CN103402158B - Dimensional sound extension method for handheld playing device - Google Patents

Abstract

The invention discloses a dimensional sound extension method for a handheld playing device. An acoustic transmission filter obtained through head-related transfer function design carries out filtering on input ordinary dimensional sound signals, and the ordinary dimensional sound signals are converted into replaying signals of a pair of narrow-opening-angle and near-listening-distance loudspeakers suitable for the handheld playing device. The method has the advantages that the dimensional sound image distribution of the handheld playing device can be extended, and the replaying effect is improved. Meanwhile, the signal processing has lower requirements on the computing and storage quantity, and the dimensional sound extension method is suitable for the signal processing hardware configuration of the handheld playing device.

Description

A kind of dimensional sound extension method for hand-hold type playing device

Technical field

The present invention relates to sterophonic technique field, be specifically related to a kind of dimensional sound extension method for hand-hold type playing device.

Background technology

Common binary channel stereophonic signal needs to reset with a pair boombox.The boombox of standard is arranged in the left front and right front position of attentive listener, and the subtended angle of relative attentive listener is 60 °.But in actual applications, also often adopt various hand-hold type playing device, as mobile phone, MP3/MP4 player, iPhone, ipad etc., play stereophonic signal.By the restriction of its structure, the distance between a pair miniature boombox of these hand-hold type playing devices is rather narrow, and usually only has several centimetres to tens cm, and thus the subtended angle of attentive listener is also very little relatively, does not far reach the standard of 60 °.When using these hand-hold type playing device playback stereophonic signals, the virtual sound source of perception will be made to concentrate on very narrow region and lose normal stereophonic effect.

Stereophonic widening technology can be used for making up the problem that narrow a pair subtended angle speaker playback stereophonic signal produces.This carries out precorrection process by the method stereophonic signal of signal transacting, makes narrow a pair subtended angle speaker playback go out the stereophonic audition effect of expansion.This kind of technology is based on head-position difficult labor and aural transmission principles of signal processing, relevant patented technology comprises the virtual speaker technology in Dolby laboratory, the technology of SRS and the patented technology (signal processing method of two-loudspeaker virtual 5.1 passageway circumvents sound of South China Science & Engineering University, national inventing patent mandate, ZL02134415.9) etc.These technology have been used to that television stereo is low voice speaking puts.

But these stereophonic widening technology for television set are also not suitable for being directly used in hand-hold type playing device.Its reason is:

(1) the stereophonic widening technology for television set designs by far-field audio speaker layout (loud speaker is greater than 1.0 meters to the distance at attentive listener head center), and what signal transacting was used is also far field head-position difficult labor.And in practical application, the loud speaker of hand-hold type playing device and the distance at attentive listener head center only have 20 to 50 centimetres (0.2 to 0.5 meter) left and right usually, belong to the distance of near field, need to use Head-Related Transfer Function for Nearby Sources process.

(2) under normal circumstances, the subtended angle of the relative attentive listener of a pair boombox of television set is greatly between 15 ° to 30 °, and its stereophonic widening process is also by this parameter designing.And the distance between a pair of hand-hold type playing device miniature boombox is rather narrow, its maximum relative to the subtended angle of attentive listener namely 10 °.When narrow subtended angle like this, extension process needs to do larger lifting (as existing " stereo dipole " technology) to the low frequency part of signal, and signal transacting is more difficult.

(3) in order to reach certain replaying effect, the stereophonic widening signal transacting relative complex of television set, has certain requirement to the calculating of signal transacting and memory space.On the contrary, in hand-hold type playing device, because the characteristic of its Microspeaker itself limits, can suitably reduce the requirement of replaying effect.But require that signal transacting hardware is fairly simple, thus will simplify as far as possible calculating and memory space.

(4) existing stereophonic widening technology supposition playback loudspeaker is placed in the horizontal plane of ears.And in the listening attentively to of reality, hand-hold type playing device is not necessarily placed in the plane of ears level.Existing stereophonic widening technology can not adapt to this change.

Although also have the stereophonic widening technology for hand-hold type playing device abroad, and be similar for the stereophonic widening technology of television set, the problem do not solved well above.

Summary of the invention

The present invention solves the defect existed in above-mentioned prior art, provides the dimensional sound extension method being applicable to hand-hold type playing device.The method can under the condition of the narrow boombox subtended angle of hand-hold type playing device, near field loudspeaker distance, the stereophonic effect of expansion of resetting out.Meanwhile, signal transacting of the present invention requires lower to calculating and memory space, is suitable for the signal transacting hardware configuration of hand-hold type playing device.

The present invention realizes in the following way:

For a dimensional sound extension method for hand-hold type playing device, comprise the steps:

Step S1, reads in original binary channel stereophonic signal, comprises left path signal L and right path signal R;

Step S2, carries out being added and subtracting each other by left path signal L and right path signal R respectively, obtains and signal M=L+R and difference signal S=L-R;

Step S3, will use a pair aural transmission filter ∑ (f) and Δ (f) filtering respectively with signal M and difference signal S, obtain filtered and signal ∑ (f) M and filtered difference signal Δ (f) S;

Step S4, be added respectively difference signal Δ (f) S after filtered and signal ∑ (f) M and filtering and subtract each other, obtaining new left path signal L` and new right path signal R`, hand-hold type playing device of feeding is reset.

