HUT59523A - Method and device for transforming sound - Google Patents

Abstract

The illusion of distinct sound sources distributed throughout the three-dimensional space containing the listener is possible using only conventional stereo playback equipment by processing monaural sound signals prior to playback on two spaced-apart transducers. A plurality of such processed signals corresponding to different sound source positions may be mixed using conventional techniques without disturbing the positions of the individual images. Although two loudspeakers are required the sound produced is not conventional stereo, however, each channel of a left/right stereo signal can be separately processed according to the invention and then combined for playback. The sound processing involves dividing each monaural or single channel signal into two signals and then adjusting the differential phase and amplitude of the two channel signals on a frequency dependent basis in accordance with an empirically derived transfer function that has a specific phase and amplitude adjustment for each predetermined frequency interval over the audio spectrum. Each transfer function is empirically derived to relate to a different sound source location and by providing a number of different transfer functions and selecting them accordingly the sound source can be made to appear to move.

Description

BACKGROUND OF THE INVENTION The present invention relates to a method and apparatus for processing an audio signal, in particular an audio signal, such that the resulting sounds are displayed in the observer such that they originate from a location other than the actual location of the speakers.

One can easily estimate the direction and range of a sound source. If multiple sound sources are distributed in space around an observer, the position of each of them can be perceived independently and simultaneously. Despite years of in-depth and continuous research, no exhilarating theory has been developed to explain all the senses of the average observer.

A method of measuring the speed and pressure of a sound wave at a single point and effectively reproducing that sound at a single point preserves the intelligibility of the speech and significantly the identity of the music. However, such a system lacks all the information needed to place the sound in space, so for example, if we were to play a band in such a system, we would perceive that all instruments are from a single point of sound reproduction,

Therefore, efforts have been made to retain information about the direction of sound during broadcast or recording and playback. Published in 1937

- U.S. Patent No. 3,293,940 to C93,540 discloses the basic details of a dual channel system, Artificial highlighting of stereo differences as a means of expanding stereo sound, which is the basis of current stereo audio reproduction,

Some known stereo enhancement systems, in one way or another, couple the stereo channel signals to emphasize the stereo position information in the stereo image. The crossover and the resulting crosstalk depend on the geometry of the loudspeakers and the acoustic space, so they must be set individually,

It is known that attempts to refine stereo systems have not resulted in significant improvements in the widespread use of consumer electronics systems. In fact, students want to sit in a comfortable position while moving or rotating their heads, and to place their speakers in line with other amenities in the room,

SUMMARY OF THE INVENTION It is an object of the present invention to provide a method and apparatus for processing an audio signal which, when performed with two audio transducers, provides the apparent location of the audio source.

adjustable so that it looks to the observer as if the location of the sound source were detached from the location of the audio transducers,

The present invention is based on the discovery that when mono audio is reproduced by means of two independent mono channels and two speakers, the imaginary locations can be produced in a high degree of clarity at different locations. Observing this phenomenon under the special circumstances of a recording studio have led to the necessary conditions determined by regular examinations to produce this surround sound.

According to the present invention, the resulting sound sensation is achieved by placing a sound source anywhere in the three-dimensional space surrounding the observer without any restriction as to the location of the speaker. By using the same two channels simultaneously, you can produce multiple images of an unlimited number of independent audio sources and their independent locations. All you need is two independent channels and two speakers, and the distance or rotation between speakers can be varied over a wide range, The listener can rotate his or her head in any plane, for example, turning to an imaginary sound source, the sound does not deteriorate,

The processing of audio signals according to the invention is characterized by converting a single channel audio signal into a two channel signal by varying the phase and amplitude difference between the two signals as a function of frequency over the entire audio frequency range. This processing is performed by dividing the mono input signal into two signals, and then transmitting one or both signals to devices having a generally uneven amplitude and phase transfer function as a function of frequency. The transmission function may include reverse and frequency dependent delay of the signal . In addition, to the best of our knowledge, the transmission functions used in the signal processing of the present invention cannot be derived from any of the prior art. They can be determined by empirical means. Each signal processing transmission function puts an audio image in a single position, which is determined by the characteristic of the transmission function, so that the position of the audio source is clearly determined by the transmission function.

A given situation may have several different transfer functions, each of which provides an imaginary location, usually at a specific location,

If a moving imaginary location is to be determined, it can be smoothly converted from one transfer function to another.

thus, the process does not need to be confined to static sound imaging,

The audio signals processed in accordance with the present invention can be directly restored after processing or by conventional stereo recording technology on various media such as optical discs, magnetic tapes, audio recording or optical discs, or broadcast by any conventional stereo broadcasting technology such as radio or cable broadcasting. transmission without adversely affecting the audio provided according to the invention,

The mapping method of the present invention can also be applied backwards. For example, if each channel of a conventional stereo signal is treated as a mono signal and the channels are mapped at two different locations in the listener's space, a complete conventional stereo mapping is obtained along the line connecting the mapped locations of the channels. recording a disc on a multi-track tape having, for example, twenty-four channels, each channel can be passed through a transmission function processor, whereby the recording engineer deliberately locates the various instruments and sounds and creates a special sound distribution.

