TWI342718B - Decoder and method for deriving headphone down mix signal, receiver, binaural decoder, audio player, receiving method, audio playing method, and computer program - Google Patents

Description

IX. INSTRUCTIONS: [Technical field of the invention] The present invention relates to decoding a multi-channel audio signal based on a parameter multi-channel representation', and particularly related to generating a headset such as a headphone compatible or a space downmix Configure the space to listen to the experience of the 2 channel downmix. %1 [Prior Art] The recent development of audio coding is the multi-channel representation of reconstructing audio signals based on stereo (or mono) signals and control signals on the corridor. Since the additional bedding is transmitted to control the surround channel based on the transmitted mono or stereo audio track, which is also referred to as upmixing, these methods are essentially older than the = Dolby Pr〇L〇gic matrix. The solution is different. Therefore, the parameter multi-channel audio decoder, such as MPEG Surround, reconstructs the ^^ channel based on the broken transmission channel and additional control data, where ν > μ. The additional control data indicates that the transmission rate is significantly lower than the transmission rate of all ν channels to make the coding very efficient while ensuring compatibility between the stern and the Ν channel. , k some parameters Weibian financial method material contains inter-channel key difference _ or channel bit difference (CLD) and frequency phase-induced (ICQ silk-based surround signal parameterization of some parameters to indicate the channel pairing in the upmix processing The power is related to this. The parameters of the previous technical towel are also included in the prediction parameters of the intermediate media or the output channel. The other developments have been provided using the stereo sound headphones. A device that listens to the effect. In order to use only two of the headphones, eight 1342718

For the spatial listening experience, the multi-channel letter sharp is downmixed into stereo sound using the head (four) transition (HRTF) of the human head's extremely complex transmission characteristics that are expected to provide space speech experience. Another-related method uses the traditional 2-channel playback environment and the channel of the wave channel of the multi-channel audio signal to achieve close proximity to the original (four) read playback. Lai processing side material machine touch construction, the appropriate space of the characteristics of the stereo sound down, '. Compared with the earphone case, the two-port eight, the signal directly deny the listener's ears, shouting the non-turning effect, _ sound effect, when = must consider the most important heavy product quality, the filter that is processed by the money ^ often called It is a crosstalk canceller. Usually, this technique aims to extend the possible range of stereo sounds by using the chirp 4 division filter 1 § to eliminate the inherent crosstalk. Because of the composite filtering, the head related transfer function filter Very long, that is, it may contain hundreds of filter taps. For the same reason, when the actual filter is used instead, it is almost impossible to find filter parameters that work well enough to degrade the sensory quality. On the one hand, the multi-channel signal saving parameter representation that facilitates efficient transmission of encoded multi-channel signals does not exist. On the other hand, only an elegant way of creating a multi-channel signal space listening experience using stereo earphones or stereo sounds is known. However, these need to be used as an application to create a multi-channel signal channel for the input of the head-related transfer function of the headphone under the headphone. Therefore, the application Before the relevant transfer function or crosstalk cancellation filter, the full set of multi-channel signals must be transmitted or the parameter representation must be completely reconstructed, so the transmission bandwidth or computational complexity is unacceptably high. 8 1342718 [Summary] Listen SUMMARY OF THE INVENTION The object of the present invention is to enable the material channel signal parameter to represent the concept of a more efficient reconstruction of a 2-channel signal that provides a spatial experience.

π according to the first feature of the present invention, this object is achieved by using a multi-channel signal, and using a level parameter of the information of the level relationship between the two channels having the multi-channel signal, and two channels of the channel signal of the shirt channel. The related head related transfer function derives the headphone downmix signal-decoder to achieve, including: - the wave calculator, which can derive the corrected head related transfer function by weighting the head related transfer function of the two channels using the level parameter. The influence of the copper phase _ shift function of the channel with the high channel 2 channel _ shift function Long Hong chat side transfer letter and the money channel is stronger; and Synthesizer, can use the = positive head side transfer function and τ mixed domain wire to export the under-headphones

The second feature of nr month, the purpose is to enable the channel signal to be down-mixed by the stereo decoding decoder 11 ', and the level between the two channels of the multi-channel signal _ information To derive the head related transfer function related to the two channels of the tiger; the head related transfer function derives the number of correction heads; the head related transfer function of the channel of the same level related to the modified head; Function '. to make. . 'This correction head related transfer function can be used and 卞 9 5 1342718 1342718