Further, aural transmission filter ∑ (f) described in step S3 and Δ (f), be obtained by 2.5 ° to 5 ° Sounnd source directions of rigid head portion model, the Head-Related Transfer Function for Nearby Sources of 0.2-0.5 rice distance and 30 ° of Sounnd source direction far field head-position difficult labor, described rigid head portion model can be calculated by following formula at the head-position difficult labor of any sound source distance:

$H(r,\mathrm{\Γ},f)=-\frac{\mathrm{rexp}\left(\mathrm{jkr}\right)}{k{a}^2}\underset{l=0}{\overset{\∞}{\mathrm{\Σ}}}\frac{(2l+1)h_l\left(\mathrm{kr}\right)P_l\left(\cos \mathrm{\Γ}\right)}{{\mathrm{dh}}_l\left(\mathrm{ka}\right)/d\left(\mathrm{ka}\right)},$

Wherein r is the distance at sound source center to the end; Γ is the angle between Sounnd source direction and ear locality; F is frequency; K=2 π f/c is wave number, and c=343 meter per second is the velocity of sound; A=0.0875 rice is average head radius; J is imaginary unit; h _l(ka) be Equations of The Second Kind ball Hankel function; P _l(cos Γ) is Legnedre polynomial,

In far field, namely sound source distance r is much larger than the far field situation of head radius a, and described head-position difficult labor will have nothing to do with distance r, and can be calculated by following formula:

$H(\mathrm{\Γ},f)=-\frac{1}{{\left(\mathrm{ka}\right)}^2}\underset{l=0}{\overset{\∞}{\mathrm{\Σ}}}\frac{(2l+1)j^{l+1}{P}_l\left(\cos \mathrm{\Γ}\right)}{{\mathrm{dh}}_l\left(\mathrm{ka}\right)/d\left(\mathrm{ka}\right)},$

Its concrete steps comprise:

Step S201, obtains the low voice speaking acoustic pressure being placed on ears place and producing during standard stereo loudspeaker arrangement, namely when attentive listener is 2 θ at a pair subtended angle ₀during the position of the left and right symmetry in the standard stereo loudspeaker arrangement of=60 °, loud speaker meets to the head-position difficult labor of picking up the ears loud speaker is to the head-position difficult labor of heteropleural ear if feed, the original stereo signal of loud speaker is L and R, and the so low voice speaking acoustic pressure being placed on the generation of ears place is:

$\left[\begin{array}{c}{P}_L\\ P_R\end{array}\right]=\left[\begin{array}{cc}{H}_{\mathrm{LL}}^1& H_{\mathrm{LR}}^1\\ H_{\mathrm{RL}}^1& H_{\mathrm{RR}}^1\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right]=\left[\begin{array}{cc}{\mathrm{\α}}_1& {\mathrm{\β}}_1\\ {\mathrm{\β}}_1& {\mathrm{\α}}_1\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right];$

Step S202, obtains the low voice speaking acoustic pressure being placed on ears place and producing in hand-hold type playing device, is in hand-hold type playing device, a pair closely, narrow subtended angle loud speaker is H to the head-position difficult labor of picking up the ears together _lL=H _rR=α, the head-position difficult labor to heteropleural ear is H _rL=H _lR=β, the signal of loud speaker of feeding is L` and R`, and the so low voice speaking acoustic pressure being placed on the generation of ears place is:

$\left[\begin{array}{c}{P_L}^{\′}\\ {P_R}^{\′}\end{array}\right]=\left[\begin{array}{cc}{H}_{\mathrm{LL}}& H_{\mathrm{LR}}\\ H_{\mathrm{RL}}& H_{\mathrm{RR}}\end{array}\right]\left[\begin{array}{c}{L}^{\′}\\ R^{\′}\end{array}\right]=\left[\begin{array}{cc}\mathrm{\α}& \mathrm{\β}\\ \mathrm{\β}& \mathrm{\α}\end{array}\right]\left[\begin{array}{c}{L}^{\′}\\ R^{\′}\end{array}\right];$

Step S203, makes and low voice speakingly in hand-hold type playing device is placed on acoustic pressure that ears place produces and low voice speaking during standard stereo loudspeaker arrangement to be placed on the acoustic pressure that ears place produces equal, obtain:

$\left[\begin{array}{c}{L}^{\′}\\ R^{\′}\end{array}\right]={\left[\begin{array}{cc}\mathrm{\α}& \mathrm{\β}\\ \mathrm{\β}& \mathrm{\α}\end{array}\right]}^{-1}\left[\begin{array}{cc}{\mathrm{\α}}_1& {\mathrm{\β}}_1\\ {\mathrm{\β}}_1& {\mathrm{\α}}_1\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right]=\left[\begin{array}{cc}{G}_1\left(f\right)& G_2\left(f\right)\\ G_2\left(f\right)& G_1\left(f\right)\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right],$

Wherein,

$\begin{array}{cc}{G}_1\left(f\right)=\frac{{\mathrm{\α\α}}_1-{\mathrm{\β\β}}_1}{{\mathrm{\α}}^2-{\mathrm{\β}}^2}& G_2\left(f\right)=\frac{\mathrm{\α}{\mathrm{\β}}_1-\mathrm{\β}{\mathrm{\α}}_1}{{\mathrm{\α}}^2-{\mathrm{\β}}^2}\end{array};$

Step S204, is introduced as the tone color equilibrium treatment improving playback tone color again in signal transacting, namely in signal L` and R`, be multiplied by a public factor relevant with frequency:

