DE19901546A1 - Electronic circuit for natural reproduction of electric audio and music signals via earphone set - Google Patents

Abstract

The circuit has two input and two output channels for processing a stereo audio signal for reproduction. The right and left audio signals of the stereo audio signal are offered to the two input signals. The signal at one output channel forms a sum of a signal, derived from left audio signal by filtering with frequency dependent amplifying and time shift of signal components.The sum contains also a signal, derived from the right audio signal by a second filtering with a frequency dependent amplifying and time shift of signal components. The sum at the other output channel is specified.

Description

The invention relates to an electrical circuit for the natural reproduction of electrical audio and music signals via headphones.

In order to achieve a natural, spatial sound image, come with one conventional stereo playback of audio and music signals two speakers for use. The loudspeakers stand in front of the left and in front of the listener, and form with the listener the so-called stereo triangle. If the acoustic signals are from the two speakers are broadcast, are different, differ also the acoustic signals perceived by the left and right ear become. Thus, an impression of direction and space can be given during playback be transmitted.

When reconstructing a spatial sound image, the listener unconsciously evaluates several properties of the acoustic signals.

• - The sound from the right speaker is mainly from the right ear recorded, this is called the "right direct transmission signal". Nearly at the same time, however, it is also perceived by the left ear. this is that so-called "left crossover signal". Likewise, sound reaches from the left Speakers the left ear, the "left direct transmission signal", as well as the right Ear, the "right crossover signal". The difference in intensity between those of the Ears perceived acoustic signals from a signal source contains information about the direction of this source.
• - The sound from the right speaker reaches because it is a longer one to the left ear The right ear has to go away earlier than the left ear. Also achieves an acoustic Signal from the left the right ear with a certain delay. This Time difference between left and right ear contains additional information about the direction of the signal source. It plays - especially at low frequencies below 1 kHz - an important role.
• - By bending and reflecting the sound waves on the head and on the ear cups the sound pressure on the eardrum not only from the direction of the sound source, but also  also depends on the frequency. In this way the sound spectrum is transmitted additional sense of the direction of the source.
• - A reflection arises from the sound waves on the walls of the listening room Reverberation that creates a feeling of space.

When playing via headphones, the above-mentioned mechanisms for the reconstruction of a spatial sound image are eliminated.

• - The acoustic signal from the right (left) transducer is only connected to the right (left) ear. There is no acoustic crossover and it is created unnaturally strong differences in the sound intensity on the left and right ear.
• - Because there is no acoustic crossover, there are no delay times with which the listener can locate a sound source.
• - Due to the direct coupling of the sound transducers to the ears, there is no reflection and bending on the head and ear cups instead.
• - There is no reverberation because the acoustic properties of the listening room do not be included.

For the realization of a more natural sound reproduction through headphones electrical circuits have been designed that give the sound experience of two speakers simulate. The electronic generation of a crossover signal is in the Foreground. The left (right) electrical audio signal is after filtering (a frequency-dependent attenuation) and with a defined time delay right (left) audio signal added. This way the acoustic crossover replaced when playing speakers.

A disadvantage of this electronically generated crossover shows up with sound sources that seem to come from the middle of the sound image. The left and right audio signals are the same for these sources; it is a so-called mono signal. With a real one Sound source located in front of the listener reaches the acoustic signal both ears at the same time. There is no time-delayed crossover signal. Thus, when playing a mono signal via speakers or headphones In principle, electronically generated crossover is also undesirable. The two are  Crossover signals on the left and right ear are the same and thus generate a right sense of direction, but the sound quality decreases due to the principle. With the crossover, the sum of the left (right) direct transmission signal is formed and the left crossover signal created from the right (left) input signal left (right) output signal. This sum signal is the left (right) Sound converter offered. When a mono signal is played back, a Interference between the direct transmission signal and the added one Crossover signal. Due to the time delay of the crossover signals, the direct Transmission signals amplified in certain frequency ranges and in others Frequency ranges weakened. The frequency response of the generated output signal is no longer flat, but shows a larger one in the medium and high frequency range Number of reductions. This is the so-called "comb effect".

The invention is a method for generating crossovers, the mono signals remain undistorted.

The crossover signal ( 3 ) is, as usual, generated by frequency-dependent attenuation from the input signal ( 1 ) (see Fig. 1). The time delay of the crossover signal also depends on the frequency and decreases at higher frequencies.

Before the direct transmission signals and the crossover signals are summed, the direct signals are filtered (a frequency-dependent attenuation) and are given a frequency-dependent, temporal shift to earlier times ( 2 ). The sum of the crossover signal ( 3 ) and the direct transmission signal ( 2 ) is equal to the original input signal ( 1 ).

With a mono signal that is offered to both inputs, two are the same Crossover signals generated. The direct transmission signals are also the same. Consequently after summation, both output signals are the same as the input signal. At a Shift the virtual sound source to the left or to the right, taking the signal is now offered in different intensities at the two entrances, crossover signals arise with natural delay times in the low Frequency range.  

As an example of the invention, the diagram of an electrical circuit is drawn in Fig. 2, with which the desired crossover signals can be implemented in a relatively simple manner. To increase the input impedance, both input signals ( 4 L, 4 R) first pass two input amplifiers ( 5 L, 5 R). First, the left input signal V _{left, in} (t) is offered on one side of a chain of three series-connected electrical networks ( 6 L, 7 , 6 R). The right signal V _{right, in} (t) is offered on the other side of the chain. The impedances of the first ( 6 L) and the last network ( 6 R) are the same; this is Z ₁ . The impedance of the middle network ( 7 ) is Z2. Both output signals are available on the middle network. To lower the output impedance, these output signals ( 9 L, 9 R) pass two output amplifiers ( 8 L, 8 R).

