Realize the method for Kara OK function in the PMP player
[technical field]
The present invention relates to a kind of PMP player system, relate in particular to the method that realizes Kara OK function in a kind of PMP player.
[background technology]
PMP player (Portable Media Player, i.e. " portable electronic device ").Become and use portable electronic product widely,, be easy to carry because its volume is little, characteristics such as music enjoyment anywhere or anytime, the people in the life that become a reality particularly receive liking of a young trend generation.But, only have the paces that music playback function still can not be caught up with modern fast pace life, can not satisfy the initiatively music experience of participatory of people, therefore, the PMP player with Kara OK function also is born thereupon.The Karaoke of this portable makes the user can enjoy the sensation of active participation music whenever and wherever possible, practices phonograph but also can work as, and further improves the performance level of oneself.
But existing PMP player with Kara OK function still has weak point; Accompaniment music must be a file of changing or download special format through exterior PC software; It is convenient anywhere or anytime to bring to the user; Simultaneously, single K song pattern can not let user experiencing different KTV sensations.For the user of Karaoke, what not only need is Kara OK function, and what more need is to experience the enjoyment that different K song pattern is brought.
United States Patent (USP) discloses for 6690799B1 number with voice and has eliminated relevant " the processing stereo signals equipment " of circuit; It is the method that the traditional voice of utilization is eliminated; Promptly when vocal component is positioned at the central passage of sound signal; To eliminate vocal component, obtain simultaneously that the signal of eliminating behind the voice adds low frequency and the radio-frequency component of primary sound and the last background sound that obtains, with the difference between L channel and the R channel when voice is not positioned at the center; Perhaps the responsive frequency band of left and right acoustic channels and people's ear is when to be 200Hz to the signal of 4KHz close, and this signal just output is compared very stiff sound with original sound.
Chinese patent 200610159501.X number, disclose with karaoke OK system in relevant " being used for eliminating the equipment and the method for the vocal component of sound signal ".Its method comprises: postpone a left side and R channel input signal; And the left side that postponed and R channel input signal feedovered to intersect with right and L channel input signal mix; With elimination first vocal component from a left side and R channel input signal, thereby generate a left side and right-channel signals; Postpone a said left side and right-channel signals, and the left side that postponed and right-channel signals feedovered to intersect with right and left channel signals mix, widening sound field, thereby generate second left side and right-channel signals; With to said second a left side and right-channel signals carry out LPF; And the high fdrequency component through the difference between the left side of LPF and right-channel signals and second left side and the right-channel signals of low frequency component mixed with the elimination vocal component, thereby generate a final left side and right-channel signals promptly to generate final accompanying music.But the frequency range of the signal of the final output of this voice removing method has only high fdrequency component to follow low frequency component, and the responsive frequency range of go-between's ear seldom or eliminate having influenced accompaniment band effect, when also existing voice not to be positioned at the center simultaneously, and the unfavorable problem of effect.
Therefore, the voice removing method that relates in above-mentioned two patents can obtain good effect at some audio file, does not have the ubiquity of consideration, and the residual people's of meeting labial was with dental after high-grade voice was eliminated.Simultaneously, the needs that noise reduction grade, reverberation pattern can not satisfy different crowd on the market are not set.
[summary of the invention]
The technical matters that the present invention will solve; Be to provide the method that realizes Kara OK function in a kind of PMP player; Not only can realize Kara OK function, also be provided with the noise reduction grade, and can utilize common audio-video document to produce the accompaniment band; And can play up different K song scenes to the user, increased the usability of Kara OK function greatly.
The present invention is achieved in that the method that realizes Kara OK function in a kind of PMP player; It is characterized in that: the obtaining step and the voice reverberant sound obtaining step that comprise the music background sound that carries out simultaneously; And the mixed step that contracts of music background sound that carries out afterwards and voice reverberant sound, wherein
The obtaining step of said music background sound comprises:
Step 110, obtain the decoded audio stream digital signal of audio-video document in the PMP player;
Step 120, the voice of the decoded audio stream digital signal in the step 110 is eliminated;
Step 130, the decoded audio stream digital signal in the step 110 is carried out former song low-frequency component keep;
The audio stream digital signal that the low frequency of step 140, the audio stream digital signal after the voice of step 120 eliminated and step 130 kept is mixed to contract and is obtained the music background sound;
Said voice reverberant sound obtaining step comprises:
Step 210, obtain user's voice digital audio and video signals through Mike;
Form the voice reverberant sound behind step 220, the voice digital audio and video signals process reverberation digital signal processing algorithm;
The mixed step that contracts of said music background sound and voice reverberant sound is specially:
Step 300, the voice reverberant sound of music background sound in the step 140 and step 220 is mixed the back of contracting form output audio signal.
In the above-mentioned technical scheme,
The obtaining step of said music background sound also comprises after said step 120:
Step 150, obtain final background music after the music background sound strengthened at least a processing that wave filter and labio-dental sound eliminate through low frequency signal.