$\mathrm{EQ}=\frac{|{\mathrm{\α}}^2-{\mathrm{\β}}^2|}{\sqrt{{|{\mathrm{\α\α}}_1-{\mathrm{\β\β}}_1|}^2+{|{\mathrm{\α\β}}_1-\mathrm{\β}{\mathrm{\α}}_1|}^2}},$

EQ is substituted into G ₁(f) and G ₂f (), obtains:

${G^{\′}}_1\left(f\right)=\frac{{\mathrm{\α\α}}_1-{\mathrm{\β\β}}_1}{\sqrt{{|{\mathrm{\α\α}}_1-{\mathrm{\β\β}}_1|}^2+{|\mathrm{\α}{\mathrm{\β}}_1-{\mathrm{\β\α}}_1|}^2}}\frac{|{\mathrm{\α}}^2-{\mathrm{\β}}^2|}{{\mathrm{\α}}^2-{\mathrm{\β}}^2}$

${G^{\′}}_2\left(f\right)=\frac{{\mathrm{\α\β}}_1-{\mathrm{\β\α}}_1}{\sqrt{{|{\mathrm{\α\α}}_1-{\mathrm{\β\β}}_1|}^2+{|\mathrm{\α}{\mathrm{\β}}_1-{\mathrm{\β\α}}_1|}^2}}\frac{|{\mathrm{\α}}^2-{\mathrm{\β}}^2|}{{\mathrm{\α}}^2-{\mathrm{\β}}^2};$

Step S205, draws Equivalent Form according to formula described in step S203 and step S204:

$\left[\begin{array}{c}{L}^{\′}\\ R^{\′}\end{array}\right]=\left[\begin{array}{cc}1& 1\\ 1& -1\end{array}\right]\left[\begin{array}{cc}{\mathrm{\Σ}}^{\′}\left(f\right)& 0\\ 0& {\mathrm{\Δ}}^{\′}\left(f\right)\end{array}\right]\left[\begin{array}{cc}1& 0\\ 0& 1\end{array}\right],$

Namely show that ∑ ` (f) and Δ ` (f) are:

$\begin{array}{c}{\mathrm{\Σ}}^{`}\left(f\right)=0.5[{G^{`}}_1\left(f\right)+{G^{\′}}_2\left(f\right)]\\ =0.5\frac{{\mathrm{\α\α}}_1-{\mathrm{\β\β}}_1+\mathrm{\α}{\mathrm{\β}}_1-{\mathrm{\β\α}}_1}{\sqrt{{|{\mathrm{\α\α}}_1-{\mathrm{\β\β}}_1|}^2+{|\mathrm{\α}{\mathrm{\β}}_1-{\mathrm{\β\α}}_1|}^2}}\frac{|{\mathrm{\α}}^2-{\mathrm{\β}}^2|}{{\mathrm{\α}}^2-{\mathrm{\β}}^2}\\ {\mathrm{\Δ}}^{\′}\left(f\right)=0.5[{G^{\′}}_1\left(f\right)-{G^{\′}}_2\left(f\right)]\\ =0.5\frac{{\mathrm{\α\α}}_1-{\mathrm{\β\β}}_1-{\mathrm{\α\β}}_1+{\mathrm{\β\α}}_1}{\sqrt{{|{\mathrm{\α\α}}_1-{\mathrm{\β\β}}_1|}^2+{|\mathrm{\α}{\mathrm{\β}}_1-\mathrm{\β}{\mathrm{\α}}_1|}^2}}\frac{|{\mathrm{\α}}^2-{\mathrm{\β}}^2|}{{\mathrm{\α}}^2-{\mathrm{\β}}^2}\end{array};$

Step S206, carries out the high-pass filtering of more than 200Hz to ∑ ` (f) and Δ ` (f), in order to the low-frequency component of filtering at below 200Hz, namely obtain aural transmission filter ∑ (f) and Δ (f).These features are differences of the present invention and prior art.

Principle of the present invention is: when narrow subtended angle boombox is arranged, the process of employing aural transmission filter stereophonic signal can by outside stereophonic sound image distributed expansion to loudspeaker arrangement subtended angle.Aural transmission filter adopts head-position difficult labor design.In order to adapt between hand-hold type playing device and attentive listener closely, Head-Related Transfer Function for Nearby Sources is adopted to carry out signal transacting.And in order to the narrow subtended angle between a pair that adapts to hand-hold type playing device miniature boombox, signal transacting is the loud speaker subtended angle design by 10 ° (or less than 10 °, as 5 °).Namely, loudspeaker arrangement, in the position of ± 5 ° (or narrower, as ± 2.5 °), adopts the head-position difficult labor process of 5 ° or narrower (± 2.5 °) Sounnd source directions.Consider that the low-frequency minimum of the Microspeaker of hand-hold type playing device itself is preferably just between 200Hz to 300Hz simultaneously, the low frequency part of signal to be processed and meaningless.Thus add high-pass filtering in the signal processing, filtering at 200Hz (or 300Hz) low-frequency component below, thus avoids low frequency signal lifting larger required for narrow subtended angle loudspeaker arrangement, decreases signal transacting difficulty.

