FR3040253B1 - METHOD FOR MEASURING PHRTF FILTERS OF AN AUDITOR, CABIN FOR IMPLEMENTING THE METHOD, AND METHODS FOR RESULTING IN RESTITUTION OF A PERSONALIZED MULTICANAL AUDIO BAND - Google Patents

Abstract

The invention lies in the development of a method for measuring PHRTF filters ("Personalized Head Response Transfer Function"), a listener and its application in a mobile device or not allowing the restitution of multichannel soundtrack in a helmet. The measurement method is implemented in a closed volume, acoustically processed and sound isolated, and comprising at least 5 loudspeakers, broadcasting an acoustic signal, all directed exactly to the head of the listener which is provided with a transducer electro-acoustic (microphone type) per ear, and in that one records by these microphones the parameters for calculating PHRTF filters. The method is characterized in that it is purely and solely acoustic. PHRTF filters are implemented and used in an application, either in a server, a terminal, a dongle or directly in a helmet of the listener by application of the convolution algorithm. The result is an individual soundtrack that renders a 360 ° spatialization effect.

Description

PHRTF filter measurement method of a listener, cabin for the implementation of the method, and methods for achieving the rendering of a personalized multichannel soundtrack

Technical sector of the invention:

The present invention relates to the technical field of measuring the acoustic behavior of an individual in a sound environment in order to integrate its unique morphological features into a sound processing method, particularly for nomadic personal listening, in particular to headphones.

Each individual is unique and perceives sounds in a modified way according to its morphological characteristics (shape of the head, shape of the flag of the ears, auditory canal, texture of the skin, nature of the hair, etc ...). A sound will not be perceived identically by two different individuals. An individual placed in a sound field perceives in three dimensions the different sources of a sound scene with their information of position, sound level, extent, distance and displacement.

An individual therefore himself causes a certain number of propagation disturbances of the sound wave that he will receive according to his own morphological characteristics. It is these disturbances that make it possible to locate the source of emission, the source, the nature, and the associated object of a sound in space. Each individual therefore has, for each of his ears, a unique transfer function (that is to say a single filter) which defines the set of disturbances. This unique filter is called "HRTF" (for "Head Related Transfer Function").

To evaluate an individual's HRTF filters, a loudspeaker is moved angularly (azimuth angle and elevation angle) around the center of the head, with a measurement microphone inside each head. flag of both ears. The microphones measure the disturbances of the sound waves emitted by the speaker in the different directions of propagation. The measurements made to obtain the HRTF filters of this individual.

New techniques, including binaural listening (related to listening with two ears), provide additional sound information in relation to stereophony, including much more spatial perception information. Where the morphological peculiarities of individuals are of little importance in stereophony (if we send the same sound signals to several individuals, they will perceive only very slight changes because the stereophonic sound sends only certain information spatial, those which are frontal), the spatial rendering obtained thanks to the measurements of the HRTF filters is much more important, and this in an audible way, during a binaural listening).

To enjoy a better reproduction of a spatial sound field, it is necessary either to place the individual in the center of the sound field (so that he can identify where each different sound comes from) using a large number of speakers to recreate a 360 ° environment; or, through binaural listening, use HRTF filters that allow the speakers placed at the ears (usually a headset) to send modified sound waves that will give the individual that feeling of spatialization.

Prior art:

For many years, home theater systems have been known to place loudspeakers in different parts of a room to create a certain spatial reproduction of the sound. These systems are obviously not nomadic (minimum two speakers, plus an amplifier). The generally adopted home theater systems are equipped with six speakers that reproduce the sensation of a sound field at 360 ° since in this configuration 2 HP are placed at the back. In more modern versions of home theater systems with 8, 10 or 12 speakers, the rendering is improved; and of course the more loudspeakers are placed, the closer the rendering approaches to a real 360 ° rendering. The problem is obviously the cost, a home theater system with 6 or 10 speakers is relatively expensive, and also requires to be arranged in a room and therefore occupies a lot of space. These home-cinema devices are not nomadic any more.