Said specifically is to carry out the grade noise reduction of left and right sound channels simultaneously in step 120, and wherein, the grade noise reduction of said R channel comprises:
The left channel signals anti-phase of step 121, decoding stereo audio stream signal with 110;
Step 122, the inversion signal in 121 multiply by a noise reduction factor, the value of this noise reduction factor is between 0~1, is worth greatly more, and grade is high more;
Step 123, the R channel addition of 122 sound signals and 110 sound signals is obtained the right-channel signals after the noise reduction;
The grade noise reduction of said L channel comprises:
The right-channel signals anti-phase of step 125, decoding stereo audio stream signal with 110;
Step 126, the inversion signal in 125 multiply by a noise reduction factor, the value of this noise reduction factor is between 0~1, is worth greatly more, and grade is high more;
Step 127, the L channel addition of 126 sound signals and 110 sound signals is obtained the left channel signals after the noise reduction.
In high-grade noise reduction pattern, be with caused dental and labial after the rejection filter elimination noise reduction.
In the said step 210, user people's acoustical signal of obtaining through Mike be through the A/D cell translation become to have with said step 110 in the audio stream digital signal equate people's acoustical signal of sampling rate.
Reverberation in the said step 220 is that the K song pattern in simulation hall, the middle Room, the little Room is carried out reverberation.
Output audio signal in the said step 220 is to do sound and mix the back acquisition of contracting with the ratio of wet sound through being provided with.
The present invention has following advantage: not only can realize Kara OK function; Also be provided with the noise reduction grade; And can utilize common audio-video document to produce accompaniment band (realizing) through step 120 to step 150; And can play up different K song scenes to the user, increased the usability of Kara OK function greatly.
[description of drawings]
Combine embodiment that the present invention is further described with reference to the accompanying drawings.
Fig. 1 is a FB(flow block) of realizing the inventive method.
Fig. 2 is the system chart of the inventive method.
Fig. 3 a and Fig. 3 b are the grade noise reduction steps flow chart block diagrams of realizing the inventive method left and right sound channels.
Fig. 4 is that the inventive method produces accompaniment band implementation structure synoptic diagram.
Fig. 5 is a reverberation modular structure synoptic diagram of realizing the inventive method reverberation step.
[embodiment]
Seeing also shown in Figure 2ly, is the system chart of realizing the inventive method.This system comprises that audio-video document decoder module 1, voice cancellation module 2, low frequency keep module 3, first and mix the sound module 4 that contracts, post-processed module 5, sound collection modular converter 6, reverberation processing module 7 and second and mix the sound module 8 that contracts.Said audio-video document decoder module 1 keeps module 3 through said voice cancellation module 2, low frequency respectively and contracts with said first mixing and be connected to said post-processed module 5 again after sound module 4 is connected; Be connected to said second in the lump with said post-processed module 5 after said sound collection modular converter 6 is connected with said reverberation processing module 7 and mix the sound module 8 that contracts.
As shown in Figure 1; Realize the method for Kara OK function in the PMP player of the present invention; The obtaining step 100 and voice reverberant sound obtaining step 200 that comprise the music background sound that carries out simultaneously; And the mixed step 300 that contracts of music background sound that carries out afterwards and voice reverberant sound, please combine illustrated in figures 1 and 2ly, the obtaining step 100 of said music background sound comprises:
Step 110, obtain the decoded audio stream digital signal of audio-video document in the PMP player, the sound signal of left and right sound channels through said audio-video document decoder module 1.
Step 120, eliminate through the voice of the decoded audio stream digital signal in 2 pairs of steps 110 of said voice cancellation module.
Step 130, the decoded audio stream digital signal that keeps in 3 pairs of steps 110 of module through said low frequency are simultaneously carried out former song low-frequency component reservation.This step is an optional step, under the companion finds pleasure in the situation that is low frequency, needs choosing.It is the signals that the binaural audio signal that said audio-video document decoder module 1 produces kept the low-frequency component in the primary sound through low-pass filter that said low frequency keeps module 3.
Step 140, through said first mix audio stream digital signal and step 130 after the sound module 4 that contracts is eliminated the voice of step 120 the audio stream digital signal that kept of low frequency mix to contract and obtain the music background sound.It mainly is said voice cancellation module 2 to be eliminated binaural signal and said low frequency behind the voice keep module 3 and carry out mixing that the binaural signal of low frequency reservation the carries out corresponding sound channel sound that contracts, and guarantees that mixed signal can not produce saturated overflowing.If under the not high situation of accompaniment background sound quality requirements, said first mix sound module 4 outputs of contracting signal can regard accompaniment tone.If dissatisfied, can further carry out the processing in later stage to existing accompaniment tone.