7 ~ can be played with a conventional playback device which, however, already has the sound-imaging features of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The invention will now be described in more detail with reference to the drawings, in which: Figure 1 is a plan view of the observation geometry for determining the mapping parameters;

Fig. 2 is a side view of the geometry of Fig. 1, Fig. 3 * is a top view of the observation geometry for determining parameters of the position of the observer,

Figure 4 · Side view of the arrangement of Figure 3 ·,

5a-5k. Figures 4A to 5B show a top view of the monitoring positions with different speaker positions, i

5m. Figure 4A is a table showing the critical dimensions of three locations;

Fig. 6 is a plan view of an arrangement of an audio transmission experiment in two separated rooms,

Fig. 7 is a block diagram of the process of the invention together with steps in the prior art,

Figure 8 is a schematic diagram of a sound mapping system according to a possible embodiment of the invention, a

Fig. 9 is a perspective view of the operator site of an embodiment of the invention, a

Figure 10 is a perspective view of a computer used to control the present invention;

Figure 11 is a computer graphic showing three perpendicular views used to control the invention.

Fig. 12 is a schematic view of the formation of the apparent sound sources of the present invention in three spaced apart views,

Figure 13 is a block diagram of the device according to the invention, a

Figure 14 shows the waveform of a brain test signal as a function of time, a

Fig. 15 is a tabular representation of a transmission function according to an embodiment of the present invention, a

Figure 16 is a block diagram of an audio mapping system according to the invention, a

Figures 17A and 17B are diagrams of typical transmission functions used in the voice processing system of Figure 16,

18A-18C. FIGS. 4A to 5B are block diagrams of an embodiment of the circuit according to the invention, and FIG

Fig. 19 illustrates a circuit diagram of a further embodiment of the present invention,

In the following, preferred embodiments of the present invention will be described. In order to provide a clear description of the sound mapping process of the present invention, first of all, the terms, dimensions and angles are defined with reference to FIGS.

Fig. 1 is a top plan view of a stereo listener showing left and right speakers 101 and 102, an observer 103, and an audio image 104 seen by the observer 103. For purposes of definition, the observer 103 is on a line 105 perpendicular to the line 106 connecting the speakers 101 and 102 and extending thereon. The position of the observer 103 is that of the reference student, but according to the invention, the listener is not restricted to this position. measuring the azimuth angle of the mapped sound (a) from the listener's position from line 105 counterclockwise to line 107 which is line 107 for the observer 103 and ΙΟΛ5. Similarly, the distance between the monitor 103 and the image 104 is designated as the field of view of the image (r). This range is the real range measured in three-dimensional space and not the projected range measured in a plane or other perpendicular projection,

It is possible with the present invention to reproduce sound images substantially outside the plane defined by the speakers. Accordingly, as shown in Figure 2, the elevation angle (b) of the image is determined. In Fig. 2, the observational location 201 corresponds to the position of the observer 103 in Fig. 1, and the position of the audio image 202 corresponds to the position of the audible image 104 in Fig. The elevation angle (b) of the image is the position of the observer 103 and the line 203 of the line 204 between the line 202 and the image 202. Note that the speakers 101 and 102 are not necessarily in the horizontal line 203,

Having determined the parameters of the location of the image relative to the position of the reference listener, we now determine the possible observation variants. 3, speakers 301 and 302 correspond to speakers 101 and 102 in FIG. 1, while lines 304 and 305 correspond to lines 106 and 105 in FIG. The distance (s) between the speakers is measured along line 304, and the distance of the observer (d) along line 305 is recorded, in the case that the observer is parallel to line 304 at position 3θ7. , a lateral displacement (e) along line 306 is measured. For each of the speakers 301 and 302, the corresponding azimuth angle (b) and (q) is determined from the line perpendicular to the line 304 extending through the speakers 301 and 302.

- 11 counterclockwise to the observer, Similarly, from the observer position an anti-clockwise (m) azimuth angle from the 3 ° 5 line to the listener's perspective is defined,

The height (h) of the loudspeaker shown in Fig. 4 · is measured from the horizontal straight line 4θ1 through the species of observer 303 to the center line of the loudspeaker 3 ° ² ,

Defined parameters allow multiple descriptions of a given geometry, For example, the position of an image can be determined by using data (180, Ο, x) or (0, 180, x) that perfectly match each other,

In traditional stereo systems, the sound is

1, so that the resulting sound image is completely free in space: the azimuth angle (a) may be in the range O-36O ⁰ , and the distance range (r) is not limited to (s) or (s). d) between distances. An audio image can be created up to the observer at or near a fraction of (d), which may be a multiple of (d), and at any one time, at any azimuth angle (a), without having to consider the position of the speakers. with this solution it is possible to stake the sound image ···· • ··· ♦ · «· ·

- also for placement of 12 verbal elevation angles, The distance (d) of the listener may range from 0.5 m to 30 m, whereby the sound image is apparently static in space during the change.

