US006975732B2_ «y United States Patent (10) Patent No: US 6,975,732 B2 Nishiguchi et al. 5) Date of Patent: Dec. 13, 2005 (6!) AUDIO SIGNAL REPRODUCING S0m217 A &1901 Brau eta. 7042064 ‘APPARATUS Bosna A motos 104290 8 ag etal MM (75) Iaventors: Masayuki Nishiguehi, Kanagawa (IP); A °3, Wadhams 4602 8995 Absit et nati Yoshihito Fujiwara, Kanagawa (1P) A eee Teena ‘ seta 8 (73) Assignee: Somy Corporation, Tokyo (I) a Niigata sie (+) Nove: Subp oy dds he oft AY 5% Abgl n, pat ended erased unde 35 if oe Be USC. 1540) by 585 drs a Ning ca Stn a Thiel tal 34 a 0. mh chetal 31a (21) Appl. No.: 10/096,621 BI 381,300 (22) Filed: Mar, 13, 2002, Bo ‘Okada et al. 235/492 5 Nong ea. at oy Prior Publication Data m Manse Sea G8 Bate Nook "oa ‘us 20020080007 AL Ju. 1, 2002 OTHER PUBLICATIONS US. Appl No, 002.0038, filed Jan, 200, Fischer, (©) issn of apticton 80. 08747910, Hie on Now 12, David Weekly; “Very Sold Audio,” The Weed Report, p. 1980, mow Fal No” 9547, which fs a.dvsion Of 1-2 ication No.4 818, Hed ont 22 190,00 PL James D, Johto, “Trnform Coding of Audio Signals Related U.S. Application Data w SeHOAS8 Using Perceptual Noise Citra" IEEE Jounal on Selected (40) Foreign Application Priority Data ‘Aceas of Communication, vol 6, No.2, Feb. 1988. 4.251989 (P) PL278909 © cd by examiner (1) mc” cons HOAR V1; HOAR 25400 Primary Examiner—Xu Mei ©) US.Cl 381/74; 381/308; 381.384 (74) Attorney, Agent, or Firm—ay H. Mol (68) Field of Search 381/74, 26,303, ‘381/310, 370, 384, 2, 98, 374; 7041226, ABSTRACT 227, 28; 7004 small-sized audio signal reproducing apparatus for heat- ing reproduced audio signals withthe aid of a headphone is (56) References Cited closed, Digitized and compression encoded audio signals, NT DOCUMENTS Stored in a semicondueloe memory, are readout s0 a8 10 US PATENT DOCUMENTS "undergo a decoding operation, which san inverse operation 4812406 A 1071088 Pavel 38750 to compression encoding, v0 reproduce the audio signal, 488500 A 61084 stean ea 370522 andthe reproduced sgaals ae heard by the headphone. The 451408 A * S/BET Honal a a SSU781—appatatus may be significantly reduced in size al weight as ier ance fous $8174 compared tothe apparatus in which tape oF a disk is used 708465 A 1887 Paker Seone comes i ms 48233 A ABAD Weiner sijan3 38 he recording medium 49MW A 61980 Host 07 44968995 4101890 Lang 58554 4 Claims, § Drawing SheetsUS 6,975,732 B2 Sheet 1 of 5 Dec. 13, 2005 U.S. PatentU.S. Patent Dee. 13, 2005 Sheet 2 of 5 US 6,975,732 B2 FIG.3 > ic & 3 — ms aes Dee Ny Ete zou Bai aio ol c 7 Me E t iU.S. Patent Dee. 13, 2005 Sheet 3 of 5 'B1B2 Bs Bs Bs Be B7_ Be US 6,975,732 B2 Bs Bio B11 Bi2 INTENSITY INTENSITY \ \4 > FREQUENCY B1B2B3 B4 Bs Bs B7 Bs Bg FIG.S SB INTENSITY MS Bio B11 Biz FREQUENCY B1B2B3B4 Bs Be B7 Be Bo FIG.7 Bio B11 B12 FREQUENCYU.S. Patent Dee. 13, 2005 Sheet 4 of 5 US 6,975,732 B2 Ome 1Olmt 10m l0lma lOlme 1Olms Re _ FREQUENCYU.S. Patent Dee. 13, 2005 Sheet 5 of 5 US 6,975,732 B2 FIG.9US 6,975,732 B2 1 AUDIO SIGNAL REPRODUCING ‘APPARATUS: ‘Thisis division of prior aplication Ser. No, 08/787,910 filed Nov. 12, 1996 now US. Pat. No. 6.695.477, which is. division of application Ser. No. 07/600,818 filed Oct. 22, 1990 now US. Pat. 5,640,458. BACKGROUND OF THE INVENTION 1. Field ofthe Invention This invention relates to an audio signal reproducing apparatus ars, more particularly, to a small sized audio signal reproducing apparatus in which reproduced audio signals may be heard by a headphone. 2, Description of the rior Art Recently, small sized audio signal reproducing apparatus, in which audio signals may be heard by a beadphone, has become popular to provide for a more extensive range of appreciation of music or the like. With the smal sized audio signal reproducing apparatus for bearing the reproduced audio signals with a headphone, there iss demand for further reduction in size and weight singe such reduction possibly leads to more widespread use and mode of application. However, with the above described small sized audio ignal reproducing apparatus, magnetic tapes or optical isk, suchas, for example, compact disks, are used 28 the signal recording media, so thatthe apparatus in its eniety ‘cannot be reduced in size beyond the sizeof these recording ‘media. On the other hand, since the mechanieal parts for , and $ is the intensity of the Burke spectrum following the convolution. In be formula (1), (ama) represents ibe allow ance function, Since i is more preferred to decrease the numberof bits allocated to a higher frequeney band with correspondingly lesser energy value to reduce the mmiber of bits in their ene, the values of a and a ave set inthe present embodiment 38 and 1, respectively (a=38 and 2-1) for which satisfactory encoding may be achieved ‘Without deterioration in the sound quality “The level «, found asdeseribed above, i transmitted 0 a divider 17, which is designed for deconvolving the level {nthe convolved region, Ths the masking spectrum may be foun from the level eby such deconvolving operation. That is this masking specrum represents the allowable noise spectrum. Although a complicated operation is necessitated for such deconvolving operation, the deconvolving opera tion is performed with he