Compared with prior art, tool has the following advantages and beneficial effect in the present invention:

1., after the present invention processes common binary channel stereophonic signal, reset by narrow a pair subtended angle loudspeaker arrangement of hand-hold type playing device, the stereophonic sound image effect of expansion can be produced;

2. the narrow subtended angle loudspeaker arrangement that the present invention is adjustable between pressing 5 ° to 10 °, and carry out signal transacting with near-field HRTF, adapt to actual loudspeaker subtended angle and the audition distance of hand-hold type playing device.Add high-pass filtering in the signal processing simultaneously, avoid the low-frequency signal processing difficulty that narrow subtended angle loudspeaker arrangement is brought.

3. the HRTF data that signal transacting of the present invention is used are calculated by rigidity spherical head model, has symmetry, adapts to the placement of the hand-hold type playing device of the plane of varying level.

4. the present invention can adopt general signal processing chip (DSP hardware) circuit or special integrated circuit to realize, and the software that algorithmic language (as VC++) also can be adopted to work out realizes on multimedia computer;

5. the present invention can be used as algorithm write hardware chip, for various hand-hold type playing device as the sound reproduction such as MP3, MP4, iPhone and mobile phone, also can be used as software to being converted to the special audio files of hand-hold type playback equipment after normal stereo Signal Pretreatment, play-over in common hand-hold type playback equipment, without the need to increasing any hardware and process.

Accompanying drawing explanation

Fig. 1 is the schematic diagram of hand-hold type playing device and loudspeaker arrangement thereof.

Fig. 2 is the schematic diagram of the stereophonic widening signal processing method for hand-hold type playing device.

Fig. 3 is to the acoustic transmission path profile of ears for the loud speaker of hand-hold type playing device.

Fig. 4 be during standard stereo is reset loud speaker to the acoustic transmission path profile of ears.

Fig. 5 is the frequency response curve of the dimensional sound extension method of hand-hold type playing device a pair aural transmission filter used.

Fig. 6 is the pretreated flow chart of dimensional sound extension method with the hand-hold type playing device of signal processing software realization in multimedia computer.

Embodiment

Below in conjunction with accompanying drawing, the present invention is further described.

Fig. 1 is the schematic diagram of hand-hold type playing device and loudspeaker arrangement thereof.The distance of its a pair loud speaker is very near, usually only has several centimetres to tens cm.In practical application, the loud speaker of hand-hold type playing device and the distance at attentive listener head center only have 20 to 50 centimetres (0.2 to 0.5 meter) left and right usually.Thus loud speaker relative to the subtended angle of attentive listener at 10 ° or less.

The schematic diagram of Fig. 2 dimensional sound extension method that to be of the present invention be for hand-hold type playing device.It uses again after common binary channel stereophonic signal is processed a pair of hand-hold type playing device closely, narrow subtended angle speaker playback, the stereophonic effect be expanded.

The present invention is obtained by following consideration and method.Be 2 θ at a pair subtended angle of Fig. 3 ₀in the standard stereo loudspeaker arrangement of=60 °, when attentive listener is when the position of left and right symmetry, two-loudspeaker is symmetrical to the acoustic transmission function (head-position difficult labor, HRTF) of ears, and namely loud speaker meets to the head-position difficult labor of picking up the ears loud speaker is to the head-position difficult labor of heteropleural ear the original stereo signal of loud speaker of supposing to feed is L and R, is so low voice speakingly placed on the acoustic pressure that ears place produces and is:

$\left[\begin{array}{c}{P}_L\\ P_R\end{array}\right]=\left[\begin{array}{cc}{H}_{\mathrm{LL}}^1& H_{\mathrm{LR}}^1\\ H_{\mathrm{RL}}^1& H_{\mathrm{RR}}^1\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right]=\left[\begin{array}{cc}{\mathrm{\α}}_1& {\mathrm{\β}}_1\\ {\mathrm{\β}}_1& {\mathrm{\α}}_1\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right]---\left(1\right)$

In hand-hold type playing device, a pair closely, narrow subtended angle loud speaker is to being H with the head-position difficult labor of picking up the ears _lL=H _rR=α, the head-position difficult labor to heteropleural ear is H _rL=H _lR=β, the signal of loud speaker of feeding is L` and R`, and the so low voice speaking acoustic pressure being placed on the generation of ears place is:

$\left[\begin{array}{c}{P_L}^{\′}\\ {P_R}^{\′}\end{array}\right]=\left[\begin{array}{cc}{H}_{\mathrm{LL}}& H_{\mathrm{LR}}\\ H_{\mathrm{RL}}& H_{\mathrm{RR}}\end{array}\right]\left[\begin{array}{c}{L}^{\′}\\ R^{\′}\end{array}\right]=\left[\begin{array}{cc}\mathrm{\α}& \mathrm{\β}\\ \mathrm{\β}& \mathrm{\α}\end{array}\right]\left[\begin{array}{c}{L}^{\′}\\ R^{\′}\end{array}\right]---\left(2\right)$

If stereophonic signal L and R being suitable for standard loudspeakers layout is become L` and R` through suitably processing, they are fed a pair closely, narrow subtended angle speaker playback, make playback acoustic pressure identical with the situation of standard arrangement, then can obtain identical stereophonic effect, closely, narrow subtended angle loudspeaker arrangement corrected.Make the ears acoustic pressure of upper two formulas equal, can obtain:

