Apparatus for providing high quality audio output by median filter in audio systems

1. Field of the Invention

The present invention relates to an apparatus for enhancing audio quality in audio systems, specifically, an apparatus for providing high quality audio output by a median filter in audio systems.

2. Description of the Prior Art

Sounds are a fundamental way in which people communicate with others. Regardless, if it is voice or music, all are sent by sounds. As new technologies are developed progressively, sounds remain an important way for people to communicate or relax. Products such as audio systems are important products for people to enjoy music and relax. This is especially true of wireless audio systems. The most convenient way to transmit sounds is via air transmission. However, there are also problems with wireless audio systems, and these problems can arise because audio signals are easily influenced by noise during the wireless transmission process. The distorted signals generate popping sounds, subsequently decreasing acoustic fidelity. Therefore, an important research target is to decrease the effect of distorted signals during the wireless transmission process.

Please refer to FIG. 1 , which is a functional block diagram of a prior art wireless audio system 10 . The wireless audio system 10 includes a transmitting apparatus 12 A and a receiving apparatus 12 B. The transmitting apparatus 12 A is used to transform an audio signal into a radio frequency signal and send the radio frequency signal via air transmission. The receiving apparatus 12 B is used to receive the radio frequency signal and transmit the audio signal, which corresponds to said radio frequency signal. The transmitting apparatus 12 A comprises two sound inputting devices 14 A, 14 B, a parallel/serial converter 16 , an encoder 18 , a burst mode controller (BMC) 19 , a modulation module 20 , and a transmitting circuit 22 . The receiving apparatus 12 B comprises a receiving circuit 24 , a demodulation module 26 , a BMC 28 , a decoder 30 , a serial/parallel converter 32 , two audio conversion devices 34 A, 34 B, and two speakers 38 A, 38 B.

In the prior art transmitting apparatus 12 A, the sound inputting devices 14 A, 14 B have a microphone and an analog-to-digital converter (ADC) installed in them. The sound inputting devices 14 A, 14 B can simultaneously receive two sounds inputted by different audio channels (such as left audio channel or right audio channel). These sounds are recognized as digital data bits (a sample value of each data bit represents an amplitude of the sound) so as to compile sequential digital signals Pa, Pb. The digital signals Pa, Pb are simultaneously transmitted to the parallel/serial converter 16 . The parallel/serial converter 16 can encapsulate the two digital signals Pa, Pb of the two sound inputting devices 14 A, 14 B into a sequential digital signal P 1 and output the digital signal P 1 to the encoder 18 . The encoder 18 adds an error protection code to the digital signal P 1 . The BMC 19 controls the clock of the digital signal P 1 and synchronizes the digital signal P 1 so as to form a digital signal P 2 . The digital signal P 2 is transmitted to the modulation module 20 . The modulation module 20 modulates the digital signal P 2 into an analog baseband signal P 3 which is capable of being transmitted via air transmission. The analog baseband signal P 3 is sent to the transmitting circuit 22 . The transmitting circuit 22 modulates the analog baseband signal P 3 into radio frequency signal P 4 and transmits the radio frequency signal via air transmission.

After receiving the radio frequency signal P 4 ′ (the corresponding received radio frequency signal relates to P 4 )transmitted from the transmitting apparatus 12 A, the receiving circuit 24 transforms the radio frequency signal P 4 ′ into a baseband signal P 5 (the baseband signal P 5 corresponds to the original baseband signal P 3 ) and sends the baseband signal P 5 to the demodulation module 26 . Note that owning to essence of radio transmittion, P 4 ′ may be effected by signal distortion, signal interference, noise, etc. Thus, P 4 and P 4 ′ may not be exactly the same. The demodulation module 26 extracts the digital data P 6 from the baseband signal P 5 . The BMC 28 controls the clock of the digital data P 6 and synchronizes the digital data P 6 so as to generate digital data P 7 . The digital data P 7 corresponds to the original digital data P 2 . The serial/parallel converter 32 splits the digital data P 7 into two digital data Pc, Pd originally identified with the different audio channels. The digital data Pc, Pd corresponding to the digital data Pa, Pb are simultaneously transmitted to audio conversion devices 34 A, 34 B of different audio channels. The audio conversion devices 34 A, 34 B are a digital-to-analog converter (DAC). The audio conversion devices 34 A, 34 B convert the digital signal into analog audio signals Pe, Pf and send the analog audio signals Pe, Pf to the speakers 36 A, 36 B. The speakers 36 A, 36 B transmit the acoustic wave corresponding to the analog audio signals Pe, Pf so users are able to hear the sound.

