1263928 玖、發明說明： 【發明所屬之技術領域】 本i明係關於一種具有立八 攜式終端裝置。 、 J说之可輸入文字之可 【先ϋ技術】 現在（行動電話或PHS (註冊商標山#、 裝置除了基本t^ 寺之可攜式終端 其收發功能戋JL # ϋ e 牙J用笔子郵件作成及 刀此次其他各種應用。例如也 網際網路逵接#此里七 ’時間表作成功能或 J峪運接功旎寺，利用該等功 或URT m 7 ^ f耶需輸入預定内容 ^ URL (Uniform Resource Locator ； - ^ . 字。當然於如PDA (可攜式資訊终端^源定址崔器）等之文 置，-般亦可輸入文字。 、㈣)〈可攜式終端裝 利用如此之文字輸入功能之愔 他女杜、& 刀此义^ ^，例如於電子郵件或其 他又件 < 製作時，伴隨著文字之， _ in. ,, m 兀具疋在仃動電話， ^吏用如圖K)所示之鍵’由於為輸人手段之鍵（按紐）之個 ，限制’各文字與各鍵無法lf"對應，伴隨按下特定次數 相同鍵等繁雜之輸入操作。例如輸入「早安」時，若是先 前’之行動電話，要如「态」鍵5次、「（i」#i次、=」 鍵j /人、「态」键3次般地進行許多鍵，來選擇輸入文字。 =—方面’鍵是否被接受，要按照其操作，以鍵單位發出 發音頻率不同之確認音或使鍵本身發光，藉由此等可=確 認。 [本發明欲解決之問題] 然而，如上所述，於鍵與欲輸入之文字未對〗對應之行動 電話等之可攜式終端裝置，僅有键單位之確認音等於該時 83900 l263928 !二法萑璉擇〈文字(輸入候選文字)，為了正確進行文 入’要以目視確認按照鍵操作所顯示之 =輸入所希望之輸入文字。另-方面，不藉由目視而 :?紐“作時之記憶之情況，如前所伴隨繁雜之輸入操 ’就會在錯誤輸人狀態下繼續輸人，結果變成要再輸入。 此外，使用者為視障者時之文字輸入，明顯地變為相當困 難。 本毛月係鑒於上述之點而成者，於可輸入文字之可攜式 終缅裝置’提供-種藉由合成之聲音可確認輸入候選文字 之可攜式終端裝置。 【發明内容】 為解決上述之課題，申請專利範圍第丨項之發明，其特徵 在於：係一種可以輸入文字之可攜式終端裝置，其包含操 作手&，其係進行為了輸入文字之特定操作者；顯示手段， 其係顯示與该操作相對應之輸入候選文字者；及輸入控制 手段’係藉由由該操作手段接受確定輸入文字之操作，將 顯示於該顯示手段之輸入候選文字做為輸入文字輸入者； 且具備聲首合成手段’其顯示該輸入候選文字於該顯示手 段時，將該輸入候選文字之發音聲音合成後輸出。 此外，申請專利範圍第2項之發明，係於申請專利範圍第 1項之可攜式終端裝置，前述可攜式終端裝置係行動電話·， 且將前述聲首合成手段使用於聲音合成之音源與使用於前 述行動電話所具備之來電音生成之音源共用。 再者，申請專利範圍第3項之發明，係於申請專利範圍第 83900 1263928 2積 < 可攜式終端裝置，前述聲音合成手段使用於聲音合成 足首源係FM音源或波形表（WT)音源。 申叫專利範圍第4項之發明，係於申請專利範圍第1項之 可攜式終端裝置’前述操作手段包含可手動操作之複數按 4丑’於各按知分擔可輸入之複數文字。 申明專利範園第5項之發明，係於申請專利範圍第1項之 可^式終端裝置，前述操作手段可指示該輸入候選文字之 幸則入確足’别述顯示手段顯示指示該輸入確定之文字。 申請專利範園第6項之發明，係輸入文字於可攜式終端裝 置 < 方法’其進行：操作程序，其係進行為了輸入文字之 心作’卜員示心序’其顯示與該操作相對應之輸入候選文字； 耳曰a成敎序’其顯示該輸入候選文字時，將該輸入候選 又字 < 發音聲音合成後輸出；及輸入程序，其藉由由該操 作私序接受確定輸入文字之操作，將該被顯示且被發音之 知入候選文字做為輸入文字輸入。 申請專利範圍第7項之發明，係於該可攜式終端裝置被執 行之程式，該可攜式終端裝置具備為了對可攜式終端裝置 輸入又字而所操作之操作器；其進行··顯示程序，其顯示 由汶“作所指定之輸入候選文字；聲晋合成程序，其顯示 該輸入候選文字時，將該輸入候選文字之發音聲音合成後 輸出，及輛入程序，其藉由由該操作器接受確定輸入文字 之权作’將該被顯示且被發音之輸入候選文字做為輸入文 字輸入。 於本發明之可攜式終端裝置，顯示與為了文字輸入之特 83900 1263928 2操作之輸入候選文字時，聲音合成手段將該輸入候選文 Z又發音聲音合成後輸出。藉此，使用該可攜式終端裝置， ^ t仃動電話之使用者，藉由聽聲音合成之輸入候選文字 、毛曰，可以確認該輸入候選文字，故無需如先前目視確 ^輪入候選文字，方便性提高。料，即使使用者為視障 立j又字輸入也容易。再者，行動電話之情形，藉由將聲 ^成手&用於聲首合成之音源與用於前述行動電話所具 、脅之來兒首生成之骨源共用，故無需為了聲音合成手段而 追加新裝置，可以抑制製造成本之增加。 【實施方式】 以下參照圖面說明本發明之實施型態。此外，於以下之 踺明中，對同一構成要素賦予相同之符號。 +於圖1顯示本發明之可攜式終端裝置一實施型態之行動 包忐之結構。於圖}中，符號丨&係cpu (中央處理裝置），藉 由執仃下述之各種控制程式來控制行動電話1之各部動 作付唬lb係R〇M (Read 〇nly Mem〇ry ;唯讀記憶體）。此 lb儲存進行cpu u執行之傳送、來電等控制之各種電話功 此秸式、電子郵件之作成或控制其收發之郵件收發功能程 式、辅助樂曲播放處理之程式、辅助聲音合成處理之程式 等之私式、或預先記錄之樂曲資料及伴奏資料、聲音合成 所必須之參數或相關資訊等資料。該程式被設計成進行： ㉙777心序’其顯示由該操作所指定之輸入候選文字；聲音 合成私序’其顯示該輸入候選文字時，將該輸入候選文字 之％首聲首合成後輸出；輸入程序，其藉由由該操作器接 83900 1263928 受確定輸入文字之操作，將該被顯示且被發音之輸入候選 文字做為輸入文字輸入。 付號 lc係 RAM (Random Access Memory ;隨機存取：己 立 體），係設定CPU la之工作區域、下載之樂曲資料或伴奏資 料之儲存區域及儲存接收之電子郵件資料之郵件資料儲存 區域等。符號Id係通信裝置，進行以天線丨丨接收之信號解 調’並將傳送之信號調變而提供給天線丨丨。此外，符號^ 係輸入裝置，具有手動操作手段，其由包含設置於行動電 話1本體之「0」〜「9」之撥號按鈕之各種按鈕（鍵；後述圖 1之le)構成；檢測由此等手動操作手段之輸入。此操作手段 用於進行為了文字輸入之特定操作，包含可手動操作之複 數按紐，於各按紐預先分配可輸入之複數文字。 符號1 f係通話裝置，以通訊裝置丨d解調之受話信號由此、雨 話裝置1 f所具備之聲音CODEC所解碼後，由同裝置所具1 = D/A轉換器（皆未圖示）所D/A轉換後由受話口（耳揚聲器）^ 輸出。另一方面，由送話口（麥克風）lh輸入之聲音信號由g 同裝置所具備之A/D轉換器所數位化，同樣由同裝】戶=: 之聲音CODEC (皆未圖示)所壓縮編碼後，由通信裝置;」向 基地台傳送。作為此通話裝置lf之編碼/解碼之方式，使^ CELP (Code Excited Lpc，·代碼激發線性預測編碼）: ADPCM (適應差分PCM編碼）方式等之聲音資料之高效轉 縮編碼/解碼方式。 料而作為來電音 電子郵件作成時 符號Π係音源裝置，播放選擇之樂曲資 或保留音，由背面揚聲器•輸出。此外， 83900 1263928 ’、接受CPU la之控制，將其輸入候選文 咬：二 I合成之耸晋由背面揚聲器U輸出。有關此 :·首二成之詳細後述。此外，符號Ik係顯示裝置，由lcd (二rystalDlspiay;液晶顯示器)所構成，進行與電話功 月匕2子郵件收發功能之項目單或撥號按 操=對叙顧示。於文字輸人時，顯示輸人候選文4 , 各力此塊經由滙流排10進行資 料或命令之授受。 J貝 圖11係顯示本發明一眘 、乂 ,^ 〃她土悲〈伃動電話之外觀形狀之 圖。如圖所示，本彳彳1 # 甘* 仃力兒忐係小型構造，其一體裝入操作 ϋ品於 订有關去電及來電之操作’·通話部，其按 照#亥操作使通話可能·一、 七 > 次、π 匕，及頦不部，其可以顯示有關至少操 作之貝訊。具體而言 y ^ 發用之天線又，行動電話1具備無線收 n#且古七 又浩盗（揚聲器）4及送話器（麥克風）lh， 同時具有包含撥號鍵等操 蓄（顯示哭）A 宁知作鍵之知入手段le及圖像顯示裝 θ ^ ^可以將個人名及電話號碼等之 ~ …員不态1k。亦可以將接收之電子郵件顧 、人+ _ 了以依據本發明之方式輸入文字。 方；’过明有關音源裝晉1 ;少·^ 4 ιΛ 4丨ΠΠ ’、裝置I1之坪細。於本實施型能，后 樣地利用使用於來電立 心原 、我+ Α ㈢寺生成之¥知音源裝置，會現於入 候選又字之發音之聲立人士 、λ同 /、规如入 結構。 “成。1圖2顯示音源裝置li之概略 於圖2中，符號21之1263928 玖, invention description: [Technical field to which the invention pertains] The present invention relates to a terminal device having an eight-port type. J said that you can enter the text can be [first technology] now (mobile phone or PHS (registered trademark mountain #, device in addition to the basic t ^ temple portable terminal its transceiver function 戋 JL # ϋ e tooth J with a pen E-mail creation and other various applications of the knife. For example, also the Internet connection #本里七' timetable function or J峪运接功旎, use the work or URT m 7 ^ fye to enter the reservation content ^ URL (Uniform Resource Locator ; - ^ . Word. Of course, such as PDA (portable information terminal ^ source location Cui device) and other texts, can also input text. (4)) <portable terminal installation Such a text input function is the result of his female Du, & knives ^ ^, for example, in email or other pieces &<; when produced, along with the text, _ in. ,, m 兀 仃 仃 仃, ^ 吏 use the key shown in Figure K) 'Because it is the key of the input means (button), limit the 'each text and each key can not lf" corresponding, with the input of a specific number of times the same key and other complicated input For example, when entering "Good Morning", if it is the previous mobile phone, it should be like The "state" key is 5 times, "(i"#i times, =" key j / person, "state" key three times to perform a lot of keys to select the input text. = - aspect 'key is accepted, according to its In operation, a confirmation sound having a different pronunciation frequency is emitted in a key unit or the key itself is illuminated, whereby it can be confirmed. [The problem to be solved by the present invention] However, as described above, the key and the text to be input are not correct. For