A good sound position can be achieved if the distance between the speakers is 0.2 to 8 meters, while the same signal is applied to the speakers at all distances. The azimuth angles (b) and (q) measured on the speakers can be varied over a wide range without would have an impact on the sound.

It is characteristic of the present invention that a slight change in speaker height (h) does not affect the height angle perceived by the listener. This is true for both positive and negative values, i.e. the speaker can be positioned below the listener's head and above. Since the resulting sound is extremely realistic, it is quite natural for the listener to turn his / her head towards the sound image. While the sound image remains stable, the azimuth angle (m) of the listener does not have a noticeable effect on the sound space. position, at least in the -i-120 ° to 120 ° angle of the angle (m). A placed sound source has such a strong influence on the listener that it easily looks at or points to the sound, and the group of listeners marks the same place, • · · · · · ···· * * • · · «·«

Figures 5 & 5k show ten observer geometries in which the stability of the honeybee is examined. 5a. Figure 1A is a village view of an observation geometry. The left and right speakers 5θ1 and 502 produce the image 504 for the observer 503. 5a ~ 5k. Figures 5a show different loudspeaker controls, which are otherwise similar to those of Figure 5a. Fig. Each of the ten geometries was examined in three different observation rooms, where the distance between the speakers (s) and. distance of students (d) were of different value as shown in 5m. Figures 1 to 4 are tabulated below. Room 1 was a small studio control room with a significant amount of equipment accumulated, Room 2 was a large wall-mounted studio room that was practically empty, while Room 3 was a small experimental room with a sound-absorbing coating on three walls.

For each experiment, the observer is asked to mark the position of the sensed image under two different conditions; in one case the angle of the head (m) was 0 and in the other the observer turned to the position of the apparent sound. Each experiment was repeated with three different observers so that sound stability was examined in a total of 180 configurations. For all 180 configurations, the same signal was applied to the speakers. In each case a

- The azimuth angle (a) of 14 images was perceptible at -60 °.

FIG. 6 is a schematic diagram of an audio transmission experiment in which the audio 601 was produced by the method of the present invention, using loudspeakers 602 and 603 in the first room 604. In Dgy 605, such as that described in German Patent Specification 1 927 4θ1, there is a left and right microphone 606 and 607 in place of the tabs. The electrical signals from the microphones 606, 607 and 608 and 609 are amplified separately from one another. Amplifiers 610 and 611 which drive a left and a right speaker 612 and 613 in the second room 614. 3g observers 615 were located in the second room 614, which was acoustically insulated from the first room 604, The observer 615 detected a sharp secondary sound 616 corresponding to the first 60Z; 601 pictures of the village.

FIG. 7 illustrates a relationship between the audio processing system of the present invention and known systems in which one or more multiband audio sources 7θ1, such as magnetic tape devices, transmit mono signals 702 from multiple sources to a mixing table 7θ3. such as changing levels and compensating for frequency turns in any way you want.

The plurality of modified mono signals 704 produced by the 7θ3 mixing table are fed to an input of the 7θ5 audio processing system of the present invention. Within this system, each input channel has a sound image position and a processing system for producing a dual channel signal from each input signal 704 is used. Each of the two-channel signals is mixed to produce a final 7 ° signal 707, which can then be traced back to the mixing table 708. It should be noted that the dual channel signals produced in accordance with the present invention do not in fact correspond to the right and left channels of the stereo signals, this designation merely facilitates reference to these signals. Thus, when all two channel signals are mixed, all left channel signals are combined into a single signal and all right channel signals are combined into a single signal. In practice, the mixing table 703 and 708 may consist of discrete units within the same table. At the control table, it is possible to input the processed signals into the speakers 7θ9 and 710 for observation. After any desired modification and leveling, the master stereo signals 711 and 712 are transferred to a master stereo recorder 713, which may be a two-band magnetic tape recorder. Devices following the processing system 705 are well known in the art.

Λ · »· ··· · ····« · »

The audio processing system 7θ5 is shown in greater detail in Figure 8, where the input signals 801 correspond to the signals 7θ4 in Figure 7 and the output signals 807, 808 correspond to the signals 711, 712 in Figure 7. Each of the input signals 801 is fed to an individual signal processing unit 802.

These signal processing units 802 operate independently of one another, with no audio frequency coupling between them. Each of the signal processors 802 generates two-channel signals having a differential phase and amplitude as a function of frequency. These transfer functions are discussed in detail below. The transmission functions, which can be described in terms of time units as real impulse response functions or the equivalent frequency unit as complex frequency response or amplitude and phase response, determine only the position of the desired image to which the input signal is assigned.

One or more processed signal pairs 803 produced by the signal processing unit are applied to the inputs of the stereo mixer 804, some or all of which may be provided to the input of a storage system 805. This system is capable of storing all processed stereo audio signals and can simultaneously play them back on the 806 outputs. · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · ·

- It has 17 input channel pairs and an outgoing oscillator pair.

Multiple outputs 806 from the storage system are routed to a further input of a stereo mixer 804. Stereo mixer 804 summed all left signals and generated a left output 807 and outputs a right output 808 from each right signal and possibly adjusts the amplitude of all inputs before summing. There is no link between left and right channels in the mixer,

The system is controlled by an operator 809 through human controls 810 and determines the desired audio situations for each input channel.