$\left[\begin{array}{c}{L}^{\′}\\ R^{\′}\end{array}\right]={\left[\begin{array}{cc}\mathrm{\α}& \mathrm{\β}\\ \mathrm{\β}& \mathrm{\α}\end{array}\right]}^{-1}\left[\begin{array}{cc}{\mathrm{\α}}_1& {\mathrm{\β}}_1\\ {\mathrm{\β}}_1& {\mathrm{\α}}_1\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right]=\left[\begin{array}{cc}{G}_1\left(f\right)& G_2\left(f\right)\\ G_2\left(f\right)& G_1\left(f\right)\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right]---\left(3\right)$

Wherein

$\begin{array}{cc}{G}_1\left(f\right)=\frac{{\mathrm{\α\α}}_1-{\mathrm{\β\β}}_1}{{\mathrm{\α}}^2-{\mathrm{\β}}^2}& G_2\left(f\right)=\frac{\mathrm{\α}{\mathrm{\β}}_1-\mathrm{\β}{\mathrm{\α}}_1}{{\mathrm{\α}}^2-{\mathrm{\β}}^2}\end{array}---\left(4\right)$

In order to improve the tone color of playback, tone color equilibrium treatment can be introduced again in the signal processing.Namely in signal L` and R` of above formula, be multiplied by a public factor relevant with frequency:

$\mathrm{EQ}=\frac{|{\mathrm{\α}}^2-{\mathrm{\β}}^2|}{\sqrt{{|{\mathrm{\α\α}}_1-{\mathrm{\β\β}}_1|}^2+{|{\mathrm{\α\β}}_1-\mathrm{\β}{\mathrm{\α}}_1|}^2}}---\left(5\right)$

Relative amplitude then between two path signal and phase relation can not change, and thus can not change the Sound image localization of playback, but change the power spectrum of signal.Like this, the G of (4) formula ₁(f) and G ₂f () is by the G` with following formula <sub>1</sub>(f) and G` ₂f () replaces:

${G^{\&prime;}}_1\left(f\right)=\frac{{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1}{\sqrt{{|{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1|}^2+{|\mathrm{\&alpha;}{\mathrm{\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1|}^2}}\frac{|{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2|}{{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2}$

${G^{\&prime;}}_2\left(f\right)=\frac{{\mathrm{\&alpha;\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1}{\sqrt{{|{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1|}^2+{|\mathrm{\&alpha;}{\mathrm{\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1|}^2}}\frac{|{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2|}{{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2}---\left(6\right)$

At this time, treated left and right path signal keeps the gross power of former left and right path signal spectrum constant, thus decreases the change of playback tone color.Some zero pole point of all right erasure signal process filter of this equilibrium treatment, make signal transacting become simple.

(3) formula and (6) formula and below (7) formula equivalence:

$\left[\begin{array}{c}{L}^{\&prime;}\\ R^{\&prime;}\end{array}\right]=\left[\begin{array}{cc}1& 1\\ 1& -1\end{array}\right]\left[\begin{array}{cc}{\mathrm{\&Sigma;}}^{\&prime;}\left(f\right)& 0\\ 0& {\mathrm{\&Delta;}}^{\&prime;}\left(f\right)\end{array}\right]\left[\begin{array}{cc}1& 0\\ 0& 1\end{array}\right]---\left(7\right)$

Wherein:

$\begin{array}{c}{\mathrm{\&Sigma;}}^{`}\left(f\right)=0.5[{G^{`}}_1\left(f\right)+{G^{\&prime;}}_2\left(f\right)]\\ =0.5\frac{{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1+\mathrm{\&alpha;}{\mathrm{\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1}{\sqrt{{|{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1|}^2+{|\mathrm{\&alpha;}{\mathrm{\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1|}^2}}\frac{|{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2|}{{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2}\\ {\mathrm{\&Delta;}}^{\&prime;}\left(f\right)=0.5[{G^{\&prime;}}_1\left(f\right)-{G^{\&prime;}}_2\left(f\right)]\\ =0.5\frac{{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1-{\mathrm{\&alpha;\&beta;}}_1+{\mathrm{\&beta;\&alpha;}}_1}{\sqrt{{|{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1|}^2+{|\mathrm{\&alpha;}{\mathrm{\&beta;}}_1-\mathrm{\&beta;}{\mathrm{\&alpha;}}_1|}^2}}\frac{|{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2|}{{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2}\end{array}---\left(8\right)$

∑ ` (f) in (8) formula and Δ ` (f) are carried out to the high-pass filtering of more than 200Hz, in order to the low-frequency component of filtering at below 200Hz, namely obtain aural transmission filter ∑ (f) and Δ (f).

The principles of signal processing figure of Fig. 2 can be designed according to (7) formula and (8) formula.Aural transmission filter ∑ (f) and Δ (f) are according to the four head-related-transfer function alpha of loud speaker to ears, β, α ₁and β ₁calculate.The head-position difficult labor that the present invention adopts rigidity spherical head model to calculate.This is that stereophonic sound image is limited in attentive listener front, also need not can not produce whole three dimensions sound virtual sound source with the head-position difficult labor comprising auricle because in the stereophonic widening of hand-hold type playing device on the one hand.The head-position difficult labor of rigid head portion model makes aural transmission filter become simple on the contrary.On the other hand, the head-position difficult labor of rigidity spherical head model has Sphere symmetry, just in time adapts to the placement of the hand-hold type playing device of the plane of varying level.