The general operation of a home theater system is to record different sound tracks which each will be sent back to the corresponding speaker taking into account the central listening position of the individual (commonly called "sweet spot").

Until now, headphone reproduction techniques have only been used on two stereo channels (sound reproduction aimed at recreating the illusion of a sound space, usually using two channels (left and right). diffused by at least two transducers (speakers or earphones)).

The current new techniques of binaural listening, allow to bring additional sound information in relation to the stereophony by allowing to reproduce the sound fields diffused in particular by home-cinema systems, as regards the information of spatial perception and it takes significantly higher proportions with HRTF filters. The advantage of binaural spatialized listening is that it can broadcast a sound field 360 ° (previously recorded) and retranscribe to at least two speakers using the HRTF filter measurement technique. The major disadvantage of this technology is that it is standardized: the spatial sound rendering systems operate on the basis of a set of "standard" binaural filters, derived from an acoustic measurement on an artificial bust or one or more standard morphology users. As noted above, perceptual acoustic location indices are not at all general for all listeners but very specific and therefore unique for each individual, resulting in a number of location errors, including increased confusion. in back.

Thus, sound rendering systems based on standard binaural filters are unreliable for a majority of listeners.

As noted, another possibility for recreating spatialization is to measure an individual's unique HRTF filters and use these filters to process sound waves so that spatialization data can be reproduced from only two high speakers placed at the level of the ears. Thus, we take into account the unique sound deformation of each individual rather than a standard deformation.

The problem is that the method of measuring individual HRTF filters (for each ear) is extremely tedious and lasts several hours. The individual must remain stationary for this duration and must support a helmet equipped with electroacoustic transducers for each ear.

France Telecom Patent No. 03 02467 is known which makes it possible to simulate the unique HRTF filters of an individual by greatly limiting the immobilization time of the individual during the measurement. This patent comes out of the acoustic-acoustic approach by proposing a process that uses photos of the individual's head. After taking pictures, a software program will calculate the HRTF filter based on parameters embedded in the morphology analysis software. HRTF filters are not directly measured but estimated algorithmically from these photos. Subsequently, the information is transmitted to a server that will modify the multichannel data broadcast taking into account the HRTF filters to send them to at least two speakers.

This solution has very important drawbacks: - not to be faithful to the acoustic reality (mathematical estimation from photographs), and - to omit certain characteristics disturbing the acoustic field (non-visible elements, shadow zones, nature of hair, skin texture especially).

The quality of the rendering absolutely can not equal the rendering of the acoustic - acoustic measurement. The measure is therefore at the expense of its quality.

Home theater-type devices are also known which are capable of sending multichannel sound to a headset. The individual then has the impression of having a spatialized rendering in a helmet. These devices have the following drawbacks: - absolutely no nomad: the headset can only work properly near the home theater system that was used to measure H RTF filters. - no customization depending on the individual HRTF filter. Whatever the individual who uses the headphones, the sound output of the headphones will be the same In general, many systems process sound before it is attached to the media (CD, DVD, Blu-Ray, servers, etc. ) on different 5.1, 7.1 or stereophonic channels and the listener can not act to adapt this sound to his or her own hearing parameters.

Technical problem posed:

There is therefore a real need for a method of rapid realization of the HRTF filters unique for an individual that is performed only acoustically to be totally faithful to the reality of acoustic diffusion, as well as for a method for using these filters for nomadic devices (headsets), all to restore a 360 ° spatialization effect equivalent to that rendered by a speaker multicast system.

As indicated above, for a spatial reproduction, we currently use several channels (and therefore several speakers that are naturally unsuitable for a reproduction in headphones, so non-nomadic). For headphone reproduction it is therefore necessary to define and characterize the behavior of our ears subjected to auditory stimuli according to different directions called "360 °" since it is a configuration used with several loudspeakers (6.8 , 12 ...) to reproduce soundtracks of audiovisual content in these same formats (5.1.7.1, ...). SUMMARY OF THE INVENTION: The invention resides in obtaining customized HRTF filters for binaural synthesis of headphone sound 360 °, referred to herein as PHRTF ("Personalized Head Response Transfer Function", "Transfer Linked Function"). to the Custom Head ").