The following applies to the output signals:
V _{left, from} (t) = (Z ₁ + Z ₂ ) / (2 * Z ₁ + Z2) * V _{left, in} (t) + Z ₁ / (2 * Z ₁ + Z ₂ ) * V _{right, in} (t)
V _{right, from} (t) = (Z ₁ + Z ₂ ) / (2 * Z ₁ + Z ₂ ) * V _{right, in} (t) + Z ₁ / (2 * Z ₁ + Z ₂ ) * V _{left, in} (t)

The desired time shifts and amplitude adjustments are realized by a suitable choice of Z ₁ and Z ₂ . If the left and right input signals are the same, the output signals are also the same:
V _{left, in} (t) = V _{right, in} (t) ⇒ V _{left, out} (t) = V _{right, out} (t) = V _{left, in} (t) = V _{right, in} (t)

A practical implementation of the circuit is shown in Fig. 3. Z ₂ consists of a single resistor R ₂ ( 12 ). Z ₁ contains a parallel connection of a resistor R ₁ ( 10 L, 10 R) and a capacitor C ₁ ( 11 L, 11 R). Suitable component values are:
R ₁ = 1000 ohms
C ₁ = 470 µF
R ₂ = 2200 ohms

Fig. 4 shows the relative amplitudes of the direct transmission signal and the crossover signal as a function of the input frequency. Crossover is pronounced at low frequencies and decreases at higher frequencies. This simulates the shadow effect of the head at higher frequencies. At low frequencies, the sound waves bend around the head. The direct transmission signal is much stronger and increases slightly at higher frequencies.

The time shifts of a signal as a function of the input frequency are shown in Fig. 5. The positive values for the crossover signal are synonymous with a delay. The negative values for the direct transmission signal are synonymous with a time shift of the signal towards earlier times. The total time shift between the direct transmission signal and the crossover signal is 320 µs with the specified component values at low frequencies. This roughly corresponds to the natural crossover time when played back using loudspeakers.

Literature and patents

[1] Asahi N., "Out-of-head localized sound reproduction system for headphone", US Patent 4136260, 1977.
[2] Bauer B., "Stereophonie to binaural conversion apparatus", US Patent 3088997, 1960.
[3] Bauer B., "Stereophonie earphones and binaural Loudspeakers", Journal of the audio engineering society 9/2, pp. 148-151, 1961.
[4] Begault RB, "Perceptual effects of synthetic reverberation on three-dimensional audio systems", Journal of the audio engineering society 40/11, pp. 895-904, 1992.
[5] Blauert J., Laws P., "Method and apparatus for imparting to headphones the sound reproducing characteristics of loudspeakers", US Patent 3920904, 1973.
[6] Cashion T., Williams S., "Apparatus for creating 3D audio imaging over headphones using binaural synthesis", US Patent 5809149, 1996.
[7] Fidi W., "Stereophony binaural recording or reproduction method", US Patent 5033086, 1989.
[8] Haramoto Y., Fujii Y. Yamazaki M., "Signal processing circuit", US Patent 4139728, 1977.
[9] Iwahara M., "Audio signal translation for loudspeaker and headphone sound reproduction", US Patent 4209665, 1978.
[10] Thomas MV, "Improving the stereo headphone sound image", Journal of the audio engineering society 25/7, pp. 474-478, 1977.
[11] Turner JC, Elliott DM, "Headphone with cross feeding ambience control", US Patent 3924072, 1974.
[12] Wightman_F.L., Kistler DJ, "Headphone simulation of free-field listening. I: Stimulus synthesis", J. Acoust. Soc. At the. 85/2, pp. 858-867, 1989.
[13] Wightman_F.L., Kistler DJ, "Headphone Simulation of free-field listening. I: Psychophysical validation", J. Acoust. Soc. At the. 85/2, pp. 868-879, 1989.
[14] Wightman_F.L., Kistler DJ, "The dominant role of low-frequency interaural time differences in sound localization", J. Acoust. Soc. At the. 91/3, pp. 1648-1661, 1992.
[15] Yamada A., Kimura Y., Kikuchi Y., "Projected sound localization headphone apparatus" US Patent 4087631, 1976.

Claims (4)

1. Electronic circuit with two input channels and two output channels for processing a stereo audio signal, in particular for playback via headphones, characterized in that the audio signals marked with left and right of the stereo audio signal are offered on the two input channels and that the signal on one the output channels form the sum of a signal obtained from the left audio signal by a first filter process F1 with a frequency-dependent amplification and time shift of the signal components, and from a signal obtained from the right audio signal by a second filter process F2 with a frequency-dependent amplification and time delay of the Signal components, and characterized in that the signal on the other output channel forms the sum of a signal obtained from the right audio signal by the same first filtering process F1 and from a signal obtained from the left audio signal by the d same second filter process F2. 2. Electronic circuit according to claim 1, characterized in that the sum of the signals obtained by the filter processes F1 and F2 from the left audio signal is equal to the left audio signal and the sum of the signals obtained with the Filter processes from the right audio signal is the same as the right audio signal. 3. Electronic circuit according to claim 2, characterized in that the circuit contains 3 electrical networks which are daisy chained, the first and the last network having the same electrical impedance, the left stereo signal at the beginning of the chain and the right stereo signal at the end of Chain are offered and the output signals are available at the connection points of the electrical networks. 4. Electronic circuit according to claim 3, characterized in that the first and the last network of the chain from a parallel connection of an ohmic Resistor and a capacitor exist and the middle network through one Resistance is formed.