Please refer to FIG. 2 , which is a clock diagram of the signals of the audio system 10 shown in FIG. 1 . The horizontal axis represents time. The vertical axis of the waveform of the audio signal Pe represents amplitudes. In the transmitting apparatus 12 A of the audio system 10 , an analog audio wave is sampled as digital signals and transformed into an analog radio frequency signal. This analog radio frequency signal is transmitted via air transmission. When the receiving apparatus 12 B receives the analog radio frequency signal, the analog radio frequency signal is reconverted into an analog audio signal. The speakers convert the analog audio signal into an acoustic wave and transmit the acoustic wave so users can hear the sound of the acoustic wave. The single audio channel in audio conversion device 34 A will be used as an example. The digital signal Pc (Pc corresponds to the digital signal Pa of the transmitting apparatus 12 A) uses the one-by-one sequential data samples to represent the amplitude of the radio frequency analog audio signal Pe waveform on each data sample point. As shown in FIG. 2 , a data PS 1 (always consisting of eight bits) within the digital signal Pc corresponds to the amplitude of the audio radio frequency signal Pe waveform at time t 1 . Similarly, another data PS 2 within the digital signal Pc corresponds to the amplitude of the audio signal Pe at time t 2 , and a data PS 8 corresponds to the amplitude of the audio signal Pe at time t 8 . The audio conversion device 34 A is used to sequentically transform data within the digital signal Pc into the amplitude of the analog waveform so as to transmit the audio signal Pe.

However, the abovementioned the analog signals are influenced by other radio signals or noise when the analog signals are transmitted via air transmission. The analog signals are influenced by the multi-path effect, meaning that some distortions may occur in the analog signal. When the distorted analog signal is received by the receiving apparatus 12 B, the corresponding digital signals Pc, Pd may also have some errors. This erroneous information causes the audio conversion device to emit popping sounds. As shown in FIG. 2 , if the data sample PS 8 at time t 8 has a bit error occurrence, the audio signal Pe at time t 8 suddenly appears as a high (or low) impulse so the original smooth audio signal Pe appears to have a suddenly change in waveform. This impulse causes users feel uncomfortable, and decreases the quality of the acoustic fidelity.

In order to prevent the above situation from happening, the prior art technology uses the error protection code to encode the sending signal so as to prevent the error of data. In the transmitting apparatus 12 A, the encoder 18 encodes the error protection code in each data of the digital signal P 1 according to a coding theorem, so as to form the digital signal P 1 ″. When the receiving apparatus 12 B receives the signal with the error protection code, the receiving apparatus 12 B transforms the signal into the digital signal P 7 and transmits the digital signal P 7 to the decoder 30 . The decoder 30 corrects the erroneous bits generated during the wireless transmission process according to the error protection code. See FIG. 2 , there is a corresponding error protection code in each data of the digital signal P 7 . For example, a corresponding error protection code e 1 is added to the data PS 1 , and a corresponding error protection code e 2 is added to the data PS 2 , and so on. The decoder 30 corrects the error of the digital signal according to the error protection code within the digital signal P 7 , so as to obtain the digital signal P 8 . The digital signal P 8 includes the sample value of each data sample. The digital signal P 8 is reconverted into an analog audio signal by the serial/parallel converter 32 and the audio conversion devices 34 A, 34 B.

A primary defect of the prior art is that the prior art wireless audio systems must have complicated encoders and decoders installed. In order to encode the error protection code, the prior art transmitting apparatus 12 A must have the encoder 18 installed and the prior art receiving apparatus 12 B must have the corresponding decoder 30 installed. Since the encoding algorithms and the decoding algorithms are complicated, the related encoder 18 and decoder 30 must have complex circuits. This is especially true for the decoder 30 . The circuit of the decoder 30 is the most complicated of the components in the receiving apparatus 12 B. Therefore, the cost and time of design, production, and maintain of the prior art audio system 10 is increased. Additionally, each data becomes longer after having the error protection code added, thereby increasing the data processing load of the audio system 10 .

It is therefore a primary objective of the present invention to provide an apparatus that uses a median filter to filter out errors in digital signals so as to provide high quality audio output. The prior art encoder and decoder are no longer required in the said invention, thereby decreasing cost of the audio system.

The claimed median filter compares the filtering data with at least one former data and at least one latter data. Abandoning the maximum sample value data and the minimum sample value data so as to obtain the median value data and output the median value data. Therefore, the median filter can efficiently filter out the erroneous data, which will generate the popping sounds. In the embodiment of this invention, the median filter compares the filtering data with one former data and one latter data. The median filter obtains the median value data within the three successive data samples and outputs the median value. This invention will efficiently filter out the erroneous data and prevent the popping sounds, and decrease the cost of the audio system.