This restitution is optimized and individualized taking into account the morphological characteristics of each individual.

The present invention also relates to a method of storing the filters in a persistent memory, online or local (in the terminal or directly in the headset)

The present invention also relates to a method of restitution in the helmet of the soundtrack adapted to the individual by calculating, in the helmet itself, the resulting soundtrack.

TECHNICAL SOLUTION The invention therefore lies in the development of a method for measuring PHRTF filters of a listener, characterized in that said method is implemented in a closed volume, treated acoustically and sound-isolated from the atmosphere sound outside said volume, and said closed volume comprises at least 5 loudspeakers (without subwoofer), broadcasting an acoustic signal, all directed to the focal point or convergence point located in its center of volume, at the exact location where find the head of the listener who is provided with an electro-acoustic transducer allowing the reception of acoustic signals (of microphone type, or other recording apparatus) by ear, and in that one records and / or transfers from these microphones, to a remote recorder such as a PC, or other digital system, the parameters for calculating the PH RTF filters, and characterized in that it is performed without taking a photograph of the head of the listener

The PHRTF individual filter measurement process is done in a specially designed cabin to be as acoustically coherent as possible. Only the sound deformation related to the individual must exist.

The enclosed volume is preferably a cabin which is substantially a rectangular parallelepiped whose height must be greater than the width so as to correspond to the characteristics defined according to the ITU R-BS775 1 and 2 standard (recommendations concerning the multichannel stereo sound system with or without associated image), and should also have an acoustic insulation to attenuate the sounds coming from the outside so as not to distort the measurement. The appropriate and commonly agreed minimum attenuation level should be between 40 dB and 50 dB

By "cabin" is meant here a closed enclosure on all sides and whose function is to contain said closed volume and ensure sound insulation and acoustic treatment, which implies the choice of obvious materials for the skilled person.

Of course other forms of cabins may be considered.

This can be removable or fixed.

In a totally preferred manner, in the case of a demountable cabin it measures 2400x1800x2100 (in mm) in order to be closer to the standards defined in the document ITU R-BS775 1

To carry out the measurements in the cabin, at least 5 loudspeakers (necessary for a 5.0 audio broadcast) are placed. Naturally, it is possible to provide more speakers if desired: 6 for broadcast in 6.0, 7 for broadcast in 7.0.

We do not place a subwoofer (dedicated to the reproduction of "serious effects"), because the wavelength of the low frequencies is too great to be relevant in the measurement of a PHRTF filter, and therefore unusable in a measuring cabin.

The arrangement of the loudspeakers is according to the ITU R-BS775 Standard 1 as illustrated in Figures 1 to 3.

In these schematic figures in top view, the listener (1) is placed in the center of the cabin and at least 5 loudspeakers (HP) are placed at the periphery of a circle of approximately 2m radius. Three loudspeakers are positioned in front of the listener (substantially facing him): one strictly opposite him and one on each side positioned at an angle of about 30 °. The two rear speakers (relative to the listener) are positioned at an angle of about 120 ° to the speaker opposite the listener. In the versions shown in Figures 2 and 3, the positions of the 3 front speakers remain exactly the same. In version 6.0, a loudspeaker is added at 180 ° to the one facing the listener (figure 2). In version 7.0, the position of the loudspeakers is changed: two loudspeakers located back left respectively having an angle of about 100 ° and 140 ° relative to the speaker in front (1) of the listener; two loudspeakers located at the right rear respectively having an angle of about 100 ° and 140 ° relative to the loudspeaker opposite (1) of the listener; The listener carries two microphones (one per ear) that are calibrated to perform accurate measurements.

According to a preferred embodiment of the cabin (6) (Figure 4), the cabin comprises an acoustic window (2) in front of the listener so that the operator (3) can be visible to the listener (1) . An acoustic door (5) is located behind the listener (1), opposite the acoustic glass (2).

The measuring system (7) is located outside the cabin (6) and is placed close to the operator (3). The listener (1) carries microphones (8).