And an integrated filter bank headphone signal. Wave signal meter out the headphone downmix signal; - minute (four) grouping, can be downmixed by the sub-band wave multi-channel signal downmix to lead the channel signal down; can be derived from the integrated headphone downmix signal to derive the time domain according to this According to a third feature of the invention, the purpose of the method is to use a multi-channel signal to downmix the table to determine the level relationship between the two channels of the channel signal and to use the position of the afi parameter and the head of the two channels of the ship Turn off the transfer function to the feeder downmix signal, including: H can be used to derive the corrected head related transfer function by weighting the head related transfer function of the two channels using the level parameter to make the face side transfer function pair of the ship with high level The effect of the modified head related transfer function is stronger than that of the modified head phase shift function. The modified head related transfer function and the downmix signal are used to derive the effect. Under the headphones, the letter is green. According to a fourth feature of the present invention, the object is achieved by a receiver or an audio player, comprising: a decoder capable of using a multi-channel signal to smash == using a positional spray between two channels having a multi-channel signal : , and use the head-related transfer function related to the two channels of the multi-channel signal to derive the headphone downmix signal, including: "to pass" = fine by the secret parameter weighting the two channels of the __ function, the Ministry: the transfer The function 'make the head-related transfer function of the channel with a high level have a stronger influence on the phase shift function of the modified head than the head phase shift function of the channel with a lower level; and - The synthesizer can use the correction header: = 1342718 and the downmix signal to derive the headphone downmix signal. According to the fifth feature of the invention, the object is by a receiving or audio = method ίί: and ίί:: Exhaustion, you can use the multi-channel signal to downmix the position loss ^ the channel between the two channels of the relationship between the two devices, the 5 tiger, the method includes: by using the bit transfer function to make the relevant transfer Function export correction header related to this _ Phase_function:::=== Headphone downmix signal. °_Transfer function and downmix signal indicate the derivation of the sixth special downmix representation according to the invention, and the use of the =2 system by using the information level Participate in the age,; "The relationship between the two channels of Xi Xi, 仏唬 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除 消除Elimination of Newcomer's secret parameter weighting two-channel elimination filter is equipped with a high (four) 2 day > Xiao filter, wave filter, so that the correction of the level of the crosstalker depends on the channel of the voice cancellation waver, The correction can be used to have a stronger influence; and - the inter-synthesis stereo τ mixed signal. The neon 11 and the down-mixed semaphore output null header are used to derive the 4 positive head correlation from the original multi-channel signal. When transferring functions, and when filtering < s >

Claims (1)