Step 150, obtain final background music after through post-processed module 5 the music background sound being strengthened at least a processing that wave filter and labio-dental sound eliminate through low frequency signal.The post-processed that said post-processed module 5 is carried out has comprised the low frequency reinforcement and has weakened labio-dental sound and disturb two modules that these two modules can be done any combination, can only carry out wherein arbitrary module, and perhaps two modules are all carried out.According to people's ear apperceive characteristic; Can know that people's ear is relatively responsive to low frequency signal, so low frequency strengthens purpose and is to cover by voice and eliminates the sound signal radio-frequency component loss that causes, after voice is eliminated; Some residual people's of song meeting dental is with labial; Can cause in the accompaniment tone noise with " neighing ", at this moment, the essential interference of noise that reduces.
Said voice reverberant sound obtaining step 200 comprises:
Step 210, obtain user's voice digital audio and video signals through said sound collection modular converter 6 usefulness Mikes, this step be through user people's acoustical signal that Mike obtains be through the A/D cell translation become to have with said step 110 in the audio stream digital signal equate people's acoustical signal (single, double sound channel also can) of sampling rate.
Step 220, through said reverberation processing module 7 with the voice digital audio and video signals through forming the voice reverberant sound behind the reverberation digital signal processing algorithm, the K song is carried out in simulation under different scene.Reverberation in this step is that the K song pattern in simulation hall, the middle Room, the little Room is carried out reverberation, and its output audio signal is to do sound and the ratio of wet sound and mix and export after contracting again through being provided with.
The mixed step 300 that contracts of said music background sound and voice reverberant sound is specially:
Step 300, said second is mixed the sound module 8 that contracts the voice reverberant sound of the music background sound in the step 150 and step 220 is mixed the formation output audio signal afterwards that contracts.
Step 120 described in the inventive method specifically is to carry out the grade noise reduction of left and right sound channels simultaneously; Multiply by the noise reduction factor after its left side (right side) sound channel anti-phase that said audio-video document decoder module 1 is produced through said voice cancellation module 2; Carry out addition with the original right side (left side) sound channel, eliminate the right side (left side) channel audio signal behind the voice by the time:
See also shown in Fig. 3 a, the grade noise reduction of this R channel comprises:
The left channel signals anti-phase of step 121, decoding stereo audio stream signal with 110;
Step 122, the inversion signal in 121 multiply by a noise reduction factor, the value of this noise reduction factor is between 0~1, is worth greatly more, and grade is high more;
Step 123, the R channel addition of 122 sound signals and 110 sound signals is obtained the right-channel signals after the noise reduction;
See also shown in Fig. 3 b, the grade noise reduction of this L channel comprises:
The right-channel signals anti-phase of step 125, decoding stereo audio stream signal with 110;
Step 126, the inversion signal in 125 multiply by a noise reduction factor, the value of this noise reduction factor is between 0~1, is worth greatly more, and grade is high more;
Step 127, the L channel addition of 126 sound signals and 110 sound signals is obtained the left channel signals after the noise reduction.And in high-grade noise reduction pattern, be with caused dental and labial after the rejection filter elimination noise reduction.
See also shown in Figure 4 again; Be that the inventive method produces accompaniment band (being that above-mentioned steps 120 is to step 150) implementation structure synoptic diagram; Among the figure; Plus sige be mix contract, PR is phase inverter, LPF is low-pass filter, LFE is that low frequency is strengthened wave filter, BSF is a rejection filter, wherein frame of broken lines is a post-processed module 5.The method that following specific explanations accompaniment tone produces because of the stereo two-channel that exists, all is identical and produce the disposal route of in the accompaniment tone two sound channels being carried out, and is example with the L channel at this:
Lin is decoded left channel audio signal among the figure, and it is pulsed modulation (PCM) signal.The Lin signal will divide three the tunnel to walk, and one the tunnel is to carry out the low-frequency component that LPF filtering obtains the Lin signal, and one the tunnel is to multiply by noise reduction factor G2 again and R channel is eliminated voice reference signal Lre*G2 is provided through obtaining Lre behind the PR phase inverter, the one tunnel keep straight-through.Then the reference signal Rre*G1 of through connect signal and R channel is mixed to contract obtaining L, low frequency composition signal is being added above-mentioned L obtains the left channel signals L1 of preliminary elimination voice.Again the L1 signal is carried out post-processed, idiographic flow be the L1 signal through the LFE low frequency strengthen wave filter with or the BSP rejection filter get the left channel signals of accompaniment tone to the end.Wherein LFE and BSP are optional.
Fig. 5 is a reverberation modular structure synoptic diagram of realizing the inventive method reverberation step.Among the figure, A is the sound signal of voice, and A1, B1 are the entrance and exit of hall pattern, and A2, B2 are the entrance and exit of middle Room pattern, and A3, B3 are the entrance and exit of little Room pattern, and B is the sound signal after the reverberation.
According to user-selected K song pattern, system can send into people's acoustical signal A the reverberation algorithm of associative mode automatically, specifically can adopt disclosed some reverberation algorithm in the existing document; Like " the reverberation algorithm of a kind of Schroeder " (Ceng Rong; The Master degree candidate of Central China University of Science and Technology academic dissertation, in April, 2003, P21-P26); Under the prerequisite of ensuring the quality of products, select the algorithm of low operand to be easy in the PMP player system, realize.