The walls inside the cabin are designed in such a way as to sufficiently mitigate reflections that interfere with the localization of the walls of the sounds coming from the loudspeakers. This also makes it possible to attenuate the "room effect" (or cocktail effect) and thus to avoid an excessive modification of the diffusion of the measurement signals.

The cabin is equipped with acoustics composed of acoustical materials placed in a manner that respects the rules of audio cabin design acoustics in accordance with the provisions indicated above. For example the first reflections will be attenuated by the placement of lateral and upper diffusers (suppress the fusion effect), as for the serious stationary frequencies in the angles which cause an increase the undesirable acoustic energy in these frequencies, they will be attenuated by absorbers positioned in the front, rear and upper lateral angles.

This allows a measurement of the individual "unmodified" PHRTF filter by the cabin itself. Once the listener is installed and the measuring microphones are positioned in his ear canals, the measurement can begin. Each loudspeaker broadcasts an acoustic signal of constant or frequently weighted level according to a continuous frequency sweep from 20,000 Hz to at least 20 Hz and following at least 5 different angles, of which at least 2 at the back of the individual (which correspond to the 5 positions of Figure 1). At each diffusion, the frequency response resulting from the diffusion that is perceived at the level of the microphones is measured simultaneously for each ear. At least ten measurements (two per ear and per diffusion chamber) of acoustic disturbances are obtained directly on the individual.

From these measurements, the number of filters corresponding to the individual (two per ear and per diffusion chamber) is generated. In the case of 5 loudspeakers, we will obtain 10 individual PHRTF filters.

For each of the scattered signals, therefore, the difference between the transmitted signal (known) and the measured signal (perturbation) is obtained by a single and individual filter PHRTF which is the exact translation of the acoustic deformation produced by the individual in the space of diffusion.

FIG. 5 is an example of a frequency representation of the individual PHRTF filters for the left ear and for the right ear.

Figure 6 is an example of time and frequency representation of different individual PHRTF filters.

The overall result of the entire measurement is a set of left and right filters according to each speaker position, ie for a signal 5.0.10 individual PHRTF filters: lg, ld, 2g, 2d, 3g, 3d, 4g, 4d , 5g, 5g.

With convention: l = left hp, 2 = right hp, 3 = central hp, 4 = left rear hp, 5 = right rear hp.

Then g = filter for the left ear and d = filter for the right ear.

The reproduction of a 5.1 sound signal (6 discrete audio channels) therefore has 6 signals. It is therefore necessary to apply for each signal out of LFE ("low-frequency effects") and for each ear the corresponding measured PHRTF filter and reinject in each left and right final signal the signal of the LFE (attenuated by 10 dB in accordance with the recommendations of adjustment of home- to preserve a comfortable bass level in the final signal (attenuated by an additional 3 dB to maintain the overall equivalent level):

For each signal of each channel: left_binaural_ signal = (sum (convolution (signal_canal_i, hrtf_gauche)) + (Signal LFe-13db)) signal_binaural_droit = (sum (convolution (signal_canal_i, hrtf_droit)) + (Signal LFE-13db))

Once the set of PHRTF measurements for an individual is obtained, a convolution algorithm, known to those skilled in the art, is used to transform the at least 5 sound signals from the loudspeakers to two sound signals for both. headphones speakers.

An example of the transformation is given in FIG. 7, in the case of a configuration in 5.1 (6 separate and non-matrixed audio channels). The above algorithm thus makes it possible to customize a multichannel soundtrack according to the specific morphological characteristics. of each listener.

It is possible to combine the individual PHRTF filters with filters specific to a given concert hall, such as, for example, the Paris Opera (Bastille or Garnier) or the MET in New York. In case, the individual PHRTF filters of the listener are corrected by addition (positive or negative, with a degree of correction adaptable according to the influence that one wants to give to the acoustics of the desired room on the final soundtrack.

It is also possible to combine the individual PHRTF filters with predefined effects such as those already present in conventional home cinema systems (for example: film, football, standard, opera, music ...) · The correction algorithm is the same as the one described above.

In both cases, the convolution algorithm remains the same, since it applies a unique set of individual PHRTF filters to the original soundtrack.