It may be particularly advantageous to include digital signal processing units 802 so that there is no limit to the position, path, or velocity of an image. These digital audio processing systems, which make the necessary phase and amplitude differential adjustments as a function of frequency, are described in more detail below. Interfacing such a digital system is not always economical for real-time signal processing, but it is completely economical. In the absence of real-time signal processing, the outputs 803 need to be connected to storage system 805, which should be capable of slow recording, and real-time playback. Conversely, if a ·· · ···· ···································································································································

- 18 corresponding number of real-time signal processors 802 are available, then storage system 805 may be omitted,

As shown in Figure 9, the operator 901 controls the 9 ° 2 control table, which is provided with a left and right stare monitor 9θ3, 9θ4 · Although the distance (s) between the speakers is 0.2 meters for stability of the final processed image however, the distance between the speakers of the mixer is preferably at least 0.5 meters. This spacing makes it much easier to achieve accurate audio capture. There is also a computer graphic 9θ5 display, a multi-axis 9 ° 6 controller and a 907 keyboard along with the appropriate computing and storage units.

With the aid of a computer graphic display 9θ5, the position or movement of the sound image can be displayed on the image, such as in Figures 10 and 11. Figure 10 is a representation of a listener position 1001 with a typical listener 1002 and an image 1003. indicated by a moving screen 1004 and perspective 1005, 1006 spatial signals.

The bottom of the display shows a cool 1007 whose data applies to parts of the path of the audio source, including recording, synchronizing, and entering information19. Menu items can be selected from the 9θ7 keyboard or by moving the cursor 1008 to the desired item using the multi-axis 9θό controller. . The selected item can be modified or blinked from the keyboard 907 by actuating the button on the multi-axis controller 906 by selecting the appropriate operation. A menu item 1009 allows the operator to move the perspective view of the perspective image after the software with the 9-axis multi-axis controller. the respective control situation or mozgáspályájá't sound image E _g y other makes another 11 shown in FIG enlarged display of the menu item 1010,

11 shows a full perspective view 1001 of FIG. 10, three views of the same image, a top view 1101, a front view 1102, and a side view 1103. To aid in the interpretation of the views, a smaller and less detailed version 1104 of the perspective image 1001 is shown in the remaining image quarter. In this case, at the bottom of the screen, like menu 1007, a menu 1105 is shown with similar functions. One of the menu items 1106 allows you to return to the picture of figure 10, »

In Figure 12, sound sources 1201, 1202 and 1203 in a first room 1204 are detected by two microphones 1205 and 1206.

which produce the right and left stereo signals recorded with a conventional stereo 1207 audio recorder.

When played back using a conventional stereo reproducing device 1208 and fed into the speakers 1209, 1210, the signals from the microphones 1205, 1206 are then placed in a second room 12115j. student is a traditional stereo 1211, 1212,

1213 detects sound images according to sound sources 1201, 1202, 1203, The sound images are located on the line connecting the speakers 1209, 1210 according to the lateral position of the sound sources relative to microphones 1205, 12 ₂ ,

In the case where the two pairs of stereo signals are processed and combined in the sound processing unit 1216 described above and played back with a conventional stereo 1217 device on a right and a left speaker 1218, 1219 in a third room 1220, the apparent audio sources for the listener 1226 are in space. away, in places independent of the actual position of the 1218 and 1219 speakers. Suppose that the processing was such that the image of the original right channel signal is 122A. and the original left channel image was at 1225, each of these hag pictures looks like a real speaker, and we can think of the sound pictures as seeming speakers,

The transmission function, which is set as a function of the differential amplitude and phase of the two-channel signal over the frequency band, must bring the sound of a mono source to the desired position. In practice, the amplitude and phase differences to determine the? are defined at no longer intervals in the two channels independently of each other throughout the entire frequency range for better image stability and coherence. In the event that there is no need for high quality and accurate placement of the sound source image, the frequency intervals can be increased. Thus, the specification of such a transmission requires about 1000 real numbers (or equivalent to 500 complex values). The human perceptual differences in the spatial location of the sound source are somewhat indefinite, as they are based on subjective measurement, but in a real three-dimensional space for students, The exhaustive characterization of all location transmissions is possible with a large number of data, which is more than 1000000 in total, and s is being collected.

It should be noted that the audio processing system of the present invention, which performs differential tuning between the two channels, is built up of pieces by · · · examining the 4θ Hz intervals of the entire audio frequency band. In addition, the transmission function of the audio processing unit places the sound source in one place relative to each other, spaced apart, that is, at a single azimuth, height, and depth,

In practice, however, it is not necessary to express all transmission functions, since mirror symmetry generally exists between the right and left channels. When the transmission characteristics modifying the .osa / V tournaments are interchanged, the azimuth angle (a) of the image is inverted while the elevation angle (b) and distance (r) remain unchanged ·

It is possible to demonstrate the method and sound sensory disappointment of the present invention by conventional equipment and by using simpler signals, where a sine wave pulse is known to be smoothly gated at relatively long intervals, the resulting signal occupying a very narrow band within the frequency unit. In fact, this signal samples the desired transmission at a single frequency, since the desired transmissions, i.e., the transmission functions, are simplified for easy control of the differential amplitude and phase (or delay) between left and right channels as a function of frequency.