the portable terminal device such as the mobile phone, only the confirmation tone of the key unit is equal to the time 83900 l263928! The second method selects <text (enter candidate text), in order to correctly enter the text, it is necessary to visually confirm the key. The operation shows the input of the desired input text. On the other hand, it is not by visual observation: the condition of the memory of the "new time", such as the complicated input operation with the previous one, will continue in the wrong input state. In addition, the result becomes that you have to re-enter. In addition, the text input by the user as a visually impaired person becomes obviously quite difficult. This Maoyue is a portable end that can input text in view of the above points. Burmese device 'provided - species by synthesis The present invention claims to be a portable terminal device that can input characters, and includes a portable terminal device that can input characters. An operator' is a specific operator for inputting a text; a display means for displaying an input candidate character corresponding to the operation; and an input control means 'receiving the input text by the operation means In the operation, the input candidate character displayed on the display means is used as the input character input device, and the voice first synthesis means "when the input candidate character is displayed on the display means, the pronunciation sound of the input candidate character is combined and output. In addition, the invention of claim 2 is the portable terminal device of the first application of the patent scope, the portable terminal device is a mobile phone, and the sound initial synthesis means is used for the sound source of the sound synthesis. It is shared with the sound source generated by the call tone used in the aforementioned mobile phone. Furthermore, the invention of claim 3 is in the patent application scope 83900 1263928 2 < portable terminal device, and the sound synthesis means is used for sound synthesis of the source FM sound source or waveform table (WT) sound source. . The invention of claim 4 is the portable terminal device of the first application of the patent scope. The foregoing operation means includes a plurality of characters that can be manually operated and can be input by each of the plurality of characters. The invention of claim 5 of the Patent Fan Park is in the terminal device of claim 1 of the patent application scope, and the operation means can indicate that the input candidate character is fortunately entered into a sufficient indication that the display means indicates that the input is determined. Text. The invention of the sixth application of the patent model is the input of the text in the portable terminal device < method', which is carried out: the operation program, which performs the display and the operation for the input of the text Corresponding input candidate characters; the earcorder a is in the order of 'when the input candidate character is displayed, the input candidate is combined with the pronunciation sound and output; and the input program is determined by the private order acceptance of the operation The operation of inputting a text, and inputting the candidate character that is displayed and input as input text input. The invention of claim 7 is directed to a program executed by the portable terminal device, the portable terminal device having an operator operated to input a character to the portable terminal device; a display program for displaying an input candidate text specified by Wen's work; a sound synthesis program that displays the input candidate text, synthesizes the sound of the input candidate character, and outputs the program, and the program is entered by The operator accepts the right to determine the input text as 'the input candidate character that is displayed and is pronounced as the input text input. In the portable terminal device of the present invention, the display is operated with the special for the text input 83900 1263928 2 When the candidate character is input, the voice synthesizing means outputs the input candidate Z and the synthesized voice, and then outputs the candidate text by using the portable terminal device. , 曰 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , It is also easy to enter the word. In addition, the case of the mobile phone is used to make the sound source and the sound source for the initial synthesis and the bone source for the first generation of the mobile phone. In addition, it is not necessary to add a new device for the sound synthesis means, and it is possible to suppress an increase in the manufacturing cost. [Embodiment] Hereinafter, embodiments of the present invention will be described with reference to the drawings. The same symbol is shown in Fig. 1. The structure of the mobile terminal device of the present invention is shown in Fig. 1. In Fig. 1, the symbol 丨 & cpu (central processing device) is executed by The various control programs described above control the actions of each part of the mobile phone 1. The lb is R〇M (Read 〇nly Mem〇ry; read-only memory). This lb stores various calls for performing cpu u execution transmission, incoming calls, etc. Private, or pre-recorded music data and companionships such as mail processing, e-mail creation or control of the mail sending and receiving function program, auxiliary music playing processing program, auxiliary sound