The present invention also relates to a method for storing PHRTF filters in a persistent memory, online or local (in the terminal or directly in the headset)

The PHRTF filters can be stored in a server containing a PHRTF filter database for, for example, the clients of an audiovisual content distribution service.

Alternatively, the storage of the PHRTF filters can also be done in a generic memory in the terminal or in the headphones or in a dongle connected to the terminal, capable of storing these filters in a persistent but erasable and rewritable manner or any other semantic storage memory. permanent (for example, a flash memory or a micro-SD card). This makes it possible to defer the generation of the personalized soundtrack at the listener level and thus to send to the listeners only the original soundtrack, which greatly simplifies the implementation on the server side.

The present invention also relates to a method of restitution in the helmet of the soundtrack adapted to the individual by calculating, in the helmet itself, the resulting soundtrack.

The processing, by application of the algorithm described above, three-dimensional sound data is done according to the different scenarios described below.

Disturbance measurement is used to adapt for each individual any single original soundtrack (in 1.0, 2.0, 4.0, 5.0, 5.1, 6.1, 7.1, and above). This adaptation is done in real time by keeping the electronic data as input, the same soundtrack for all users.

The present invention also relates to a method of restitution in the headset of the soundtrack adapted to the individual by calculation, in the headset itself, of the resulting soundtrack

There are several options for using individual PHRTF filters for application to headphones.

The preferred implementation mode developed according to the present invention is the individualization of the original soundtrack is performed directly in the headphone of the listener. See diagram shown in Figure 8.

The computer server containing the original media content distributes the same original content to all listeners. The terminal (nomadic or not) of the listener receives the original soundtrack from the server, possibly stores it locally and sends it to the listener's headset when the listener wishes to access the content. The terminal does not change the original soundtrack.

The individual PHRTF filters are stored in the listener's headphones and it is this equipment that calculates, in real time, the personalized soundtrack, before broadcast in the speakers of the headphones. The individual PHRTF filters can be stored as in a terminal, in a fixed or removable permanent memory of the SD or micro-SD card type (re-writable or not) with a specific slot accessible from the outside by the listener. This slot allows you to customize the headset for each listener, by selecting the right PHRTFs filters stored in the headset or by inserting the removable memory card containing the right filters.

The convolution is calculated in the helmet, which implies that the helmet has the ability to integrate the electronic device for this purpose. This imposes design constraints. For in-ear earpieces, the electronic device must integrate directly on the cable (like the currently existing remotes). This has the consequence that this method and system works with all headphones capable of directly receiving a digital signal (in this case a dematriate 5.1 digital audio signal). Other possible modes of implementation are described below:

Variation 1: The individualization of the original soundtrack is performed in a computer server before distribution of the multimedia content in the terminal. See diagram shown in Figure 9.

In this mode, the individual PHRTF filters are stored in the computer server containing the original media content.

The personalized soundtrack is produced in this server at the moment when the listener wishes to access the content for immediate listening ("streaming") or for storage in his terminal (PC or mobile device).

The terminal (nomadic or not) receives the individualized soundtrack, stores it locally if necessary and sends it as it is, without further manipulation to the listener's headphones.

The convolution takes place in the server, the terminal receives a digital signal which is converted into analog in the reader and sent as is to the headphones. The headset therefore receives this analog signal (which represents the original soundtrack individualized for the listener), and there is no new constraint on the headphones. This has the consequence that this method and system work with all existing helmets.

Variant 2: The individualisation of the tape are original is performed in the mobile terminal of the listener. See diagram shown in Figure 10.

In this version, the individual PHRTF filters are stored in the listener's terminal after a prior download procedure. Note that the same terminal can store several individual PHRTF filters corresponding to different listeners. The choice of the right PHRTF filters will then be done by a software offering this selection through an appropriate human-machine interface.

The computer server containing the original media content distributes the same original content to all listeners.

The terminal (nomadic or not) calculates the personalized soundtrack directly into the player when the listener wants to access the content, then sends the personalized soundtrack to the listener's headphones.