Briefly, the claimed invention discloses an apparatus for enhancing audio quality of an audio system. The apparatus comprises a receiving circuit, a demodulation module, a frame synchronization control module, a filter, and an audio conversion device. The receiving circuit is used to receive a radio frequency signal and generate a corresponding baseband signal. The demodulation module is electrically connected to the receiving circuit, and is used to demodulate the baseband signal and correspondingly output sequential data. The frame synchronization control module is electrically connected to the demodulation module, and is used to synchronize the data and outputs sequential data. The filter is electrically connected to the frame synchronization control module, and is used to filter out erroneous data transmitted from the frame synchronization control module. The audio conversion device is connected to the filter for transferring an output of the filter into a corresponding audio signal.

It is an advantage of the claimed invention that said invention does not need the encoder and the decoder installed, as was the case with the prior art. This invention only needs to have the simple and cheap median filter installed. The median filter can efficiently filter out the erroneous data within the digital audio signal, thereby decreasing popping sounds and increasing the acoustic fidelity. The said invention also decreases the cost of the audio system.

These and other objectives of present invention will no doubt become obvious to those of ordinary skill in the art after reading the following detailed description of the preferred embodiment which is illustrated in the various figures and drawings.

FIG. 1 is a functional block diagram of a prior art wireless audio system.

FIG. 2 is a clock diagram of signals of the audio system shown in FIG. 1 .

FIG. 3 is a functional block diagram of the present invention wireless audio system.

FIG. 4 is a clock diagram of each related signal of the present invention apparatus.

FIG. 5 is a functional block diagram of a present invention filter.

Please refer to FIG. 3 , which is a functional block diagram of the present invention wireless audio system 40 . The audio system 40 includes a transmitting apparatus 42 A and a receiving apparatus 42 B. The transmitting apparatus 42 A includes two sound inputting devices 44 A, 44 B, a parallel/serial converter 46 , a frame synchronization control module 49 , a modulation module 50 , and a transmitting circuit 52 . The modulation module 50 includes a modulation circuit 48 A and a spreading circuit 48 B. The receiving apparatus 42 B includes a receiving circuit 54 , a demodulation module 56 , a frame synchronization control module 60 , a serial/parallel converter 62 , two filters 64 A, 64 B for different audio channels, two audio converter devices 66 A, 66 B for the said different audio channels, and two speakers 68 A, 68 B for the said different audio channels. The demodulation module 56 includes a de-spreading circuit 58 A and demodulation circuit 58 B. Both of the sound inputting devices 44 A, 44 B for the said different audio channels, each device has a microphone and an analog to digital converter (ADC) respectively installed for converting the analog audio signal into the digital audio signal. The sound inputting devices 44 A, 44 B can obtain the digital audio signal from other sound sources (such as from a music CD). The speakers 68 A, 68 B can be earphones.

The sound inputting devices 44 A, 44 B generates digital signals Sa, Sb and outputs the digital signals Sa, Sb to the parallel/serial converter 46 . The parallel/serial converter 46 arranges the digital signals Sa, Sb of two different audio channels into a single sequential digital signal and transmits this sequential digital signal to the frame synchronization control module 49 . The frame synchronization control module 49 controls the clock of the digital signal and synchronizes the digital signal so as to form a digital signal S 1 . The digital signal S 1 is transmitted to the modulation module 50 . The modulation circuit 48 A of the modulation module 50 can be a pi/4-DQPSK modulation circuit so as to modulate the digital signal S 1 into a digital signal S 2 . The spreading circuit 48 B performs convolution and multiplication operations on the digital signal S 2 and a spreading code Ss 1 so as to form a baseband signal S 3 . The spreading circuit 48 B can make use of direct-sequence spread spectrum (DSSS). That means each bit of the digital signal S 2 is represented by several bits. The baseband signal S 3 is outputted to the transmitting circuit 52 . The transmitting circuit 52 converts the baseband signal S 3 into a radio frequency signal S 4 and transmits the radio frequency signal S 4 via air transmission.