synthesizing processing program, etc. Data, parameters necessary for sound synthesis, or related information. The program is designed to perform: 29777 heart order 'which displays the input candidate text specified by the operation; sound synthesis private order' when displaying the input candidate text The first input of the input candidate text is first synthesized and output; the input program is operated by the operator to receive the input text, and the displayed input candidate text is used as the input text input. The payment number lc is RAM (Random Access Memory; random access: three-dimensional), which is the storage area of the CPU la, the storage area of the downloaded music material or accompaniment data, and the mail data storage area for storing the received e-mail data. . The symbol Id is a communication device that performs signal demodulation by the antenna ’ and transmits the transmitted signal to the antenna 调. Further, the symbol input device has a manual operation means composed of various buttons (keys; le of FIG. 1 described later) including dial buttons of "0" to "9" provided in the main body of the mobile phone 1; Wait for the input of manual operation means. This means of operation is used to perform a specific operation for text input, including a manually operable multiple button, and pre-assign the plural characters that can be entered in each button. The symbol 1 f is a communication device, and the received signal demodulated by the communication device 丨d is decoded by the audio CODEC provided by the rain device 1 f, and the same device has a 1 = D/A converter (all are not shown). The D/A conversion is output by the receiving port (ear speaker) ^. On the other hand, the sound signal input from the mouthpiece (microphone) lh is digitized by the A/D converter provided by the device, and is also the same as the sound code CODEC (not shown). After compression coding, it is transmitted by the communication device;" to the base station. As a method of encoding/decoding the communication device lf, ^ CELP (Code Excited Lpc) is an efficient encoding/decoding method for audio data such as ADPCM (Adaptive Differential PCM Coding). As a caller's tone, when the e-mail is created, the symbol is the source device, and the selected music or reserved sound is played back by the rear speaker. In addition, 83900 1263928 ’, accepts the control of CPU la, and inputs it into the candidate bite: the second I synthesis is output by the rear speaker U. This is the case: The details of the first two are described later. Further, the symbol Ik is a display device, which is composed of an LCD (two rystal Dlspiay; liquid crystal display), and performs a menu item for dialing and receiving functions of the telephone power, or a dialing operation. When the text is entered, the input candidate 4 is displayed, and each block is transmitted or received by the bus 10 via the data or command. J. Fig. 11 shows the appearance of the invention in a careful, 乂, ^ 〃 土 伃 伃 伃 伃 伃 伃 。 。 。 。 。 。. As shown in the figure, this 彳彳1 #甘* 仃力儿忐 is a small structure, which is integrated into the operation of the product to order the operation of the call and the caller's call, and the call is possible according to the #亥 operation. One, seven > times, π 匕, and 颏 ,, which can display at least the operation of the news. Specifically, the y ^ antenna is used again, and the mobile phone 1 has a wireless receiving n# and an ancient seven-tooth thief (speaker) 4 and a microphone (microphone) lh, and has an operation including a dialing button (showing crying). A is better known as the key to the key and the image display device θ ^ ^ can be personal names and phone numbers, etc. It is also possible to enter the text in accordance with the method of the present invention by receiving the email and the person + _. Fang; 'Tong Ming about the sound source installed Jin 1; less · ^ 4 ιΛ 4丨ΠΠ ’, the flat of the device I1. In this embodiment, the use of the sound source device used in the call of the original heart, I + Α (three) temple, will be in the voice of the candidate and the pronunciation of the word, λ with /, rules into structure. "成.1 Figure 2 shows the outline of the sound source device li. In Figure 2, the symbol 21
1 τ ^ ^ 輻出入I/F(介面）經由滙流排10,由CPU la接又為了播放來電 田 疋律寺首樂<樂曲序列資料或命人， 839〇〇 -10- 1263928 並且係為了將下述FIFO 22之狀態通知輸出於cpu ia之介面 電路。FIFO 22係包含 FIFO記憶體（First In First 〇ut mem〇ry ; 先進先出記憶體）之電路，暫時保持被給之樂曲序列資料 (①）’ k供給顯示於順次符號23之排序器（②）。此外，FIFO 22 通知CPU la記憶體之空間狀況（⑤），記憶體變空（Empty)之前 接受繼續之樂曲序列資料之轉送。 排序咨23由CPU 1 a接受發音開始/發音結束等命令（⑥），於 開始發音時，解釋由FIF〇 22接受之樂曲序列資料之同時， 计f時機而將各種參數或控制信號提供給Fm音源24 (詳細 後述）或WT音源25 (③、④），驅動該等音源。WT音源25如眾 所皆知，係將各種樂器聲或聲音等數位錄音，藉由大概或 反覆讀出預先儲存之波形記憶體26之波形資料，化實地再 現原本之樂器聲或聲音等。 FM首源24及WT (波形表）音源25之輸出以加法器27加 异，其輸出於數位/類比轉換器（未圖示）轉換為類比資料， 提供給背面揚聲器lj (圖丨）。一般於音源裝置Η，各音源係 經由FIFO 22及排序器23驅動，惟被要求即時性（即時響應性） 之效果音之種類，CPU la不經由FIFO 22及排序器23，而直 接驅動FM音源24或WT音源25。於本實施型態中之聲音合成 亦同樣地，CPU la直接驅動各音源。再者，波形記憶體％ 係使用ROM構成。 其次說明有關FM首源24。FM音源24—般組合複數圖3所示 之運算器30及加法器所構成。如圖3所示，}個運算器川包 含·· SIN波形表31，其記憶於SIN波形（正弦波之波形）之各相 83900 -11 - 1263928 位角點之波形振幅值；相位產生器（PG) 32，其由排序器幻 或CPU la接受頻率參數，基於此頻率參數生成為了 _波形表31使其輸出之_波形資料之頻率及相位：相位 位址信號後輸出加法器33，其將輸入信號及上述相位位址 加起來後提供給SIN波形表31;色絡產生器（EG)34，其由排 序器23或CPU la接受振幅參數，生成為了控制由該運算器 輸出之波形之振幅之色絡信號（振幅係數）後輸出，及乘法 為35，其乘以SIN波形表31之輸出及色絡產生器（EG)料之輸 出。 於如此構成之運算器30,記憶於SIN波形表31之sin波形之 振幅值係依據包含經由加法器33所提供之相位位址信號之 信號依序被讀出。因此，於此運算器％，藉由使讀出記憶 於SIN波形表31之波形振幅值之速度變化，即藉由適當控制 :供於SIN波形表3 1之相位位址信號，可以改變音高。例 如，放慢讀出速度則可生成低的音，提高讀出速度則可生 成高的音。