The convolution takes place in the terminal. As in version 1 above, the headset receives an analog signal (which represents the original soundtrack individualized for the listener), and there is no new constraint on the headset. This has the consequence that this method and system work with all existing helmets.

Variation 3: The individualisation of the band are original is performed in a dongle connected to a nomadic equipment or not the listener. See diagram shown in Figure 11.

This variant is an alternative to variant 2.

The computer server containing the original media content distributes the same original content to all listeners. The listener's terminal receives the original soundtrack from the server, and optionally stores it locally.

The individual PH RTF filters are stored in a "wireless" dongle that includes a connector (USB for example) that allows it to connect to an external port of the terminal, and paired or not with the listener's headphones .

In the headset either via a wired connection (type an analog audio output) or via a wireless connection (type Bluetooth, or wi-fi).

The terminal (nomadic or not) sends to its external port, the original soundtrack, which is personalized in real time by the dongle "wireless", which sends the personalized soundtrack to the headphone of the listener. The "wireless" dongle contains both the individual PHRTF filters, the electronics needed to execute the convolution algorithm.

Note that the "wireless" dongle can also connect to any digital 5.1 audio broadcast source (unregistered).

The server and the terminal have no information about the customization of the original soundtrack and therefore make no difference between a listener (and his headphones) that uses the individual PHRTF filters and a listener who uses the original soundtrack. The system for broadcasting the original soundtracks does not require any modification or adaptation and can therefore be used in a completely transparent manner for "standard" users and users using the method described here.

The convolution takes place in the "wireless" dongle and sends to the headphones an analog signal either by wired connection or by "wireless" connection, which contains the original soundtrack individualized for the listener. This version therefore works for "wireless" headphones able to connect according to "wireless" standards, but also those that have dual wired and "wireless" connectivity. This has the consequence that this method and system work with all existing "wireless" headsets.

Naturally, the present invention is not limited to the embodiment described above by way of example; but it extends to other variants, depending on the number of speakers for rendering the number of audio streams generated, without limitation (5.0, 5.1, 5.2, 7.1, 10.2, ...). The present invention is compatible with all mono, stereo, quadraphonic broadcast formats.

The present invention also relates to the adaptation of one sound diffusion format to another - such as mp3 "surround" or HEAAC ("High-Efficiency Advanced Audio Coding", which are not diffusable in a headset. the present invention is capable of transforming the soundtracks from each of these formats (to the extent that they are decoded beforehand) to make them broadcastable to the headphones and this in a completely personalized way.

The present invention is not limited to the distribution system of the multimedia content including the soundtrack. The invention is completely compatible with CD, Blu-Ray, Internet streaming or satellite or cable distribution formats or Video on Demand (VOD), Near Video on Demand (NVOD) or Television on demand.

The present invention is also completely compatible with compression algorithms used for storage or broadcasting of multimedia content.

The present invention is also compatible with all content protection technologies, DRM system or conditional access (such as Viacces, Nagra, Irdeto, Conax or NDS / Cisco)

The method works with all existing helmets in versions 1, 2 and 3. The nomadic device comprising the electronic device is selected from the non-exhaustive list of: telephones, tablet computers, laptops or other music storage devices, films , video, can be connected to a helmet.

It also adapts to the 3 types of existing helmets: closed, semi-open, in-ear and headband. The invention also covers all the embodiments and all the applications which will be directly accessible to the person skilled in the art upon reading the present application, and of his own knowledge.

Claims (18)