1342718 Revised replacement page to apply for a patent circle: • - Decoding to use multi-channel signal downmix representation; multi-channel signal between two channels of positional information: quasi: shift function and (four) - display downmix domain, including correction Correlation transfer; =:, it is weighted by the level parameter and the phase factor is applied to the head related transfer function of the channel of 6 hai, and the positive phase = transfer function is used to make the correction head have a high level ^ = function The effect of the related transfer function:: The low-level error of the solution side function _ correction related transfer function has a stronger influence, and the phase compensation of the head two-off transfer function of the two channels in the two channels The weighting is achieved before the phase compensation head phase_shift function is combined; and the - synthesizer' uses the miscellaneous phase transfer function and the downmix k number table to derive the headphone downmix signal. 2. The Republic of China 100 2 曰 曰 曰 = = = = = = = = = = = = = = = = = = = = = = = = = = = = 专利 专利 专利 专利 专利 专利 专利 专利 专利 专利 专利 专利The correlation transfer function is such that the master phase shift function of the channel having a low level is shifted closer to the average phase of the two correlation transfer functions than the channel having the -high level. A decoder as claimed in claim 1 wherein the filter calculator is operative to cause the number of derived header related transfer functions to be less than the number of associated header related transfer functions of the two channels. 34. 1342718 A decoder for correcting a face (such as the scope of claim patent), which causes the chopper calculator to derive a modified head related transfer function applied to a set of the wave group of the downmix signal 5. The decoder of claim 2, for decoding the downmix signal derived from a chopper group domain. 6. The decoder of claim i, wherein the chopping calculator is operative to derive a modified head related transfer function using a head related transfer function characterized by more than two parameters. 7. The decoder of claim 1, wherein the filter calculator is operative to derive a head related transfer function weighting factor for the two channels using the same level parameter. 8. The decoder of claim 7 of the patent scope, the Yin Chopper calculator is used to derive the first using the level parameter CLDi according to the following formula a first weighting factor Wlf of the channel f and a second weighting factor w of the second channel s, s, ... 2 _ 1 〇 α, 1 9. The decoder of claim i, wherein the chopper calculator The modified head related transfer function is derived by operating an applied-gain factor to the head related transfer function of the two channels to preserve energy when the modified head related transfer function is derived. 10. The decoder of claim 9, wherein the gain factor is within the interval [1/H]. 11. The decoder of claim 2, wherein the filter calculator 35 1342718 is revised in the Republic of China, 迓 P P is replacing the page system operation using the head related transfer function _ response of the two channels A delay between the time to derive the phase parameter. 12. Please ask for the solution of U of the patent scope, its Weibo calculator: _ repair has L fresh sense, money delay. • Time to derive the phase parameters of each frequency band. 3. The decoder of claim 11, wherein the chopper calculator operates in an age-receiving domain having more than two fresh numbers, and derives each frequency band index using the delay time Γχ γ according to the following formula Each of the phase parameters φ χ γ : 14. The decoder of claim 2, which causes the chopper meter to operate using a normalized composite of the head-related transfer function impulse responses of the first and second channels The phase angle is derived by correlating the associated phase angle. 15. As claimed in claim i, the first channel of the two channels is the left or right front channel of the multichannel signal, and the second channel of the two channels is one of the same side After the channel. 16. (4) The decoder of item 15 of the special fiber circumference, wherein the Weibo calculator operation uses the following complex linear combination to transfer the head related transfer function HY(Xf) from the front channel and the head of the subsequent channel. ^Hy(Xs) derives the modified head related transfer function Ηγ(χ): Meaning (4)=W exp(-K(6)(8)+ W exp(Kzhen where φχγ is the phase parameter '% and ~ is used to use the level parameter 1342718 In October 1999, η had corrected the weighting factor for the correction of the replacement page, and g is the gain factor derived using the level parameter. 如. The decoder of claim 1 is used to use the mixed signal. The downmix signal has a left and a right channel derived from one of the multichannel signals and the multichannel signal has a left front, a left loop, a right front, a right surround, and a central channel. The decoder of clause 1, wherein the synthesizer operates to apply a linear combination of a modified header related transfer function to the downmix representation of the multichannel signal to derive a channel of the headphone downmix signal. Decoding of the first item Wherein the synthesizer operates to use the coefficients of the linear combination according to the level parameter. 20. The decoder of claim 18, wherein the synthesizer operates according to an additional spatial characteristic of the multichannel signal The multi-channel parameter is used to use the coefficient of the linear combination. 21. A stereophonic decoder comprising: a decoder as in claim 1 of the patent application; an analysis filter bank for using the multi-channel The signal is mixed under the sub-band filter to extract the channel signal under the mixed expression; and - the integrated filter m is used to derive a time domain headphone signal by synthesizing the headphone downmix signal. 22· - a kind of decoder, which uses more than one One of the channel signals is downmixed, the level parameter of the level information between the two channels having the multichannel signal is used, and the canceling wave n associated with the two channels of the multichannel signal is used to derive the spatial stereo downmix The signal, comprising: 37 1342718 _ - the wave calculation ϋ, is used to guide the crosstalk cancellation chopper by weighting the _ level parameter and applying a phase factor to the two channels Correcting the crosstalk canceling filter 'make the crosstalk canceling chopper with a high level channel affect the modified crosstalk canceling chopper. The crosstalk canceling waver with the low level channel corrects the crosstalk The effect of the cancellation filter is stronger, and the phase compensation of the crosstalk cancellation filter of the two channels is achieved before the combination of the weighted and phase compensated crosstalk cancellation of the two channels; and - synthesis, which makes _Correcting the crosstalk cancellation_waveper and the downmix signal representation to derive the spatial stereotone downmix signal. 23·=A method for deriving a headphone downmix signal, which uses a multichannel signal downmix representation and use Having the level-to-information between the two channels of the multi-channel signal, the number of revolutions, and the head related transfer function associated with the two channels of the multi-channel signal to derive the headphone downmix signal 'This method includes: by using The level parameter is weighted and a phase factor is applied to the head related transfer function of the two channels to derive a modified head related transfer function to correlate the head of the channel with a high level The effect of the shift function on the modified head related transfer function is better than the influence of the head related transfer function of the channel having the low level on the modified head related transfer function = strong ▲ ' and the head of the two channels is related to the transfer The phase compensation of the function is achieved before the combination of the weighting of the two channels and the phase compensation head related transfer function; and the correction of the head related transfer function and the downmix signal is used to derive 38 1342718 Republic of China 100 February 15 correction Correct the replacement page for the headphone downmix signal. A receiver for receiving a signal, wherein the receiving device has a decoder for deriving a headphone downmix signal as described in one of claims 1 to 2. An audio player for playing an audio signal, wherein the audio player has a decoder for deriving a headphone downmix signal as described in one of claims 2-3 . A method of receiving a signal, wherein the method of receiving has a method for deriving a headphone downmix signal as described in claim 23 of the patent application. 27. A method of audio playing an audio signal, wherein the method of audio playback has a method for deriving a headphone downmix signal as described in claim 23 of the patent application. 28. A computer program that, when executed on a computer, has a code for executing the food (four) code of items 23 and 26 to 27 of the Shenjing patent. 39 1342718 9f3 month j correction replacement page 4/5 7th picture Output signal r214 Fig. 8 306 /312 Level parameter 7 Downmix signal table 彔 -304 Headphone downmix signal f 308- ym r310 HRTF-J Filtering calculator Ib Γβ > M Fig. 9 408 406 '410 Analytical Wave Decoder — > Integrated Filter Set =i —ϊ -· Wave Set Lb〆 •Rb 400