再者，相位產生器（PG) 32接受重設信號，則將 輸出之相位位址信號重設（使由SIN波形表M讀出之位址回 到初始值）。 FM音源24如圖4(a)所示，將如此之運算器3〇複數串級連 接，或如同圖（b)所示，再進一步使用加法器，加上運算哭 30之輸出，以各式各樣地組合複數運算器儿及加法器，可 以生成無限多樣種類之聲音。於本實施型態中，利用揭示 於特公昭58-53351號公報等之所謂CSM聲音合成之技術，使 用具備於此行動電話1之FM音源24實現於該行動電話1之聲 83900 -12- 1263928 首合成。 於此，說明有關上述CSM聲音 短時間内可視為大致穩定。 0成疋原理。一般聲音於 間内聲音之頻 。此’ CSM聲音合成係在短時1 τ ^ ^ Radiated into I/F (interface) via bus bar 10, connected by CPU la and in order to play the caller Tian Yu Lawi first music &music; music sequence data or life, 839〇〇-10- 1263928 and The status notification of the FIFO 22 described below is output to the interface circuit of the cpu ia. The FIFO 22 is a circuit including a FIFO memory (First In First 〇 me mem〇ry; FIFO memory), and temporarily holds the music sequence data (1)' k supplied to the sequencer displayed in the sequential symbol 23 (2) ). In addition, the FIFO 22 notifies the CPU la memory of the spatial condition (5), and accepts the transfer of the continued music sequence data before the memory becomes empty. The sorting protocol 23 accepts the command (6) such as the start of pronunciation/end of pronunciation by the CPU 1 a, and explains the data of the music sequence received by the FIF 〇 22 at the same time of starting the sound, and provides various parameters or control signals to the Fm at the same time. The sound source 24 (described later in detail) or the WT sound source 25 (3, 4) drives the sound sources. The WT sound source 25 is well known for recording digital sounds or sounds of various instruments, and by roughly or repeatedly reading the waveform data of the pre-stored waveform memory 26, and realizing the original instrument sound or sound. The outputs of the FM head source 24 and the WT (waveform table) sound source 25 are added by the adder 27, and the output is converted into analog data by a digital/analog converter (not shown) and supplied to the rear speaker 1j (Fig. 。). Generally, in the sound source device, each sound source is driven by the FIFO 22 and the sequencer 23, but the type of sound effect required for immediacy (immediate responsiveness), the CPU la does not directly drive the FM sound source via the FIFO 22 and the sequencer 23. 24 or WT source 25. Similarly to the sound synthesis in the present embodiment, the CPU la directly drives the respective sound sources. Furthermore, the waveform memory % is constructed using a ROM. Next, the FM source 24 is explained. The FM sound source 24 is generally composed of an arithmetic unit 30 and an adder shown in Fig. 3. As shown in FIG. 3, the arithmetic unit includes a SIN waveform table 31, which is stored in the SIN waveform (the waveform of the sine wave), and the waveform amplitude value of each phase of the 83900 -11 - 1263928 bit angle; the phase generator ( PG) 32, which receives the frequency parameter by the sequencer phantom or the CPU la, and generates the frequency and phase of the _ waveform data for the _ waveform table 31 based on the frequency parameter: the phase address signal is outputted to the adder 33, which will The input signal and the above phase address are added up and supplied to the SIN waveform table 31; a color burst generator (EG) 34, which receives the amplitude parameter from the sequencer 23 or the CPU la, and generates an amplitude for controlling the waveform output by the operator. The color signal (amplitude coefficient) is output and multiplied by 35, which is multiplied by the output of the SIN waveform table 31 and the output of the color burst generator (EG). In the operator 30 thus constructed, the amplitude value of the sin waveform stored in the SIN waveform table 31 is sequentially read in accordance with the signal including the phase address signal supplied via the adder 33. Therefore, the operator % can change the pitch by reading the speed of the waveform amplitude value stored in the SIN waveform table 31, that is, by appropriately controlling the phase address signal supplied to the SIN waveform table 31. . For example, slowing down the read speed produces a low sound, and increasing the read speed produces a high sound. Further, when the phase generator (PG) 32 accepts the reset signal, the output phase address signal is reset (the address read by the SIN waveform table M is returned to the initial value). As shown in Fig. 4(a), the FM sound source 24 connects such an arithmetic unit 3 to a plurality of cascades, or as shown in Fig. (b), and further uses an adder, plus an output of the operation crying 30, in various forms. By combining a plurality of arithmetic units and an adder, it is possible to generate an infinite variety of sounds. In the present embodiment, the FM sound source 24 equipped with the mobile phone 1 is realized by the technology of the so-called CSM sound synthesis disclosed in Japanese Patent Publication No. Sho 58-53351, and the sound of the mobile phone 1 is implemented in the sound of the mobile phone 1 83900 -12-1263928 First synthesis. Here, it is explained that the above-mentioned CSM sound can be regarded as substantially stable in a short time. 0 is the principle of growth. The general sound is the frequency of the sound in the room. This 'CSM sound synthesis system is in a short time