1. A method for measuring personalized PHRTF filters of a listener ("Personaiized Head Response Transfer Function"), characterized in that it comprises the following steps: a step of diffusion of an acoustic signal by between 5 and 7 high speakers arranged according to the ITU R-BS775 1 standard and all directed to the focal point of a closed volume treated acoustically and sound isolated from the sound environment external to said volume, said focal point being located at the center of volume of said closed volume; a step of simultaneous measurement for each ear of a head of a listener placed at said focal point of said closed volume of the frequency response resulting from 1a. broadcast, using a microphone per ear; and - a step of generating a set of customized PHRTF filters for said listener by a digital system from said measurements; and where: said process is performed without taking a photograph of the listener's head. 2. Method according to claim 1, characterized in that each speaker diffuses an acoustic signal of constant or frequently weighted level according to a continuous frequency sweep from 20000 Hz to 20 Hz minimum and following at least 5 angles defined by the position of the high Speakers. 3. Method according to any one of claims 1 and 2, characterized in that, at each diffusion of an acoustic signal, is measured simultaneously for each ear frequency response resulting from the diffusion that is perceived at the microphones, and characterized in that the difference between the acoustic signal emitted by the loudspeakers and that received by the two microphones is compared. 4. Method according to any one of claims 1 to 3, characterized in that from the measurements, two PHRTF filters are generated per ear and by diffusion loudspeaker, ie at least 10 PHRTF filters for at least 5 loudspeakers. speakers. 5. Cabin for implementing the method according to any one of claims 1 to 4, characterized in that it comprises between 5 and 7 loudspeakers arranged according to the ITU R-BS775 1 standard, in that it encloses said closed volume acoustically and acoustically isolated from the external environment and in that said cabin is substantially a rectangular parallelepiped whose height is greater than the width, and has walls allowing a level of attenuation the external sound environment is between 40dB and 50dB. 6. Cabin according to claim 5, characterized in that it can be removable or fixed. 7. Cabin according to any one of claims 5 to 6, characterized in that it measures 2100x1800x2400 (h x 1 x L) (in mm). 8. Process for converting from 5 to 7 sound signals from 5 to 7 loudspeakers to 2 listening loudspeakers using the PHRTF filters measured according to the method according to any one of claims 1 to 4, characterized in that that he uses a convolution algorithm. 9. Transformation method according to claim 8 characterized in that it comprises: an application step for each signal out of LFE and for each ear of the corresponding measured filter PHRTF; an injection step in each left and right final signal of the attenuated LFE signal of! 3dB is for each signal of each channel: - signa! left binaural :::: (sum (convolution (signal_canal_i, hrtf_gauehe)) + (Signal LFe-13dB)); - signal binaural_droit :::: (sum (convolution (signal_canal_i, hrtf_droit)) + (Signal LFE ~ 13dB)). 10. A method of transformation according to claim 8 or 9, characterized in that PHRTF filters are combined with specific acoustic transformation filters which can be including acoustic characteristics of a concert hall, or predefined effects: opera, rock , classic, music, etc 11. A method of storing in a persistent memory, online or local (server, PC or other) PHRTF filters obtained according to the measuring method according to claims 1 to 4, characterized in that the storage is performed in a server, or alternatively in a generic memory in a terminal or in the mobile device, or in a dongle connected to the terminal, or in an SD or micro-SD card. 12. Method of restitution in a nomadic device of the personalized soundtrack, modified by the transformation method according to one of claims 8 to 10 characterized in thatindividualisation of the original soundtrack is performed directly in the headphone of the listener . 13. Method for rendering in a nomadic device of the personalized soundtrack, modified by the method of transformation of claims 8 to 10 characterized in that the individualization of the original soundtrack is performed in a computer server before distribution of the multimedia content in a mobile terminal of the listener. 14. Method of restitution in a nomadic device of the personalized soundtrack, modified by the method of transformation of claims 8 to 10 characterized in that the individualization of the tape is original is performed in a mobile terminal of the listener. 15. Method of restitution in a nomadic device of the personalized soundtrack, modified by the method of transformation of claims 8 to 10 characterized in that Γ individualization of the tape are original is performed in a dongle connected to a listener equipment . 16. Nomadic device characterized in that it comprises at least two listening loudspeakers of a soundtrack and an electronic device for: storing the PHRTF filters obtained according to any one of claims 1 to 4; - Transformation and other filters according to claims 8 to 10 via PHRTF filters of the sound signals from 5 to 7 speakers to 2 listening speakers; and - the restitution of the individualized soundtrack. 17. Nomadic device according to claim 16, characterized in that it is a helmet, in particular of two high-wager type, one applied to each ear. 18. Mobile device according to claim 16 characterized in that it is a telephone, a tablet computer, a laptop or other device for storing music, movies, video, which can be connected to a computer. helmet.