present embodiment by using the simplified divider 17 ‘The masking spectrum is transmited via a symtesis circuit 18 toa subiracor 19, to which the ovtput from the sum detection eitcuit 14, that isthe Burke spectrum SB from the sum detetion czeuit 1, i supplied by means of a delay cireit 21. Thus the masking spectrum is subirected atthe subtractor 19 from the Bue spectrum SB, otha, asshown in FIG. 7, the Burke spectrum SB is masked at lower than a Fevel indicated by the respective levels of the masking speciram MS. ‘The output ofthe subractor 19 is supplied to quater 24 by means of a ROM 20. Inthe quantizer 24, the amplitude value Am supplied thereto by means of a delay circuit 23 is, «quantized with the numberof bits allocated in accordance ‘ith the output of the subiactor 19. In other words, the components of cach critical band are quantized in the «quantizer 24 with the number of bits which i allocated in accordance with the level of the difference between the energy of each critial band and ibe level ofthe ouput ofthe rise level setting means, The delay circuit 21 is designed for delaying the Burke spectrum SB from the sum detection circuit 14 in consideration of the delay caused in each circuit ‘component upsteam of the synthesis iui 18, whereas the delay circuit 23 is designed for delaying the amplitude value ‘Am in consideration of the delay caused in each circuit ‘componeat upsteam ofthe ROM 20, This ROM is provided for transient storage of the output ofthe sutactor 19 at 2 predetermined time interval during quantization for sbse~ quent eadout to quantizer 24 It wl be noted thatthe syihess in the above mentioned synthesis circuit 18 is performed for synthesizing or com- ining the data fora so-called minimum audible curve or qual loudness eure RC, characteristic ofthe human au tory sense, supplied from & minimum audible cuve genera- tor 2, a8 shown ia FIG. 8, with tbe above meationed ‘masking spectrum MS. By combining the minimum audible curve RC withthe masking spectrum MS in this manner, the allowable noise level may be set upto the upper limit of the hatched lin zone in FIG. 8, s0 that it becomes possible to reduce the aumierof bits for the hatched line zne in FIG 8 at the time of quantization. Meaniil, erica! bands,US 6,975,732 B2 7 similar to those shown in FIG. 4, are shown in FIG. 8, alongewith the signal spectrum SS. “The dat fom the buller memory 28 ae wansmitted 10 a ata volume operating circuit 26 wo finda data volume which is then transmited toa comparator 27. In the comparator, the ata volume i compared toa desired value ofthe aumber of bits per frame supplied from terminal 3 for bit rate adjust- ‘ment, The result of comparison is transmitted to the above ‘mentioned function controller 28, This function controller 28 controls the fupetion generator 29 10 compute the alow- ‘ance funetion to find the level «supplied to the subtractor 16, With changes inthis allowance function, the masking spectrum MS of FIG. 8 s variable controled in level. More specifically, the number of allocated bits may be increased ‘or decreased in thee entirety by parallel displacement of the masking spectrum MS in the direction ofthe sigaal level, that is, vertically in FIG, 8, withthe aumber of bits trans: sited during a predetermined time interval remaining con- Slant In this manner, the funtion generator 29 generates not ‘only the allowance Gunetion for finding the level e but also the function for bitrate adjustment, With the above described high eicieney compression encoding apparatus of the present emibodiment, the allow- able aose level i increased towards the side of the higher Frequencies with lower energy values to decrease the num- beraf bits allocated forthe side of the higher frequencies, so that it becomes possible to reduce the number of bis allocated for qusaization bwill be noted thal, inthe present embodiment, the above ‘mentioned bit ate adjustment andor the minimum audible curve syathesis may be eliminated. That is, i the bitrate agjusiment isto be eliminated, the data volume operating circuit 26, comparator 27 and the function contolle 28 are eliminated and the allowance funetion from function gen- erator 29 is fixed to, for example, (38). If the minimum audible curve is not synthesized, the minimum curve gen- erator 22 andthe synthesis circuit 18 are eliminated, so that the output from subtracior 16 is supplied dizcelly to suburactor, 19 after deconvolution a the divider 17 With the shove described high eliciency compression encoding, high fidelity audio signals may be transmitted at, Tor example, 64 kbps per channel, so that audio data about 44 minutes of longer may be stored in, for example, « 16 M bit make ROM chip. If four of such mask ROM chips are arranged 88 4 package, music software goods with «play- back time of about 17 minutes may be provided as.an article cof commerce. Although the playback time may be substan tilly halved fora stereophonic sound source, the number of bits per channel may be reduced futher by taking advantage ofthe cortelation between the lft and right channels, so that, by packaping four 16 M bit ROMS, a playback time of about twa and several minute may be realized. In view of the tendency