When the receiving apparatus 42 B receives the radio frequency signal S 4 , the receiving circuit 54 transforms the radio frequency signal S 4 into a baseband signal S 5 and transmits the baseband signal S 5 to the demodulation module 56 . The de-spreading circuit 58 A of the demodulation module 56 performs de-spreading on the baseband signal S 5 (performs the convolution and multiplication operations on the baseband signal S 5 and a spreading code Ss 2 ) so as to generate a digital signal S 6 . The demodulation circuit 58 B performs the inverse operation of the modulation circuit 48 A so as to demodulate the digital signal S 6 into a digital signal S 7 . The digital signal S 7 is transmitted to the frame synchronization control module 60 . The frame synchronization control module 60 controls the clock of the digital signal S 7 and synchronizes the digital signal S 7 so as to generate a digital signal S 8 . The digital signal S 8 is transmitted to the serial/parallel converter 62 . The serial/parallel converter 62 splits the digital signal S 8 into two digital signals Sc, Sd respectively for different audio channels. The filters 64 A, 64 B filter the digital signals Sc, Sd so as to generate corresponding digital signals Se, Sf. Finally, the audio conversion devices 66 A, 66 B respectively transform the digital signals Se, Sf into analog audio signals Sg, Sh and transmit the analog audio signals Sg, Sh to the speakers 68 A, 68 B. The speakers 68 A, 68 B transmit the acoustic wave corresponding to the analog audio signals Sg, Sh. The audio conversion devices 66 A, 66 B can be digital to analog converters (DACs). In addition, it is noteworthy that each of the frame synchronization control modules 49 , 60 can be a burst mode controller (BMC).

FIG. 3 shows that the primary difference from the prior art is that the present invention uses the filters 64 A, 64 B instead of the prior art encoder and decoder so as to filter out the erroneous data of the different audio channel digital signals Sc, Sd. The present invention uses simple median filters to be the filters 64 A, 64 B. Please refer to FIG. 4 . FIG. 4 is a clock diagram of each related signal of the present invention apparatus. The horizontal axis of FIG. 4 represents time. The following uses the filter 64 A as an example so as to illustrate the operating principle of the median filter. The operating principle of the filter 64 B is same as that of the filter 64 A. Similar to the prior art receiving apparatus 12 B, the present invention receiving apparatus 42 B also uses the sequential data of the digital signal to represent the amplitude value (sample value) of the analog waveform on each data samples. The analog waveform corresponding to the digital signal Sc, which inputted to the filter 64 A, is a waveform Wc shown in FIG. 4 . The vertical axis represents the amplitude of the waveform Wc. The analog waveform corresponding to the digital signal Se, which is processed by the filter 64 A, is a waveform We shown in FIG. 4 . The vertical axis represents the amplitude of the waveform We.

As shown in FIG. 4 , each data (always is eight-bit length) within the digital signal Sc corresponds to the sample value of the waveform Wc on each data sample. A data D 1 corresponds to the amplitude of the waveform Wc at time t 1 . A data D 2 corresponds to the amplitude of the waveform Wc at time t 2 . Data D 3 , D 4 and D 7 , D 8 , D 9 simultaneously correspond to the amplitudes of the waveform Wc at times t 3 , t 4 and t 7 , t 8 , t 9 . The relation between the digital signal Se and the waveform We is similar to the relation between the digital signal Sc and the waveform Wc. The abovementioned radio signal is influenced by noise during the transmission process. Therefore the digital signal Sc received and processed by the receiving apparatus 42 B will carries erroneous data, so that the sound outputted from the speakers has popping sounds. For example, the data D 8 within the digital data Sc is erroneous data. This erroneous data causes the waveform Wc to have a protruding wave at time t 8 . The filter 64 A uses the function of the median filter to filter out the erroneous data within the digital signal Sc so as to generate the digital data Se. In the present embodiment, when the median filter wants to update a protruding data, the median filter uses the median value data of three successive data samples (the data itself, and the former data of the data, and latter data of the data) instead of the original data. That means the data that has the maximum sample value or the data that has the minimum sample value are replaced by the data with median sample value, so as to filter out the erroneous data. For example, when the filter 64 A processes the data D 2 corresponding to time t 2 , the filter 64 A compares the value of the data D 1 , D 2 , D 3 (corresponding to time t 1 , t 2 , t 3 ). That means comparing the sample values of the three data samples of the waveform Wc at times t 1 , t 2 , t 3 . The waveform Wc shows that the amplitude at time t 2 is between the amplitude at time t 1 and t 3 . Regarding the data D 2 of the digital signal Sc, the median filter still outputs data D 2 in digital signal Se. After finishing processing the data D 2 at time t 2 , the median filter processes the data D 3 corresponding to time t 3 within the digital signal Sc. At this time the median filter compares the value of data D 2 (the former data), D 3 and D 4 (the latter data). After comparing, the median filter outputs the median value data D 3 to the digital signal Se. Then the median filter continues and processes the data D 4 corresponding to time t 4 within the digital signal Sc. The analog signal is sampled as digital signal, the sampling frequency is usually higher than a Nyquist frequency of the analog signal. That means an interval between the data samples is very small. The sample values of two neighboring data samples do not have large change. In normal situations, if there is no erroneous data within the audio signal, the filtering data is equal to the median value when it is compared with the former data and the latter data. With regards to the waveform Wc and waveform We shown in FIG. 4 , there is no erroneous data before time t 7 , meaning that the waveform Wc is same as the waveform We.

