seminar will be preceded or followed by a management meeting.

These seminars will concern audio effects and their application to sound and music. It makes sense to divide these effects in four categories which can be studied successively but with only one working group:

-time effects (1),

-filtering and colouring (2),

-spatialisation and reverberation (3),

-morphing mimicking and tracking (4).

For each theme, each team will follow the methodological approach (see chapter C) and will:

(a)establish and categorize the list of audio effects that have already been worked on or evaluated by each team or other well known sound transformation techniques;

(b)provide information about each effect in terms of implementation, control, graphic and musical examples, and perceptual attributes.

E.Economic dimension

This Action implies that the different teams will have to support different costs:

-the personnel cost,

-the instruments necessary for the research (computers, sound interfaces),

-telecommunications cost for the use of Internet dial up connections,

-technical, software and maintenance costs of the experimental and computer facilities. In particular, sonic equipment has to be maintained and electronic sites have to be created and maintained,

-computer costs at exterior academic facilities,

-translation fees.

The following COST countries have actively participated in the preparation of the Action or otherwise indicated their interest:

France

Germany

Italy

Spain

Norway

Czech Republic

These national costs have been evaluated at ECU 300 000 a year by adding the estimation of the six countries having already responded to the proposal. So the overal cost for national costs for 4 years is: ECU 1,2 million.

Accordingly, the coordination costs paid by the Commission for four years are estimated to

60 000 x 4 = ECU 240 000 for the seminars and workshops

11 500 x 4 = ECU 46 000 for the short term missions.

Tema(s)

Data not available

Convocatoria de propuestas

Data not available

Régimen de financiación

Data not available

Coordinador

N/A
Aportación de la UE
Sin datos
Dirección


Francia

Ver en el mapa

Coste total
Sin datos