towards « higher inlegration degree of semicon- ductor memories, mask ROMs with 128M bitsor 256 M bits, ‘would be commercially available in a not distant future Since further improvement in the high efiiency encoding twebnique can be reckoned, it would be possible to transmit high quality high fidelity audio signals at 64 Kbps per stereo-channel and to provide a one-chip audio signal soft- ware article wth a playback time of 33 or 66 minutes, I willbe noted that the audio sigaal reproducing appa ‘atus according tothe present invention may be applied not only to the ustal headphone device shown in FIG. 2, but to ‘headphone device 60 in which the headphone driver units, areexpased to ouside, as shown for example in FIG 9. With the headphone devioe 69, shown in FIG. 9, the headphone riven unit 61L, 61R are directly attached inthe users ears, 0 ™ 2 s 0 a 8 and a main body of the playback circuit section 63 is connected to these driver units 61L., 61R by means of signal transmission co¢d 62, The semiconductor memory 41, decoder 42, D/A converter 43 and batteries as the power source, shown in FIG. 1 are accommodated inthe main body of the playback circuit section 43. On the other hand, an ‘operational mode selector circuit 64 for controlling the playback operation and a volume dial 68 are provided on the ‘outer surlace of the main body of the playback cipcuit section 63, which may be implemented with te size and the shape of, for example, a pendant ‘The present invention is not limited to the above described embodiment. For example, an audio signal repro- ducing apparatus may be constituted by circuit elements including and downstream of the semiconductor memory 41 in FIG. I or may additionally include the A/D converter 32 and the encoder 33. The playback circuit section or the batteries may also be accommodated in a thickened head- phone band or in a headphone driver unit, The high eff- civacy compression encoding may be implemented by, for ‘example, sub-band encoding, instead of by adaptive trans- ‘form encoxing as described in the foregoing embodiment. If such sub-band encoding technique is used, the signal may be divided by, for example, band-pass filters, into a plurality of frequency bands, and the allowable nose level may be set so as to be progressively higher toward the side of the higher ‘requencies for the same energy level ‘What is claimed is: 1. A compact audio signal reproducing apparatus com- prising ‘4 memory unit for storing digitized and high-efliciency ‘compression encoded audio datas decoder for reading out data stored in said memory unit and decoding the read-out data by performing an inver- sion of the compression encoding; 4 digital’nalog converte for converting output signals from said decoder into analog signals; headphone unit for said analog signals from said digital analog converter and converting said analog signals ‘rom said digital/analog converter into acoustic signals; and « housing for containing said memory unit, said decoder, ‘and siid digitaanalog converter, said housing being ‘compact, ‘wherein said audio data is compression encoded by sub- band dividing audio data by dividing said audio data into a plurality of frequeney bands, and ‘quantizing a frequeney component transformed from audio data for cach of sad plurality of frequency bands according to a masking spectrum associated with said frequency component for each of said plurality of frequency bands and a minimum audible curve based ‘on the human auditory sense. 2. The compact audio signal reproducing apparatus according to claim 1 wherein said audio data is compression encoded by allowable noise level setting based on said masking spectrum and ssid minimum audible curve for exch of said plurality of frequency bands and quantizing said frequency component by the number of bits allocated in accordance with the level of a diller- ence between said lrequency component and said allowable nvise evel 3. The compact audio signal reproscing apparatus according to claim 1, furber comprising arate controller for controlling abit rate of said compression encoded audi data by increasing or decreasing said allocated mumber of bits associated with said frequency component,US 6,975,732 B2 9 4A portable audio player comprising: 4 memory unit configured to store digitized high- efficiency encoded audio dats; 4 decoder configured to read out data stored in said ‘memory unit and to decode the read-out data by por= forming an inversion of « compression encoding used to obtain the high-ellicency encoded audio data, 4 digital/analog coaverter configured to coavert output signals from said decoder inio analog signals for ‘cansmission toa headphone; and 4 portable housing containing said’ memory unit, said ‘decoder, and said digial’analog converter, 10 wherein said audio data is compression encoded by sub-band dividing audio data by dividing the audio data into a plurality of frequency bands, and quantizing a frequency component transformed from audio data foreach of the plurality of frequency bands according t© a masking spectrum associated with the frequency component for exch of the plurality of fre- quency band and a minimum audible curve based on ‘human auditory sense,