When the median filter processes the data D 7 corresponding to time t 7 within the digital signal Sc, the filter compares the data D 6 , D 7 , D 8 corresponding to time t 6 , t 7 , t 8 . Since there is no erroneous data, the filter still sends the data D 7 in the digital signal Se. Then the median filter processes the data D 8 at time t 8 within the digital signal Sc. The median filter compares the data of D 7 (the former data), D 8 , D 9 (the latter data). After comparing, the media filter transmits the median value data D 7 to the digital data Se. Therefore the data within the digital data Se at time t 8 is changed to D 7 , but not the original data D 8 within the digital data Sc. Thus, the erroneous data D 8 corresponding to time t 8 within the digital signal Sc is filtered out by the median filter. The median filter continues to process the data D 9 corresponding to time t 9 within the digital signal Sc and transmits the median value data D 9 to the digital signal Se. The waveform Wc and waveform We shown in FIG. 4 show that the median filter really can filter out the erroneous data from the audio signal so as to make the waveform much more smooth. The filtered digital signals Se are transmitted to the audio conversion device 66 A. The audio conversion device 66 A transforms the digital data Se into the audio signal and transmits the audio signal to the speaker 68 A. The speaker 68 A transmits the acoustic wave corresponding to the audio signal. Since the erroneous data has been filtered out by the median filter, users will no longer hear the popping sounds.

In conclusion, if the data samples do not have erroneous data, the sample values of two successive data samples do not have large change. The filtering data is the same as the median value data when comparing the filtering data with the former data and the latter data. In this situation, the median filter maintains the original waveform. However, when the sample value of one data sample suddenly becomes higher or lower, that means this data sample is an erroneous data. In this invention, the erroneous data is not the median value data when comparing with the former data and the latter data. The median filter chooses the former data or the latter data instead of this erroneous data so as to make the waveform of the output signal much more smooth, thereby preventing the popping sounds.

Please refer to FIG. 5 , which is a functional block diagram of the present invention filters 64 A, 64 B. The median filter will be used as an example. The median filter 66 shown in FIG. 5 has three delay units 70 . The function of each delay unit is to performz −1 operation. The three delay units 70 can obtain three successive data samples from the inputted digital signal. The three successive data samples are transmitted into the median value selector 68 so as to choose the median value data and output the median value data. Since the probability that two successive data samples both contain erroneous data is very small, the present invention median filter which compares three successive data samples and outputs the median value data can efficiently filter out the erroneous data. Of course, the present invention median filter can also use a median filter which compares five (or more) successive data samples. The median filter which compares five successive data samples, compares the data itself, the former two data, and the latter two data so as to obtain the median value data. In which, this median filter has five delay units.

In contrast to the prior art audio system which uses the complicated encoder and decoder to add the error protection code so as to filter out the erroneous data, the present invention audio system uses the simple median filter to filter out the erroneous data. The transmitting apparatus of the present invention wireless audio system does not need the encoder installed, and the receiving apparatus also does not need the decoder installed. The present invention only needs two simple and inexpensive median filters installed for different audio channels so as to efficiently filter out the erroneous data within the digital signal, thereby decreasing the occurrence of popping sounds and increasing the acoustic fidelity. The present invention can be used not only in wireless audio systems which have frequency bands between 2.4 GHz to 2.5 GHz, but also can be used in frequency bands between 5.15 GHz to 5.35 GHz. Since these frequency bands are commonly used by people, these signals are easily influenced by noise. The present invention can efficiently filter out the erroneous data generated during the transmission process with low cost, and decrease the popping sounds. Since the wireless transmission signals do not need to have error protection codes added, the load of the wireless transmission is decreases. The abovementioned embodiment used the wireless audio system as an example. However, the present invention is not limited to that. The present invention can be used in general digital audio systems to filter out the erroneous data within digital signals so as to increase the acoustic fidelity.

Those skilled in the art will readily observe that numerous modifications and alterations of the apparatus may be made while retaining the teachings of the invention. Accordingly, the above disclosure should be construed as limited only by the metes and bounds of the appended claims.