几’—疋而進行馨立、A 將數⑽至數十ms之短時間之聲立^<合成。具體而言， 之和來表現聲音。依據離散時間：為Μ ’以數個正弦波 表示成： 1表現，耸音之時間系列{xt} 其中t係表示離散時刻之整數，不 Λ 4 M 係正弦波成分之個數（一般 為4〜6個左右），叫係第丨正弦 Π：玢、士 士八、上 夂风刀足角頻率（(Κωθπ)，Ai係 正弦波成分义振幅。 於此CSM聲音合成，對於以 夂I 、上述（1)式表示之模式，給予 多敷{〇卜"οοη、Ai...An}，藉由 η、斗、 x ^ 猎由（1)式，就各時刻t求出合成聲 曰系列{xt}。此時，對於有聲音立 二 曰（母曰或濁子音等），因有聲 洫具有周期性’故每此週J；月f p H]阪、 、1 (間^週期）將於（1)式之時刻t重 设成零而使相位初始化，另一女 - 力 万面，對於無聲音，因不具 周期性，故給予隨機週期，即*人卩、左 、 J即於队機 < 週期將時刻t重設而 使相位隨機地初始化。如此人成 > 敕丄 此口成'^耸首信號之時間系列接 近於人之聲音。 其次，說明有關此CSM聲音合成技術適用於FM音㈣（史 照圖十於⑴式所表示之各正弦波之成分可以使用前述之 運算器30生成。即’藉由對應於各正弦波之成分之_波形 表31，於時間系列使正弦波輸出（此時，各運算器如之輸入 信號足為零，相位產生器（PG) 32提供為了由SIN波形表3〗讀 83900 13 1263928 、^弦波之波形㊄料之相位位址信號（位址）），藉由下一級 :乘法器35，使其具有由色絡產生器⑽）34所提供之振 :可以由各運异态30得到（丨）式之各正弦波成分之信號之 輸=。然後，藉由以加法器5〇加上此等之輸出，可得到合 成聲首訊號之系列{Xt}。於CSM聲音合成，對於有聲音，每 其週期將時刻t重設成零，使相位初始化，並斜於無聲音， 於Ik機週期將時刻點t重設成零，使相位初始化，此相位之 初始化可藉由對於相位產生器（PG) 32於各週期提供重設信 號’使相位初始化來進行。 如以上，於使用FM音源24之CSM聲音合成，藉由合成複 數共振峰音，其係由提供給相位產生器（PG) 32之頻率參數 或重設信號及提供給色絡產生器（EG) 34之振幅參數之3要 素所合成者，決定音素，可進行聲音合成。例如聲音合成 樓花」時，藉由每數mS至數十mS設定複數組之上述3要 素’合/S/~>/A/->之6音素後發音。再者，小 的「。」、「妒」等或英文字之小寫等提高音程等來區別， 有關其他之記號亦預先決定易了解之說法使其發音即可。 提供給各運算器30之上述3要素，每個音素預先定義，登 記於ROM lb。此外，有關構成各文字之各音素之資訊，例 如「3」之情形，該文字由音素/S/^/A/所構成等之資訊， 亦同樣地登記於ROM lb。行動電話1於文字輸入時（後述之 文字輸入模式時），與先前相同地顯示對應於鍵操作之輸入 候選文字。然後，再進一步於顯示該輸入候選文字時，參 照有關登記於ROM lb之構成該輸入候選文字之音素之資 83900 -14- 1263928 訊，由所得之資訊再參照上述3要素之參數，其對應於構成 孫輸入候選文字之音素，每數mS至數+mS，提供頻率參數 或重設信號給相位產生器（PG) 32，並提供振幅參數給色絡 產生器（EG) 34，將輸入候選文字之發音聲音合成後輸出。 此外，於本發明之型態，使用FM音源24實行csm聲音合 成，惟當然顯然使用WT音源25亦可聲音合成。例如聲音合 成「櫻花」時，「夂，、「γ 「产 ^ 曰口 、人、 、」 〈」、&」數位錄音後儲存 ^己’撥放此等即可。然而，使用剛音源24進行c⑽ 聲首合成，必要之參數（資料）少即可，較為有利。 、其次二對於如此構成之本實施型態之可攜式終端裝置1 之來電等待模式時之動 說明。於此來電等待rr:::作流程圖加以 而起作用。 4 ’首源裝置11係為播放來電旋律 首先’ CPU la判斷有盔來 有來電為止（被判定為/為二(步知S61)，反覆此判斷直到 步驟S61之判斷判定為θ、' A此，作為有來電。則於 〗疋為疋，移至步騾S62。於步萨rmT , 將預先選擇設定作為來泰 > ^ ^S62 CPU la 至音源裝置ii。、人、、⑨疋律< M曲之樂曲相資料轉送 成來電旋律，持續播放該來電旋：〜曲序列資料’合 二二斷，於開狀態或關狀態(步驟叫。於此牛 I S63#疋相鍵為關狀態時（ ）I此步 線路斷路（步驟S6 為口争），再判斷是否 (判定為是時）回到步驟S64之判斷爾線路斷路時 回到步驟S63。 〜 ^非線路斷路時（判定為否時） 83900 -15· 1263928 另一方面，於步驟S63判定通話鍵為開狀態時（被判定 是時）’ CPU la給予音源裝置U終止來電旋律播放之 | 驟S65)。於此階段，音源裝置π終止目前正在播放之來二 律之播放。然後，進行一般通話時之處理（步驟％6)。 下來之步驟S67判斷結束通話鍵係開狀態或關狀態，反声此 判斷直到結束通話键成為開狀態為止（被判定為是為^ 然後，於此判斷，判定結束通話鍵為開狀態時（判定為是時 移至步驟S68。然後，於步驟S68進行結束通話時之處=(噪 路斷路），回到步驟S61。以上說明了於來電等待模式之由 未電至線路斷路之動作。 、其次，對於如此構成之本實施型態之可攜式終端裝置工 之文字輸入(文字輸入模式)時之動作，參照圖7所示之動作 流程圖加以說明。於此係經由使用者特定之操作，該可样 式終端裝置i處於文字輸入模式者。此外，於以下及圖7: “NKN”（新鍵號碼）及“〇KN”（舊鍵號碼）為變數，‘‘〜，，係表示 游標輸送鍵。再者’於_作為初始值設定數值鍵之代： 、、外之代碼。此外，變為文字輸入模式的係例如電子郵件 或時私表或其他文件之作成時’或網際網路連接時之侃[ <輸入時等，文字輸入成為必要時變成此模式。 首先，於步驟S71判斷鍵是否被碰觸（键開）。然後，反覆 此判斷直到使用者碰觸鍵為止。於此，輸入裝置u檢測： 用者之鍵碰觸’檢測出鍵碰觸時，將表示鍵被碰觸之鍵號 碼=知給CPUla。CPUla&輸人裝^到收到鍵號碼通知， 判疋為沒有键碰觸。於此係檢測出鍵碰觸（於步驟s7i，是 83900 -16- 1263928 的判足）此時，cpU la由輸入裝置丨6接到鍵號碼通知，將 接到之鍵號碼設定於變數NKN (步驟S72)。 其次，判斷設定於變數NKN之代碼（於此為鍵號碼）是否為 數值键之代碼（步驟S73)。於此，判定被設定於變數NKN之 代碼非數值鍵之代碼，則移至步驟S74。於步驟§74，再判 斷被設定於變數NKN之代碼是否為游標輸送鍵（「—」）之代 碼。於此步驟S74之判斷’判定被設定於變數nkn之代碼非 游標輸送鍵之代碼（否之判定時），則執行對應於另外規定 之其他鍵之處理（步驟S7y。然後，移至步驟s76。 於步驟S76’將變數NKN之代碼設定於變數〇kn後，回到 步知S71此外，使用者之鍵操作為特定之模式變更操作 時，即被設定於變數ΝΚΝ之代碼為對應於此模式變更操作 之鍵之代碼時，於步驟S75脫離圖7之文字輸入模式之流 程，終止文字輸入模式。 另一万面，於步騾S74之判斷，判定被設定於變數NKN之 代碼為游標輸送键之代碼（是之判定時），移至下個步驟 S77。然後，於步騾S77再判斷被設定於變數〇kn之代碼是否 為數值鍵之代碼。於步驟S77之判斷，判定被設定於變數 ΟΚΝ之代碼非數值鍵之代碼（否之判定時），則移至步驟 S79，於此步驟S79做使游標移動之處理。另一方面，於步 驟S77之判斷，判定被設定於變數〇ΚΝ之代碼為數值鍵之代 碼（是之判定時），此時確定已被輸入被顯示之表示候選文 字（顯示文字）作為輸入文字（步驟S78)。然後，移至步驟 S79，於此步騾S79做使游標移動之處理。步騾S79之處理結 83900 1263928 束，則移至步驟S76,將變數NKN之代碼設定於變數〇kn， 回到步騾S71。 於步騍S73之判斷，判定被設定於變數νκν之代碼為數值 鍵之代碼（是之判定時），則再判斷被設定於變數νκν之代碼 被設定於變數ΟΚΝ之代碼是否一致（步驟S8〇)。於此，判定 被設定於變數NKN之代碼與被設定於變數〇KN之代碼不一 致（否之判疋時）’則移至步騾S81。於步騾S81，再判斷被設 定於變數OKN之代碼是否為數值鍵之代碼。 於步驟S81之判斷，判定被設定於變數〇KN之代碼非數值 鍵之代碼（否之判定時），則於步驟S82使對應於被設定於變 數NKN之代碼之輸入候選文字（第丨候選）顯示於顯示裝置 lk，移至步騾S86。另一方面，於步驟S81之判斷，判定被設 定於變數OKN之代碼為數值鍵之代碼（是之判定時），於步驟 S83，將對應於現在作為輸入候選文字而被顯示之被設定於 變數OKN之代碼之文字確定作為輸入文字，以特定之樣態 使其顯示於顯示裝置lk。然後，於步騾S84，再使顯示於顯 不裝置lk之游標（於此，此游標係顯示輸入候選文字於被顯 示之位置者）顯示於下一個文字顯示位置，使對應於被設定 於變數NKN之代碼之輸入候選文字顯示於顯示裝置以之對 應位置（游標位置）後，移至步驟S86。 另一方面，於步驟S80之判斷，判定被設定於變數NKn之 代螞與被設定於變數OKN之代碼一致（是之判定時），則此時 因同一鍵又被鍵碰觸，故將現在顯示之輸入候選文字變更 成下一個輸入候選文字（S85)。具體來說，例如現在顯示之 839〇〇 -18- 1263928 輸入候選文字為「务」之情形，將此輸入候選文字變更成 「、、」，再顯示。然後，移至步驟S86。 於以上之步驟S82、S84、S85之各階段顯示輸入候選文字’ 然而與此此輸入候選文字之顯示同時，於步驟S86，將對應 於該輸入候選文字之頻率參數及振幅參數與於特定之定時 將重設信號轉送至晋源裝置1 i内之FM音源24，使該輸入候 選文字之發音聲音合成、輸出。之後，於步騾S76將設定於 變數NKN之代碼設定於變數ONK後回到步驟S71，以後於文 字輸入模式之間，反覆以上之處理。以上說明了於文字輸 入模式之動作。 如此，於本實施型態可以使用同一音源裝置^進行來電 等待模式時之來電旋律之播放及藉由文字輸入模式時之輸 入候選文字之聲音合成之播放。再者，於上述說明各動作 流程係一例，當然不限定於上述處理之流程。 於此，作為本實 " π调八^丨夭1¾又字之 不例及其發音例顯示於圖8、9加以說明。圖8係假名文字 入時之-例。大字候選文字顯示於符號81所示之輸入： 又=之輸入欄（假名漢字變換前之輸入攔）。同圖⑷顯示 入則之狀態。此外，最終確定之文字、 位置。 卞蜎不於付戒82所示 万；此，使用者按 圄令I% 调艾游標（顯示於 圖 < 下線）位置顯示「态」之文字，並 、 击人山 如其發骨/Α/聲 則（圖8(b))。再者，使用者按「1 输 一 & - 」鍵’則於輸入欄4 ㈣不位置顯示下一個文字之「V 、 ^ 」的又竽，與此同听 83900 -19 - 1263928 其發音/1/聲音合成、輸出（圖8(c))。其次，使用者按「6」键， 則先前輸入之輸入候選文字之「ρ」確定作為平假名之輸 入文字’游標移動一個文字量。然後，於此位置顯示下一 個出入候選文字之「泣」，將其發音/H/4/A/聲音合成、輸 出（圖8(d))。其次，使用者按「*」键，則於輸入欄之同一 游標位置顯示「试、」作為輸入候選文字，將其發音/BAVA/ 聲骨合成、輸出(圖8(e))。再者，按「*」鍵時之處理於圖7 所示之動作流程中，係於步驟S75之其他鍵處理被進行，此 進形’為了進行聲音合成，步驟S75之處理後，非步驟Μ 而是移到步騾S86。 其次’忒明英文字輸入時之一例(參照圖9)。於此例，賴 ^候選文字顯示於符號91所示之游標位置。_顯示輸Λ 則《狀悲。首先，使用者按「2」鍵，則於游標位置顯示「A 二二：且作為其發音將「“、」，即_(發音符號 Ο尸耳h成、輸出。再者，使用者按「2」鍵，則於同一游 標位置顯π英文字「B」，作為其發 (發音符號bi :)聲音合成、輸出。 」P ”―八 以上參照圖面詳細說明了本發每 > 处^ 體結構不限於此實施型態， 二她土恐。當然，具 之範圍之結構等。例如不脫離本發明要旨 英文字輸人，北㈣、上:m逑!：假名文字輸入或 或韓語、德語、法語、西班牙^、東1、台灣崎甲國話， 可以同樣實施。再者，於σ:::语等其他國語亦 之發音聲音合成，惟利用電話二=將輸入候選文字 刀此％入電話號碼之情形， 83900 -2〇> 1263928 =觸^而輸入之輸入文字（此情形為號碼），非輸入候選文 字而是輸入文字本身，於此情形亦與輸入候選文字同樣地 將该輸入文字聲音合成其發音即可。 [發明之效果] 如以上所詳細說明，依據本發明，顯示與為輸入文字之 特定操作相對應之輸入候選文字時，聲音合成手段將該輸 入候遠又字〈發首聲骨合成後輸出。藉此，使用該可攜式 :端裝置例如行動電話之使用者，因聽聲音合成之輸入候 選文竽之發音可以確認該輸入候選文字，故無需如同先前 般地目視輻入候選文字之顯示來確認，便利性提高。此外， 即使使用者為視障者，纟字輸人亦容易。再者，行動電話 < 6形，將聲首合成手段使用於聲音合成之音源與使用於 前述行動電話所具備之來電音生成之音源共用，無需為了 聲音合成手段而追加新裝置，可以抑制製造成本之增加。 【圖式簡單說明】 胃 圖1為顯示本發明一實施型態之行動電話結構之圖。 圖2為顯示同實施型態之音源裝置結構成之方塊圖。 圖3為頒不包含於同實施型態之FM音源裝置之運算器钟 構之方塊圖。 α 圖4(a)、圖4(b)為顯示於FM音源之運算器組合例之圖。 圖5為顯示藉由。“聲音合成執行來電旋律合成之立 源結構之圖。 曰 圖0為來電等待模式時之動作流程圖。 圖7為文字輸入模式時之動作流程圖。 83900 1263928 圖8(a)〜圖8(e)為顯示假名文字輸入時之輸入候選文字之 顯示例及其發音例之圖。 圖9(a)〜圖9(c)為顯示英文字輸入時之輸入候選文字之顯 示例及其發音例之圖。 