Rigid head portion model can be calculated by following formula at the head-position difficult labor of any sound source distance:

$H(r,\mathrm{\&Gamma;},f)=-\frac{\mathrm{rexp}\left(\mathrm{jkr}\right)}{k{a}^2}\underset{l=0}{\overset{\&infin;}{\mathrm{\&Sigma;}}}\frac{(2l+1)h_l\left(\mathrm{kr}\right)P_l\left(\cos \mathrm{\&Gamma;}\right)}{{\mathrm{dh}}_l\left(\mathrm{ka}\right)/d\left(\mathrm{ka}\right)}---\left(9\right)$

In above formula, r is the distance at sound source (loud speaker) center to the end; Γ is the angle between Sounnd source direction and ear locality; F is frequency; K=2 π f/c is wave number, and c=343 meter per second is the velocity of sound; A=0.0875 rice is average head radius; J is imaginary unit; h _l(ka) be Equations of The Second Kind ball Hankel function; P _l(cos Γ) is Legnedre polynomial.

In far field, namely sound source distance r is much larger than the far field situation of head radius a, and the head-position difficult labor of (9) formula will have nothing to do with distance r, and can be calculated by following formula:

$H(\mathrm{\&Gamma;},f)=-\frac{1}{{\left(\mathrm{ka}\right)}^2}\underset{l=0}{\overset{\&infin;}{\mathrm{\&Sigma;}}}\frac{(2l+1)j^{l+1}{P}_l\left(\cos \mathrm{\&Gamma;}\right)}{{\mathrm{dh}}_l\left(\mathrm{ka}\right)/d\left(\mathrm{ka}\right)}---\left(10\right)$

To hand-hold type playing device, closely, the head-position difficult labor of narrow subtended angle loud speaker needs to calculate by (9) formula.As an example, consider the distance at loud speaker and attentive listener head center in usual hand-hold type playing device, r=0.3 rice can be chosen.In FIG, if the distance d=0.1 rice between a pair loud speaker of hand-hold type playing device, then the relative attentive listener of two-loudspeaker subtended angle be 2 θ ₁=10 °.Then loud speaker and be Γ=90 °-θ with the angle between picking up the ears ₁=85 °, and the angle between heteropleural ear is Γ=90 °+θ ₁=95 °.(9) formula that these parameters substituted into can obtain loud speaker to the same head-position difficult labor α picked up the ears ₁with the head-position difficult labor β of heteropleural ear ₁, namely obtain the Head-Related Transfer Function for Nearby Sources of Sounnd source direction 5 ° (relative dead ahead), 0.3 meter of distance.In like manner, the head-position difficult labor of standard stereo loudspeaker arrangement can calculate by (10) formula.Assuming that the relative attentive listener of two-loudspeaker angle be 2 θ ₀=60 °, then loud speaker and be Γ=90 °-θ with the angle between picking up the ears ₀=60 °, and the angle between heteropleural ear is Γ=90 °+θ ₀=120 °.These parameters are substituted into (10) formula and can obtain loud speaker to the head-position difficult labor β with the head-position difficult labor α picked up the ears and heteropleural ear, namely obtain the far field head-position difficult labor in Sounnd source direction 30 ° (relative dead ahead).

Solid line in Fig. 5 is the amplitude response by above-mentioned head-position difficult labor data and a pair aural transmission filter ∑ ` (f) calculated and Δ ` (f).The amplitude of filter response ∑ ` (f) has larger lifting at the low frequency being less than 200Hz.This is that subtended angle in hand-hold type playing device between a pair loud speaker only has caused by 10 °, also brings difficulty to signal transacting simultaneously.But consider that the low-frequency minimum of the Microspeaker of hand-hold type playing device itself is preferably just between 200Hz to 300Hz, the low frequency part of signal to be processed and meaningless.Thus add high-pass filtering in the signal processing, filtering at the low-frequency component of below 200Hz, thus avoids low frequency signal lifting larger required for narrow subtended angle loudspeaker arrangement, decreases signal transacting difficulty.Dotted line in Fig. 5 is the amplitude response of a pair aural transmission filter ∑ (f) through high-pass filtering and Δ (f).Final filter can realize by a pair 64 or 128 FIR (finite impulse response (FIR)) structures, thus requires lower to signal transacting to calculating and memory space.

The hardware circuit that the present invention can utilize the general signal processing chip in hand-hold type playing device to make realizes; Also can be designed to special integrated circuit (IC) chip realize; Software can also be designed on multimedia computer, realize preliminary treatment, the audio files handled well (as mp3 file) is reset in common hand-hold type playing device.The present invention can be used for the sound reproduction of the hand-hold type such as mobile phone, MP3, MP4, iphone playing device.

The application of mobile phone or MP3 player

Obtain two paths of signals after common binary channel stereophonic signal is carried out stereophonic widening process by (7) formula or Fig. 2, the miniature loudspeaker of player of then feeding respectively is reset.Wherein stereophonic widening process is as a part of hardware circuit in player.

The pretreated application of multimedia computer

The audio files (as WAV or mp3 file) of ordinary stereo acoustical signal is read in computer, then after carrying out stereophonic widening preliminary treatment with computer software by the method for (7) formula or Fig. 2, the two paths of signals obtained, and save as new audio files.Directly will reset at various hand-hold type playing devices such as mobile phones through pretreated audio files.

Fig. 6 is the pretreated flow chart of dimensional sound extension method with the hand-hold type playing device of signal processing software realization in multimedia computer.

