DE2414527A1 - DEVICE FOR HOLOGRAPHIC STORAGE AND REPLAY OF VOICE INFORMATION - Google Patents

Description

24H527

Nippon Electric Company, Limited 33-1, SMba Gochome, Minato-ku Tokyo / Japan

Device for the holographic storage and reproduction of voice information

The present invention relates to a device for the holographic storage and reproduction of speech information, which are converted into voiceprints.

With conventional, holographic speech synthesizers becomes a unit of language information, e.g. the expression "I want to go" (I will go) in different syllables, Subdivided syllables, phonemes or words, such as "I" (I), "will" (want), and "go" (go)). These vowel segments or parts of speech are PCM-coded (pulse code modulated) and stored in a magnetic drum memory Read out sequentially as required, with one speech information unit being reproduced at a time.

However, such language converters still have considerable disadvantages.

409841/0777

24U527

split up. For example, the number of in a magnetic Drum memory stored voice information on some a hundred limits the amount of language information required is considerably larger. If, for example, several thousand language information units have to be stored, thus, various magnetic drums are required, making the entire apparatus bulky and expensive. Since the in Information stored on a magnetic drum can only be read out sequentially and the speed of movement the magnetic drum is relatively small, it is suitable for cultiplex arrangements in which a certain number of Language information units is read out at the same time, unusable.

The object of the invention is therefore to provide a device for the holographic storage and reproduction of voice information, that are converted into voice expressions, and with that several thousand linguistic information units in the smallest of spaces can be stored, so that an efficient and easy-to-use multiplexing arrangement is created will.

To solve this problem, it is proposed according to the invention that the device emit coherent, monochromatic light having emitting light source, a light deflector with a control device for any deflection of the light beam, a fixed two-dimensional one. Data mask on which the voice information to be stored is arranged in a matrix are, a sliding lens system for Fourier transformation, a light-sensitive storage medium on which One-dimensional Fourier transform holograms are stored, a variety of photodetectors for detecting the from a one-dimensional Fourier hologram reconstructed image and an arrangement for further processing of the Photo detectors derive electrical signals into voice expressions.

409841/0777

* ° "2AU527

Further details of the arrangement according to the invention can be found in the subclaims.

A notable advantage of the device according to the invention is that the data to be stored are arranged two-dimensionally on the data mask, and that one of the Data mask designed, one-dimensional, Fourier-transformed image is holographically stored on the storage medium.

The production of one-dimensional Fourier transform holograms on a storage medium is already known. One-dimensional Fourier transform holograms of digital data produced in a one-dimensional arrangement, e.g. B. in the form of butted perforations shown and the conversion into memory information by means of a Computer that controls an electron beam that is used to expose photographic film * One such process is known from "Proceedings Fall Joint Computer Conference" 1966, pages 729-734, FIG. 9 in particular shows that grid patterns are produced on photographic film by means of a cathode ray tube. A similar Method is described in the journal "IEEE Journal of Quantum Electronics" June 1969, pages 333-334, wherein a binary hologram is written on a rotating body by means of an electron beam.

In these known methods, the one-dimensional Fourier transform images recorded in binary form, whereby the resolution is limited and on the other hand, a poor signal / noise ratio is achieved during the reconstruction, so that the number of to storing bits per hologram is limited. Since the information to be stored is also one-dimensional per hologram are present, a large number of holograms arranged next to one another for storing two-dimensional information necessary.

409841/0777

-A-

24U527

In the device according to the invention, on the other hand, the two-dimensional information is recorded in a one-dimensional Fourier transformation halogram as PCM-coded speech units. The individual language elements are recorded in the columns corresponding to the y-axis, while the vowel intensities of the individual language elements are recorded in the rows corresponding to the x-axis. The one-dimensional Fourier transformation in the x-direction takes place by means of a lens system. If the two-dimensional image contains m-lines and η-columns, the recorded hologram is that of the one-dimensional Fourier-transformed image of the number m. The number of recordings required in the device according to the invention is therefore only 1 / m of that obtained according to the known method required are. According to the invention, holograms of a certain length can thus be obtained simply by moving the storage medium sequentially in the y-direction and thereby exposing it.

During the reconstruction of the information recorded in this way, the hologram is exposed to a playback light beam, while at the same time the rotating body carrying the storage medium rotates so that a unit of speech information is reconstructed sequentially over time. If the rotational speed of the rotating body is chosen so that for each revolution a speech information is reconstructed, it is possible without the use of keystrokes by sequential Lay out whole parts of speech together.

The invention is explained in more detail below with reference to the figures. Show it:

FIG. 1 shows an arrangement for generating the one-dimensional holograms,

FIG. 2 schematically shows the reconstruction process of the stored holograms and

Figure 3 shows an embodiment of the device according to the invention.

409841/0777

24H527

In the arrangement shown schematically in Figure 1, 1 denotes a monochromatic, coherent light beam, which is focused on a storage medium 4 by means of a cylindrical lens 2. An essentially more transparent one Body 3 on which the to. storing information are shown two-dimensionally, as well as'eine lens 2, are in Arranged beam path. The transparent body has a matrix of m rows and η columns, with the temporal Speech parts to be stored sequentially are arranged in the lines and the vowel intensities of the individual speech parts are included in the columns. An image, which is a pseudo Fourier transform image, is generated on the storage medium in the x direction. contains. A hologram 5 is now created on the storage medium 4 by superimposing the coherent beam ΐ generated with a coherent monochromatic reference beam 6, the latter originating from the same light source. This The hologram is spatially assigned to a specific part of the transparent body 3, i.e. part of the information 3a is a part of the hologram 5a saved. In the y-direction a Fresnel diffraction image is generated, which as an image of the in the y-direction means a telescope system imaged transparent body can be viewed. After saving a hologram, the storage medium 4 displaced in the y-direction by a piece corresponding to the length of a hologram, so that a new hologram can be recorded, whereby a whole row of holograms is created. A second row of holograms can be added in an analogous manner by appropriately shifting the lens and the reference beam 6 in the x-direction. Since the transparent body 3 is arranged fixed in space, you can see immediately that the reconstructed images of all holograms a row in the x-direction at the same place. This fact is essential for holographic storage of language information.