• · · ·

In this way, the transmission function can be empirically assembled with different differential phase and amplitude settings to accommodate a given audio source at each selected frequency interval over the entire audio frequency range. According to Fourier's theory, every signal can be generated as the sum of a series of sine waves, so the applied signal is landscape.It is original,

An exemplary embodiment of a system for demonstrating the present invention is shown in FIG. 13, wherein a 13θ2 audio frequency synthesizer (a LieLess-Packard Multifunction Synthesizer model 8904A) is controlled by a computer 1301, a Hewlett-Packard model 33OM, and thus a mono generating an audio signal which is applied to the inputs 1303, 1304 of two channels of an audio delay line 13 ° 5. The delay line 1305 is a Fventide Precision Delay type PD86O. From the delay line 1305, the right channel signal is transmitted to a switchable inverter 1306, then the left and right channel signals pass through a variable attenuator 1307, 1308, and then the next two power amplifiers 1309, 1310, left and right 1311, 1312 connected to speakers,

Synthesizer 1302 has a smoothly gated sine wave

Generates 24 pulses at any desired test frequency 1401, the envelope of which is shown in Figure 14, the sine wave is gated along a first linear incline 1402 for 20 ms, then held at a constant amplitude 1403 for 45 ms, followed by a 20 ms gates downward along a second linear slope 14 04. These pulses are repeated at 14θ5 time intervals of 1-5 sec.

In addition, using the system of FIG. 13, the waveform of FIG. 14 can be constructed by adjusting the delay function of the delay line 13θ5 and adjusting the amplitudes with the dampers 1307, 1308 over the entire frequency range. who observes the location of the audio image and determines whether it is in the correct position. If so, the next frequency range should be examined. If not, further adjustment is required and monitoring should be repeated. In this way, the transmission function can be compiled over the entire audio frequency range.

Figure 15 is a brain table that contains data for the transmission function that allows the generation of audio images, much different from the direction of the loudspeakers • ····················································································· · · * · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · For multiple sine frequencies. The figure is shown by a larger number of observers in FIG. 4A, who have turned their heads over a wide range, including turning to the image.

The placement of narrowband signals may be generalized as detailed above, allowing for the audio representation of broadband signals, which are complex audio sources such as speech or music. If the differential amplitude and phase shift in the two channels are derived from a single input signal defined for each frequency in the full frequency range, then the full transmission function is specified. In practice, for a given number of frequencies in the band, amplitudes and delays. At any intermediate frequency between these frequencies, the amplitudes and delays can be determined by interpolation. If the frequencies at which the transmission is determined are not too far apart and the smoothness or extent of the change in the real transmission is taken into account, the interpolation procedure is not too critical. In the table of Figure 15, the signal amplitudes and delays of each channel are shown; this can be seen essentially in Figure 16.

*, Where separate 1500, 1501 audio processing units are used. The single-channel audio frequency signal is fed through input 1502 to each of the audio processing units 1500, 1501, adjusting their amplitude and phase as a function of frequency such that the difference between the right channel 1503, 150A output of the slit is the appropriately determined and detailed above . The control parameters are transmitted via line 1505, which is used to adjust the differential phase and amplitude so that the image is in a separate fixed location, e.g., a digital audio processing unit may consist of a finite impulse response (FIR) filter whose coefficient can be changed by the control parameter signal. in such a way that different transfer functions come up the ladder.

The system shown in Figure 16 can be simplified as will be apparent from the following analysis. First, the difference between the delays of the two channels, or with differential importance, Suppose that the left and right delays are t £ l) and t (r), the delays of the new channels t '(1) and t' (r) are any fixed t (a) by adding a delay:

t '(1) = t (1) + -6 (a) (1) t * (r) = t (r) + t (a) (2)

The result is that the full effect is heard t or a later or earlier, provided t (a) is negative • · · · · · · · · · · · · · · · · · · · · · · · · ·· · · · · · · ··· ··

- It's worth 27. This general expression is true in the special case where t (a) = -t (r). Assuming that t '(l) = t (1) - t (r) (3) t' (r) = t (r) - t (r) = 0 (4) with that transformation, the delay in one channel for each In case of practical implementation, we always have to make sure that the smaller delay is subtracted so that a negative delay occurs. To overcome this problem, it is preferable to leave a constant residual delay in one channel and change the delay in the other channel. If the fixed residual delay is large enough, the variable delay must be negative.