The first step, the audio files (as WAV or mp3 file) of ordinary stereo acoustical signal is read in computer, obtain left path signal L and right path signal R.

Second step, left path signal L and right path signal R is carried out being added and subtracting each other respectively, obtain and signal M=L+R and difference signal S=L-R.

3rd step, a pair aural transmission filter ∑ (f) and Δ (f) filtering will be used respectively with signal M and difference signal S, obtain filtered and signal ∑ (f) M and filtered difference signal Δ (f) S.

4th step, difference signal Δ (f) S after filtered and signal ∑ (f) M and filtering be added respectively and subtracted each other, obtaining through pretreated left path signal L` and right path signal R`.

5th step, will be combined into through pretreated left path signal L` and right path signal R` and directly reset at the various hand-hold type playing device such as mobile phone through pretreated audio files.

As mentioned above, the present invention can be realized preferably.

Listening testing experimental verification actual effect of the present invention.Subjective experiment mainly plays stereophonic signal on mobile phone, allows attentive listener qualitatively judge stereo widening effect.All 8 experimenters confirm stereo widening effect.

Research of the present invention obtains the subsidy of state natural sciences fund " numbering: 11174087 " and subtropical zone architecture science National Key Laboratory Funded Projects.

The above embodiment of the present invention is only for example of the present invention is clearly described, and is not the restriction to embodiments of the present invention.For those of ordinary skill in the field, can also make other changes in different forms on the basis of the above description.Here exhaustive without the need to also giving all execution modes.All any amendments done within the spirit and principles in the present invention, equivalent to replace and improvement etc., within the protection range that all should be included in the claims in the present invention.

Claims (1)

1. for a dimensional sound extension method for hand-hold type playing device, it is characterized in that, comprise the steps:

Step S1, reads in original binary channel stereophonic signal, comprises left path signal L and right path signal R;

Step S2, carries out being added and subtracting each other by left path signal L and right path signal R respectively, obtains and signal M=L+R and difference signal S=L-R;

Step S3, will use a pair aural transmission filter ∑ (f) and Δ (f) filtering respectively with signal M and difference signal S, obtain filtered and signal ∑ (f) M and filtered difference signal Δ (f) S;

Step S4, be added respectively difference signal Δ (f) S after filtered and signal ∑ (f) M and filtering and subtract each other, obtaining new left path signal L` and new right path signal R`, hand-hold type playing device of feeding is reset;

Aural transmission filter Σ (f) described in step S3 and Δ (f), be obtained by 2.5 ° to 5 ° Sounnd source directions of rigid head portion model, the Head-Related Transfer Function for Nearby Sources of 0.2-0.5 rice distance and 30 ° of Sounnd source direction far field head-position difficult labor, described rigid head portion model can be calculated by following formula at the head-position difficult labor of any sound source distance:

$H(r,\mathrm{\&Gamma;},f)=-\frac{\mathrm{r\; exp}\left(\mathrm{jkr}\right)}{k{a}^2}\underset{l=0}{\overset{\&infin;}{\mathrm{\&Sigma;}}}\frac{(2l+1)h_l\left(\mathrm{kr}\right)P_l\left(\cos \mathrm{\&Gamma;}\right)}{{\mathrm{dh}}_l\left(\mathrm{ka}\right)/d\left(\mathrm{ka}\right)},$

Wherein r is the distance at sound source center to the end; Γ is the angle between Sounnd source direction and ear locality; F is frequency; K=2 π f/c is wave number, and c=343 meter per second is the velocity of sound; A=0.0875 rice is average head radius; J is imaginary unit; h _l(ka) be Equations of The Second Kind ball Hankel function; P _l(cos Γ) is Legnedre polynomial,

In far field, namely sound source distance r is much larger than the far field situation of head radius a, and described head-position difficult labor will have nothing to do with distance r, and can be calculated by following formula:

$H(\mathrm{\&Gamma;},f)=-\frac{1}{{\left(\mathrm{ka}\right)}^2}\underset{l=0}{\overset{\&infin;}{\mathrm{\&Sigma;}}}\frac{(2l+1)j^{l+1}{P}_l\left(\cos \mathrm{\&Gamma;}\right)}{{\mathrm{dh}}_l\left(\mathrm{ka}\right)/d\left(\mathrm{ka}\right)},$

Its concrete steps comprise:

Step S201, obtains the low voice speaking acoustic pressure being placed on ears place and producing during standard stereo loudspeaker arrangement, namely when attentive listener is 2 θ at a pair subtended angle ₀during the position of the left and right symmetry in the standard stereo loudspeaker arrangement of=60 °, loud speaker meets to the head-position difficult labor of picking up the ears loud speaker is to the head-position difficult labor of heteropleural ear if feed, the original stereo signal of loud speaker is L and R, and the so low voice speaking acoustic pressure being placed on the generation of ears place is:

$\left[\begin{array}{c}{P}_L\\ P_R\end{array}\right]=\left[\begin{array}{cc}{H}_{\mathrm{LL}}^1& H_{\mathrm{LR}}^1\\ H_{\mathrm{RL}}^1& H_{\mathrm{RR}}^1\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right]=\left[\begin{array}{cc}{\mathrm{\&alpha;}}_1& {\mathrm{\&beta;}}_1\\ {\mathrm{\&beta;}}_1& {\mathrm{\&alpha;}}_1\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right];$