The reconstruction of the speech information stored in the hologram is shown schematically in FIG. On the partial holo-

409841/0777

-C-

24H527

gram 5a, a monochromatic playback beam 7 is directed namely in the opposite direction to the reference beam 6, so that the partial information 3a is reconstructed. The image reconstructed in this way hits a number of photodetectors 8 which convert it into an electrical signal. If the storage medium 4 is moved in the y-direction, the adjacent partial holograms become 5b. and 5c one after the other the playback beam 7 illuminated so that the corresponding Partial information are reconstructed one after the other, after which they are converted by the photodetectors 8 into electrical signals will.

FIG. 3 is a schematic representation of the device according to the invention for speech composition, the light beam 10 originating from a laser 9 being deflected one-dimensionally by a deflector and a control device 11. The deflected beam passes through a lens 1Jj parallel to its optical axis and hits the partial hologram ^ a _{1 of} a hologram track 15a of the hologram 15 in which the speech information is stored as one-dimensional Fcure transformation images. The hologram 15 is located on a rotatable body 14. The image reconstructed from the partial hologram 15a .. hits the photodetectors 17 and is converted by them into electrical signals. These signals are converted into vocal parts in a digital-to-analog converter. The mode of operation will now be described in more detail in connection with FIG. 3, with the hologram arranged on the rotating body 14 being denoted by 15, which consists of a plurality of hologram tracks 15a to 15f and a plurality of PCM-coded speech information items are stored in each track . The position of a partial hologram 15a,. is assigned to a certain point of the transparent body and thus to the multitude of speech information, while the information in the partial hologram 15a,. stored information contain the vowel intensity of the speech information of the corresponding point. There eight

409841/0777

24H527

bits are sufficient to express the vowel intensity, P information can be read out simultaneously, if in each partial hologram 15a,. 8 P bits are stored can.

In the following it is assumed that 16 bits (i.e. two language elements) in partial hologram 15a. are stored; a language element be "good morning" in lane 15a and the other is "thank you / nice." (thank you) in lane 15a, where the former from the photodetectors 17a (four detectors) and the latter from the photodetectors 17b (four detectors) will. It is also assumed that another language element "My Lord" (gentleman) is stored in the hologram track 15f and that this reconstructed image from the photodetector 17b is received. These language elements can be added to the language information "good morning, sir" (good morning, gentleman) in the following way will. A monochromatic light beam 10 is incident on the hologram track 15a, which is transmitted by the control device 11 controlled light deflector 12 is deflected accordingly, so that an image 16 is reconstructed. A Vowel signal, which corresponds to the PCM-coded speech element "Gute Morgen" (good morning), is detected by the photodetector 17a received in that when the rotating body 14 rotates, the reconstructed image 16 falls on the photodetector. The hologram track 15f is then illuminated by the monochromatic light beam 10, which in turn is illuminated by the beam deflector 12 is directed accordingly, so that the data contained in the reconstructed image I6b with one revolution of the rotating body 14 can be received, so that a vocal signal with the PCM-coded speech element "mein Herr" (gentleman) is obtained. By a . twice rotation of the rotary body 14 is so by the photodetector arrangement 17 the Linguistic information composed of "good morning, gentleman". In the same way you can the language information "danke / schö ^ mein Herr" (thank you,

409841/0777

24H527

gentleman) by reading out the information of the reconstructed image i6b in the hologram track 15a will. The PCM encoded vocal signals are conventionally vocalized by a digital-to-analog converter 18 'further processed.

From this embodiment it is clear that the inventive holographic speech composer has significant advantages, since the .vocal data as one-dimensional Fourier transform holograms are stored so that a 'plurality of language elements are stored in a small space can be, whereby the handiness of an effective holographic multiplex system is guaranteed.

409841/0777

Claims (7)

24H527 Claims Device for the holographic storage and reproduction of speech information converted into voice expressions are characterized by a coherent, monochromatic light emitting Light source, a light deflector with a control device for any deflection of the light beam, a spatially fixed two-dimensional data mask on which the speech information to be stored is arranged in a matrix are, a sliding lens system for Fourier transformation, a light-sensitive storage medium, are stored on the one-dimensional Fourier transform holograms, a variety of photo detectors for Recognition of the image reconstructed from a one-dimensional Fourier hologram and an arrangement for further processing the electrical signals coming from the photodetectors in vocal expressions. 2. Apparatus according to claim 1, characterized in that the one-dimensional Fourier transform holograms contain a variety of digitally coded speech information, 3. Device according to claims 1 and 2, characterized in _t that the one-dimensional Fourier transform holograms are arranged in a plurality of parallel tracks on the storage medium. 4. Apparatus according to claim 1, characterized in that the storage medium on a rotating body, and in particular on the outer periphery of a cylindrical drum. 5. Apparatus according to claim 1, characterized in that the arrangement for further processing of the electrical signals is a digital-to-analog converter. 409841/0777 24H527 6. Gate direction according to claim 3, characterized in that the one-dimensional iOurier transformation holograms are arranged in the parallel tracks in the direction of movement of the storage medium. 7. Apparatus according to claim 1, characterized in that in the vertically extending Columns of the data mask individual language elements and vowel intensities in the horizontal lines the individual language elements are arranged or vice versa. 409841/0777 Blank page