Secondly, it is not necessary to adjust the amplitude of the channels independently of one another. It is a common practice in the audio frequency technique to change the amplitude of the signals by amplification or attenuation. As long as the stereo channels have the same rate of change, there is no change in the content of the positional information. The ratio or difference of the amplitudes is important and should be maintained. As long as this difference is maintained, all of the effects and sensations described herein are completely independent of the level of overall sound produced. Accordingly, all amplitude controls • · ·

It can be performed in 28 single channels similar to the timing and phase control described above, while leaving the other channel at a constant amplitude value. Again, it is desirable to leave a fixed residual damping in one channel in order to adjust the desired damping ratios in the other channel. Full control is thus achieved with a variable damping device in one channel,

In this way, all the necessary information can be determined by determining the differential attenuation and delay in a single channel as a function of frequency. For the second channel, a fixed, frequency independent attenuation and delay can be determined; if this is not specified, assume unit gain and zero delay.

In this way, for any audio position, and accordingly for any left / right transmission function, the differential phase and amplitude adjustment (filtering) can be arranged in one or the other channel or in any combination thereof, one of the audio processing units Ι5θθ, 1501 can be simplified it contains variable impedance or simply a jumper. This cannot be an open circuit, Assuming phase and amplitude adjustments are only in one channel.

• ·· ·

- 29, in order to create the necessary difference between the two channels, the corresponding transmission functions are shown in Figures 17A and 173,

17A. Figure 1A shows a typical transfer function _? for the differential phase of the two channels, in which the left channel is unchanged and the right channel is phase tuned in the frequency range as a function of frequency. Similarly, Fig. 173 shows the differential amplitude of the transmission function between two channels, where the left channel amplitude is unchanged and the right channel attenuation is varied as a function of frequency in the frequency range 0

The audio processing units 1500, 1501 shown in FIG. 16 may be, for example, analog or digital, which may include one or more of the following: filters, delayers, inverters, sumers, amplifiers, and phase shifters. the transfer function is feasible. There are many equivalent embodiments that are customary in the art.

For example, a delay can be defined as a phase change at a given frequency using the following equations:

phase (degrees) = 360 3 (delay time) x frequency phase (radians) = 2 xTx (delay time) x frequency

Caution should be exercised in applying these relationships, as it is not sufficient to determine the phase default value; in the above equations the total phase shift shall be taken into account,

It is common practice in electronic design practice to represent complex s shikers. All filter characteristics that can be realized with real analog components can be defined (arbitrarily) as the quotient of two polynomials in the s Laplace complex frequency variable. The general formula:

_T (s) = ₌ (5)

Bout (s) D (S)

T (s) is the transfer function in plane s, Ein (s)

Sout (s) is the input and output signal as a function of s, and in the above formula the H (s) of the counter and the heater

D (s) takes the following form:

⁰ π

IT (s) = a _Q + a .jS 4- a £ S 4 · a ^ s' 4- ... + a ^ s (6)

D (s) = b _Q -> b ^ s 4 · b ₂ s ² 4- b ^ s ^ 4- ... 4 · b _b s ⁿ (7)

The advantage of the dnnek s relation is its simplicity, In order to determine the transmission function completely at all frequencies, without the need for interpolation, it is only necessary to determine the coefficient of n4-1, * ·· · * ··· · · · · · · · · · · · · · · · · · · · · · · · ···

- 316s is the coefficient of b n + 1. With these defined coefficients, the amplitude and phase of the transmission function can be readily determined at any frequency by well-known methods. Another advantage of this determination is that this form is most easily deduced by analyzing an analog circuit and therefore the most natural, simplest and generally accepted method to determine the transmission function of such a circuit,

Another suitable illustration of the method of the invention is the representation in the z-plane. In a preferred embodiment of the present invention, the signal processing unit may be configured with digital filters to take advantage of its advantageous flexibility. Since the location of each image can be determined by the transmission function, we need a filter in which the transmission function is easily and quickly with minimal restrictions. feasible, A fully programmable digital filter is available to meet this need.

Such a digital filter may operate in the frequency unit, which ash is first transmitted by the Tourier transform from the time unit to the frequency unit. By one of the methods described above, the amplitude and phase stroke of the filter is achieved by complex multiplication to represent the signal in the frequency unit. Finally, with the inverse Fourier transform, the signal is put back into the time domain of the digital-> · · · · · · · · · · · · · · · · · for analog conversion.

Alternatively, the transmission may be directly determined in the unit of time as a real impulse response. This response function is mathematically equivalent to the phase amplitude and / or the response function in the frequency domain and can be determined from this by the Fourier inverter transformation. The impulse response function can be applied directly in the time domain by plotting the signal in the time domain. It can be shown that the convolution operation in the time domain is mathematically equivalent to the frequency domain multiplication, so that the direct convolution is completely equal to the frequency domain operation described above.

Thus, all digital operations are discrete rather than continuous, discrete expression is preferable to continuous, A response function is usually defined by direct coefficients applied in a recursive direct convolutional digital filter that can be easily performed in parallel with the plane expression so that if T (z) is the response function of the time domain s = with the expression T (s) in the frequency domain then l (Z) is -D (z) where IT (z) and D (z) are of the form:

(8)

- 33 <· · · · · · · · · · · · · · · · · · · · · · · · · · · · · · ··· ·

Ν (ζ) = οθ -.> · C ₁ z ¹ * ο _£ ζ ² +

B (z) = d _Q * d ^ ^-1 + d ₂ z ” ² +

+ o

(9) (10)

In this expression, the coefficients c and d are sufficient to determine the function, as the coefficients a and b were sufficient to determine in plane s, so they can be equivalent. The z siku filter can be directly inserted if the operator is interpreted as zn sampling interval delay.