Step S202, obtains the low voice speaking acoustic pressure being placed on ears place and producing in hand-hold type playing device, is in hand-hold type playing device, a pair closely, narrow subtended angle loud speaker is H to the head-position difficult labor of picking up the ears together _lL=H _rR=α, the head-position difficult labor to heteropleural ear is H _rL=H _lR=β, the signal of loud speaker of feeding is L` and R`, and the so low voice speaking acoustic pressure being placed on the generation of ears place is:

$\left[\begin{array}{c}{P_L}^{\&prime;}\\ {P_R}^{\&prime;}\end{array}\right]=\left[\begin{array}{cc}{H}_{\mathrm{LL}}& H_{\mathrm{LR}}\\ H_{\mathrm{RL}}& H_{\mathrm{RR}}\end{array}\right]\left[\begin{array}{c}{L}^{\&prime;}\\ R^{\&prime;}\end{array}\right]=\left[\begin{array}{cc}\mathrm{\&alpha;}& \mathrm{\&beta;}\\ \mathrm{\&beta;}& \mathrm{\&alpha;}\end{array}\right]\left[\begin{array}{c}{L}^{\&prime;}\\ R^{\&prime;}\end{array}\right];$

Step S203, makes and low voice speakingly in hand-hold type playing device is placed on acoustic pressure that ears place produces and low voice speaking during standard stereo loudspeaker arrangement to be placed on the acoustic pressure that ears place produces equal, obtain:

$\left[\begin{array}{c}{L}^{\&prime;}\\ R^{\&prime;}\end{array}\right]={\left[\begin{array}{cc}\mathrm{\&alpha;}& \mathrm{\&beta;}\\ \mathrm{\&beta;}& \mathrm{\&alpha;}\end{array}\right]}^{-1}\left[\begin{array}{cc}{\mathrm{\&alpha;}}_1& {\mathrm{\&beta;}}_1\\ {\mathrm{\&beta;}}_1& {\mathrm{\&alpha;}}_1\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right]\left[\begin{array}{cc}{G}_1\left(f\right)& G_2\left(f\right)\\ G_2\left(f\right)& G_1\left(f\right)\end{array}\right]\left[\begin{array}{c}L\\ R\end{array}\right],$

Wherein,

$\begin{array}{cc}{G}_1\left(f\right)=\frac{{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1}{{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2}& G_2\left(f\right)=\frac{{\mathrm{\&alpha;\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1}{{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2}\end{array};$

Step S204, is introduced as the tone color equilibrium treatment improving playback tone color again in signal transacting, namely in signal L` and R`, be multiplied by a public factor relevant with frequency:

$\mathrm{EQ}=\frac{|{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2|}{\sqrt{{|{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1|}^2+{|{\mathrm{\&alpha;\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1|}^2}},$

EQ is substituted into G ₁(f) and G ₂f (), obtains:

${G^{\&prime;}}_1\left(f\right)=\frac{{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1}{\sqrt{{|{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1|}^2+{|{\mathrm{\&alpha;\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1|}^2}}\frac{|{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2|}{{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2}$

${G^{\&prime;}}_2\left(f\right)=\frac{{\mathrm{\&alpha;\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1}{\sqrt{{|{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1|}^2+{|{\mathrm{\&alpha;\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1|}^2}}\frac{|{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2|}{{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2}$

Step S205, draws Equivalent Form according to formula described in step S203 and step S204:

$\left[\begin{array}{c}{L}^{\&prime;}\\ R^{\&prime;}\end{array}\right]=\left[\begin{array}{cc}1& 1\\ 1& -1\end{array}\right]\left[\begin{array}{cc}{\mathrm{\&Sigma;}}^{\&prime;}\left(f\right)& 0\\ 0& {\mathrm{\&Delta;}}^{\&prime;}\left(f\right)\end{array}\right]\left[\begin{array}{cc}1& 0\\ 0& 1\end{array}\right],$

Namely show that Σ ` (f) and Δ ` (f) are:

$\begin{array}{c}{\mathrm{\&Sigma;}}^{\&prime;}\left(f\right)=0.5[{G^{\&prime;}}_1\left(f\right)+{G^{\&prime;}}_2\left(f\right)]\\ =0.5\frac{{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1+{\mathrm{\&alpha;\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1}{\sqrt{{|{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1|}^2+{|{\mathrm{\&alpha;\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1|}^2}}\frac{|{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2|}{{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2}\end{array}$

$\begin{array}{c}{\mathrm{\&Delta;}}^{\&prime;}\left(f\right)=0.5[{G^{\&prime;}}_1\left(f\right)-{G^{\&prime;}}_2\left(f\right)]\\ =0.5\frac{{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1-{\mathrm{\&alpha;\&beta;}}_1+{\mathrm{\&beta;\&alpha;}}_1}{\sqrt{{|{\mathrm{\&alpha;\&alpha;}}_1-{\mathrm{\&beta;\&beta;}}_1|}^2+{|{\mathrm{\&alpha;\&beta;}}_1-{\mathrm{\&beta;\&alpha;}}_1|}^2}}\frac{|{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2|}{{\mathrm{\&alpha;}}^2-{\mathrm{\&beta;}}^2}\end{array};$

Step S206, carries out the high-pass filtering of more than 200Hz to Σ ` (f) and Δ ` (f), in order to the low-frequency component of filtering at below 200Hz, namely obtain aural transmission filter Σ (f) and Δ (f).