圖10為顯示一般之行動電話之鍵（按鈕）之一例之圖。 圖11為顯示本發明一實施型態之行動電話外觀形狀之 圖。 【圖式代表符號說明】 1 行動電話（可攜式終端裝置） la CPU (聲音合成手段之一部分） lb ROM (聲立合成手段之一部分） lc RAM Id通信裝置 I e輸入裝置 If 通話裝置 lg耳揚聲器 lh麥克風 li音源裝置（聲音合成手段之一部分） lj 背面揚聲器 lk顯示裝置 II 天線 10 匯流排A few '-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Specifically, the sum expresses the sound. According to the discrete time: Μ ' expressed as a number of sine waves: 1 performance, the time series of the towering {xt} where t is the integer of discrete moments, not the number of 4 M sine wave components (generally 4 ~6 or so), called the 丨 丨 Π Π 玢 玢 玢 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 士 玢 玢 玢 玢 玢 玢 玢 玢 玢 玢 玢 玢 玢 玢 玢 玢 玢 玢 玢The pattern represented by the above formula (1) is given by a plurality of {〇卜"οοη, Ai...An}, and the synthetic sonar is obtained at each time t by η, bucket, and x ^ hunting by equation (1). Series {xt}. At this time, for the sound of the second 曰 (mother 浊 or turbid sound, etc.), because the vocal chord has a periodicity, so every week J; month fp H] saka, 1 (between ^ cycle) will At time t of (1), reset to zero and initialize the phase. Another female-powered face, for no sound, because there is no periodicity, so give a random period, that is, *人卩, left, J is the team The machine < cycle resets the time t and causes the phase to be randomly initialized. This is the result of this time. Secondly, it is explained that the CSM sound synthesis technique is applicable to the FM sound (4) (the composition of each sine wave represented by the equation (1) can be generated using the aforementioned operator 30. That is, by corresponding to each sine The wave component _ waveform table 31, the sine wave output in the time series (at this time, the input signal of each operator is zero, the phase generator (PG) 32 is provided for reading 830000 13 1263928 by SIN waveform table 3] The phase address signal (address) of the waveform of the chord wave is obtained by the next stage: multiplier 35, which has the vibration provided by the color generator (10) 34: 30. The signal of each sine wave component of the (丨) type is obtained. Then, by adding the output of the adder 5〇, the series of synthesized first sound signals {Xt} can be obtained. For the presence of sound, the time t is reset to zero every cycle, the phase is initialized, and the sound is oblique to no sound. The time point t is reset to zero in the Ik machine cycle to initialize the phase, and the phase can be initialized by Phase generator (PG) 32 provides reset in each cycle The signal 'initializes the phase. As described above, the CSM sound synthesis using the FM sound source 24, by synthesizing the complex formant peaks, is provided by the frequency parameter or reset signal supplied to the phase generator (PG) 32 and The three elements of the amplitude parameter of the color burst generator (EG) 34 are combined to determine the phoneme, and the sound can be synthesized. For example, when the sound is synthesized, the three elements of the complex array are set by mS to several tens of mS. '合/S/~>/A/-> 6 pronunciation after pronunciation. In addition, small ".", "妒", etc. or lowercase of English characters to improve the interval, etc., the other marks are also Predict the easy-to-understand statement so that it can be pronounced. The above three elements are supplied to each of the arithmetic units 30, and each of the phonemes is defined in advance and registered in the ROM lb. In addition, as for the information of the phonemes constituting each character, for example, in the case of "3", the information of the character composed of the phoneme/S/^/A/ is also registered in the ROM lb. The mobile phone 1 displays an input candidate character corresponding to the key operation as in the case of character input (in the character input mode described later). Then, when the input candidate character is further displayed, referring to the information on the phoneme constituting the input candidate word registered in the ROM lb, 83900 -14-1263828, and the obtained information refers to the parameters of the above three elements, which corresponds to The phoneme constituting the candidate character of the sun input, each mS to the number + mS, provides a frequency parameter or reset signal to the phase generator (PG) 32, and provides an amplitude parameter to the color network generator (EG) 34, which will input the candidate text. The pronunciation sound is synthesized and output. Further, in the form of the present invention, the csm sound synthesis is carried out using the FM sound source 24, but of course it is obvious that the WT sound source 25 can also be used for sound synthesis. For example, when the sound is combined into "Sakura", "夂," "γ", "曰口,人, ," <", &" digits are recorded and stored. However, it is advantageous to use the rigid sound source 24 for c(10) initial synthesis, and the necessary parameters (data) are small. And the second description of the call waiting mode of the portable terminal device 1 of the present embodiment thus constituted. Here, the caller waits for rr::: to work as a flowchart. 