'Then the coefficients c and d are directly coefficients of the insertion. The exponents of z must be limited to negative values, since this corresponds to a positive delay. The positive exponent of z corresponds to the negative delay, which is a transmission that would be prior to excitation.

With these terms and the device already described, it is possible to place broadband audio images such as speech and music. From this, the audio processing unit of the present invention, such as the processing unit 802 of FIG. 8, may be implemented with a two-way variable analog filter, with variable coupling dampers as shown in FIG. 18A.

18A. 1A, a mono input signal is applied to input 16O1, which is the input of two filters 1610, 1630, and which input is directed to two potentiometers 1651, 1652, The output of filters 161O, 1630 is the 1653, 1654 po4 · ·· * 4 · ·· ··· · * · »« · 9 · * ·· 4 «·« ·· «» r * ··

- 34 tonne otné joins squares. The four potentiometer hubs 1651-1654 are arranged in a so-called joystick control so they work in a differential manner. it operates mechanically in the opposite direction and transmits a smaller signal from its input signal to its output. 1653, 1651}. potentiometers likewise have differential operation on a second independent joystick axis. In 1653, '

The signals displayed at the output of the potentiometers 1654, through the unit-amplified separating stages 1655, 1656, reach the potentiometers 1657, 1658, which cooperate with each other by incrementally increasing or decreasing the ratio of the signal input to the input. The output signal of the potentiometers 1657, 1658 is provided to an inverter switch 1659, which allows the filter signals to be directly or interchanged to the first input of the summing elements 1660, ,67θ.

The second input of a summing element 1660, 1670 has an output signal of potentiometers 1651, 1652.

The summing element 1670 controls the inverter 1690, and the switch 1691 allows the input 1684 of the attenuator 1689 to be directly or inverted.

“35 The output signal of the attenuator As 1689 is the so-called right-hand signal. Similarly, the summing element 1660 controls the inverter 168 1 and the switch 1682 allows the signal to be directly or inverted to the 1683 point. The switch 1685 provides the possibility that the signal at point 1683 or the signal at input 16O1 is applied to the attenuator 1686, which produces the output 1688 of the left channel,

The filters 1610, 1630 are identical to each other, one of which is shown in greater detail in FIG. 18B. Single Unit 1611 Separation Stage Receives a signal at input 1601 which is capacitively coupled to the filter element 1613 on the capacitor 1612. Similar filter elements 1614 1618 are connected in series, and the last filter element 1618 drives an inverter 1621 through a capacitor 1619 through a unit gain 1620. The switch 1622 allows the output 1623 to be provided with either the output of the isolation stage 1620 or the output of the inverter 1621.

The filter elements 1613 to 1618 have the same structure, one of which is shown in Figure 18C. They differ only in the value of the capacitor 1631 contained therein. Input 1632 is connected to capacitor 1631 and resistor 1633, and resistor 1635 is a

It is connected to the inverting input of an operation amplifier 1634, and the output 1636 is the output of the filter element. A feedback resistor 1635 is mounted to the operational amplifier 1634 in the conventional manner, The non-inverting beaen of the operational amplifier 1634 is connected to one of the resistors as 1631 and 1637 to 1642. This filter is an all-through filter, the phase shift of arBBly varies with frequency depending on the position of switch 1643,

Table 1 lists the values of capacitor 1631 used in each filter element 1613-1618, and Table 2 lists the resistances selectable by switch 1642, These resistivity values are the same for all filter elements 1613-1618.

A detailed embodiment of one of the summing elements 166O, I67O is shown in Fig. 18D, where the two inputs 1661-1662 to be summed are summed on an operation amplifier 1663, and the summed signal appears on the output 4664 of the amplifier 1665i 1667. , as well as the feedback resistance of s 1666. In either case, switch 1659 is connected to input 1662 and joystick potentiometers 1651, 1652 are connected to input 1661. Table 3 shows the setup values for an audio placement

- 37 as an example, in which the soundtrack “flies in the direction of a helicopter, which is well above the loudspeaker and the loudspeaker. To produce the mono signal required for the method of the present invention, the stereo bands of a noise effect disk were summed with the apparatus shown and the tabular settings. there will be audio images in space such that the listener will detect a helicopter

1, Table a

filter 1 2 3 4 5 6 1631 capacitor value in nF 100 47 33 15 10 4.7 2, table a t 1642 switch standings 1 2 3 4 5 Resistor 1637 1638 1639 1640 1641 Resistance value Chm in h7co 1000 470 390 120

···· · · · · · · · · · · · · · · ···

Table 3

1630 filter elements, 1 switch position 5 5 1630 filter elements, 2 switch positions 5 5 1630 filter elements, 3 switch positions 5 5 1630 filter elements, 4 switch positions 5 5 1630 filter elements, 5 switch positions 5 5 1630 filter elements, 1630 inverter switch normal normal 1652 potentiometer ratio 0.046 0,054 1654 potentiometer ratio 0.90 0.76 1653 potentiometer ratio 0.77 0.77 1691 inverters switch position invert. invert. 1685 selectors switch position 1601 1601 1686 output damping ratio 0.23 0.23 1687 output damping ratio 1.0 1.0 azimuth image angle (a) in degrees -45 -30 height (b) of sound image + 21 + 17 sound distance (r) away from away from

Note to Table 3: In both cases, the position of the inverter switch 1659 is such that signals from element 1657 drive element 1660 and signals from element 1658 drive element 1670.