4 'The first source device 11 is to play the incoming call melody first' CPU la judges that there is a helmet to have an incoming call (it is judged as / is two (step S61), and this judgment is repeated until the judgment of step S61 is θ, 'A As there is an incoming call, then 疋 疋 疋, move to step S62. In step rmT, pre-selection settings as comet> ^ ^S62 CPU la to the sound source device ii., person, 9 law < The music of the M song is transferred to the incoming call melody, and the call is continuously played: ~ The sequence of the music is 'combined two or two off, in the open state or off state (step call. This cow I S63# 疋 phase key is off state) When ( ), the line is disconnected (step S6 is the contention), and it is judged whether or not (when the determination is YES), the process returns to step S63 when the circuit returns to the step S64. When the non-line is disconnected (the determination is no) 83900 -15· 1263928 On the other hand, when it is determined in step S63 that the call key is on (when judged to be YES), the CPU la gives the sound source device U to terminate the incoming melody play | Step S65). At this stage, the sound source device π terminates the playback that is currently playing. Then The process of the normal call is performed (step %6). Step S67 is judged to end the call key open state or the closed state, and the judgment is reversed until the end call key is turned on (it is determined to be ^ and then, When it is determined that the end call key is on (when the determination is YES, the process goes to step S68. Then, when the call is terminated in step S68 = (noise path is broken), the process returns to step S61. The call waiting mode is explained above. The operation from the absence of electricity to the disconnection of the line. Next, the operation of the character input (character input mode) of the portable terminal device of the present embodiment configured as described above is performed with reference to the action flow chart shown in FIG. Note that the user-specific operation is performed in the character input mode by the user-specific operation. In addition, in the following and FIG. 7: "NKN" (new key number) and "〇KN" (old key number) For the variable, ''~,' means the cursor transport key. In addition, 'the default value is set as the initial value of the value key: , , and the code outside. In addition, the text input mode is e.g. e-mail. When a piece or a private table or other file is created' or when the Internet is connected [ < when inputting, etc., the text input becomes necessary when necessary. First, it is judged in step S71 whether the key is touched (key open) Then, the judgment is repeated until the user touches the key. Here, the input device u detects: the user's key touches 'when the touch key is touched, the key number indicating that the key is touched = knowledge to CPUla CPUla& input device receives the notification of the key number, and judges that there is no key touch. This detects the key touch (in step s7i, it is the judgment of 83900 -16-1263928). At this time, cpU la The key number notification is received by the input device 丨6, and the received key number is set to the variable NKN (step S72). Next, it is judged whether or not the code set to the variable NKN (here, the key number) is the code of the numeric key (step S73). Here, if the code of the code non-numeric key set to the variable NKN is determined, the process proceeds to step S74. In step §74, it is determined whether the code set to the variable NKN is the code of the cursor transport key ("-"). In the judgment of the step S74, "the code of the code non-cursor transport key of the variable nnk is determined (if it is determined), the processing corresponding to the other predetermined key is executed (step S7y), and then the process proceeds to step s76. After the variable NKN is set to the variable 〇kn in step S76', the process returns to step S71. When the user's key operation is a specific mode change operation, the code set to the variable 为 corresponds to the mode. When the code of the operation key is changed, the flow of the character input mode of FIG. 7 is terminated in step S75, and the character input mode is terminated. On the other side, in the judgment of step S74, it is determined that the code set to the variable NKN is the cursor delivery key. The code (when it is determined) moves to the next step S77. Then, in step S77, it is judged whether or not the code set to the variable 〇kn is a code of the numerical value key. In the judgment of step S77, the determination is set to If the code of the variable ΟΚΝ code is not a numerical value key (NO at the time of determination), the process proceeds to step S79, where the process of moving the cursor is performed at step S79. On the other hand, in the judgment of step S77, the determination is determined. The code set in the variable 为 is the code of the numeric key (when it is determined), and at this time, it is determined that the display candidate character (display character) has been input as the input character (step S78). Then, the process proceeds to step S79. In step S79, the processing of moving the cursor is performed. If the processing of step S79 is 83900 1263928, the process moves to step S76, and the code of the variable NKN is set to the variable 〇kn, and returns to step S71. In the judgment of S73, it is judged whether the code set to the variable νκν is the code of the numerical value key (when it is determined), and it is determined whether or not the code set to the variable νκν is set to the variable ΟΚΝ (step S8 〇). Therefore, it is determined that the code set to the variable NKN does not match the code set in the variable 〇KN (if it is determined), then the process proceeds to step S81. In step S81, the code set to the variable OKN is determined again. Whether