- 39 By adding two separate circuit elements to the above arena, it is possible to shift the student space in the height direction. Note, however, that this is not essential to the creation of the audio images. These separate elements are shown in Figure 19, where the left and right signals 17θ1, 1702 are taken from the outputs 1688, 1689 of the signal processing unit shown in Figure 16. 33, 1704 delayers are inserted, and the output signal of the delayers 17θ3, 1704 becomes the audio processing unit

Output 17C5, 1706,

With this auxiliary unit, the delay in each channel is frequency independent. In this way, this can be perfectly characterized by a single real number. The left channel delay t (l) and the right channel delay t (r), As in the case above, only the difference between the delays is of interest and the device can be controlled by determining the difference between the delays. A fixed delay is inserted into each channel to ensure that at least no negative delay is required to achieve the desired difference. The delay t (d) defines the difference. You can drink it as t (d) = t (r) - t (l) (11)

If t (d) is equal to C, the effect produced is not substantially affected by this auxiliary unit,

In the case that t (d) is positive, the center of the listening space shifts to the right in the direction (3) shown in Figure 3, the positive value of t (d) corresponds to the positive value of (e), typically producing a right shift. equals the negative value of (e), which can be obtained by the negative value of t (d), This procedure allows the entire field of view in which the student senses the effect to be laterally displaced between or beyond the speakers, Dimension (e) can easily exceed the wall of the (s) dimension, and good results were obtained with a shift where the size of (e) was 83 fs, but Ξζ is not a technical constraint, but is limited to the present experiments

Claims (12)

A method for generating and locating the apparent origin of a selected tone from an electrical signal corresponding to the selected tone at any predetermined and delimited location within a three-dimensional space containing an observer, the electric signal being separated into first and second channel signals, respectively. changing the amplitude and shifting the phase of at least one of the second channel signals as a function of frequency, generating at least one modified first or second channel signal and generating a differential phase and amplitude difference between the two channel signals, then at least the first and second modified channel signal within the three-dimensional space to a first and second voice transducer located on either side of the observer and producing a sound that is Nation space comes from an apparent predetermined location that may be different from the location of the audio transducers. The method of claim 1, wherein the amplitude change and phase shift are performed on both the first and second channel signals, while the phase and amplitude between them is β β · · • · • • • •. · · 4 · • · 4 • · · · · · · · · · · · · · · · - We maintain 42 differences. The method of claim 1, wherein at least one of the signals of the first and second channels is provided with an operation amplifier to an all-pass filter having a specific frequency path and a topology derived empirically for the Laplace complex frequency variable (s). transfer function. The method of claim 3, wherein the at least one signal applied to the filter is further applied to a plurality of filters coupled in series. The method of claim 1, further comprising storing the signals of the first and second channels and generating, at a later selected time, signals modified by a suitable device from the stored signals. 6. The method of claim 1, wherein during amplitude change and phase shifting, at least one of the first and second channels is fed to a sound processor that produces a differential phase transfer, a specific phase transfer function and a differential amplitude shift, a defined amplitude transfer function. across. 7. The method of claim 6, wherein the frequency response of said phase and amplitude transfer function is 40 Hz inter9 · Λ. · - A. "•the *· • · · · ·· · · · ··· ·"· ····· · · " ·· · ··· ·· - 43 based on religions. A system for treating a signal to produce a sensory illusion of sound and to produce and locate the apparent origin of at least one sound in a three-dimensional space containing one observer using two transducers disposed in the open space from an electrical signal corresponding to the selected sound. and a second channel means, one of the first and second channel means comprising means for varying the amplitude and phase shift of the electrical signal in response to frequency, producing a corresponding modified signal such that the amplitude and phase difference between the two channels is each frequency having a first and a second channel to the two converters. The system of claim 8, further comprising a second means for generating a second modified signal in the second channel, wherein the second modified signal also generates amplitude change and phase shift as a function of frequency and maintains the amplitude and phase difference. The system of claim 8, further comprising a storage means which is connected to a means for generating modified signals from the stored signals at a later selected time. · · · · · · · · · · · · · · · · · · · · · · · · · · · · ~ 44 " 11. The system of claim 8, wherein the amplitude change and phase shift generating means comprises a sound processor for generating a differential phase transfer with a defined phase transfer function and for producing a differential amplitude shift with a specific amplitude transfer function. System according to claim 11, characterized in that the frequency dependence is based on 40 Hz intervals.