it is a code of a numeric key. In the judgment of step S81, when the code of the code non-numeric key set to the variable 〇KN is determined (NO determination), the code corresponding to the variable NKN is made in step S82. Input waiting The selected character (the second candidate) is displayed on the display device lk, and the process proceeds to step S86. On the other hand, in the judgment of step S81, it is determined that the code set to the variable OKN is the code of the numeric key (when it is determined), In step S83, the character corresponding to the code set to the variable OKN, which is currently displayed as the input candidate character, is determined as the input character, and is displayed on the display device lk in a specific state. Then, in step S84, The cursor displayed on the display device lk (where the cursor displays the input candidate character at the displayed position) is displayed at the next character display position, so that the input candidate character corresponding to the code set to the variable NKN is displayed. After the display device is at the corresponding position (the cursor position), the process proceeds to step S86. On the other hand, in the judgment of the step S80, it is determined that the generation of the variable NKn is the same as the code set to the variable OKN (when the determination is made), then the same key is touched by the key again, so now The input candidate characters to be displayed are changed to the next input candidate characters (S85). Specifically, for example, when 839〇〇 -18-1263828 is displayed, the candidate text is "service", and the input candidate text is changed to ",," and displayed. Then, the process moves to step S86. The input candidate text is displayed in each of the above steps S82, S84, and S85. However, at the same time as the display of the input candidate character, the frequency parameter and the amplitude parameter corresponding to the input candidate character are associated with the specific timing. The reset signal is forwarded to the FM sound source 24 in the source device 1 i, and the sound of the input candidate character is synthesized and output. Thereafter, in step S76, the code set to the variable NKN is set to the variable ONK, and the process returns to step S71, and thereafter, the above processing is repeated between the text input modes. The above describes the action in the text input mode. Thus, in the present embodiment, the same source device can be used to play the incoming call melody in the incoming call waiting mode and the sound synthesis of the input candidate characters by the character input mode. In addition, an example of each operation flow described above is of course not limited to the flow of the above processing. Here, as an example of the actual " π 调 丨夭 丨夭 丨夭 ⁄ ⁄ 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其 及其Figure 8 is an example of the pseudonym text. The large character candidate text is displayed in the input indicated by the symbol 81: the input field of the == (the input block before the kana kanji conversion). The same figure (4) shows the status of the entry. In addition, the final text, location.卞蜎 卞蜎 付 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 82 The sound is (Fig. 8(b)). In addition, the user presses the "1 to lose one & -" button to display the "V, ^" of the next character in the input field 4 (4), and listen to 83900 -19 - 1263928. 1/Sound synthesis, output (Fig. 8(c)). Next, when the user presses the "6" key, the "ρ" of the previously input input candidate character determines that the input text as a hiragana "the cursor moves by one character amount. Then, at this position, the next crying of the candidate text is displayed, and the pronunciation/H/4/A/sound is synthesized and outputted (Fig. 8(d)). Next, when the user presses the "*" key, "Test," is displayed as the input candidate character at the same cursor position in the input field, and the pronunciation/BAVA/synthesis is synthesized and output (Fig. 8(e)). Further, when the "*" key is pressed, the processing in the operation flow shown in FIG. 7 is performed, and the other key processing in step S75 is performed. This is performed for the sound synthesis, the processing of step S75, and the non-step processing. Instead, move to step S86. Next, an example of the input of the English word (see Figure 9). In this example, the candidate text is displayed at the cursor position indicated by symbol 91. _ display loses, then "sorrow." First, when the user presses the "2" button, "A 22 is displayed at the cursor position and " pronunciation" is ",", that is, _ (pronounced symbol Ο 耳 成 成 , , , , , , , , , , , , , , , , , When the 2" key is displayed, the English word "B" is displayed at the same cursor position, and the sound (combination symbol bi:) is synthesized and outputted as a sound. "P"--eight or more detailed description of the present hair is performed. The body structure is not limited to this embodiment, and she is afraid of it. Of course, the structure of the scope, etc. For example, without losing the essence of the present invention, the English word input, North (four), upper: m逑!: Kana input or Korean, German, French, Spanish^, East 1, Taiwanese, and Sakamoto can be implemented in the same way. In addition, in other languages such as σ:::, the sound is synthesized, but the phone number == will be entered as a candidate word knife. In the case of entering a telephone number, 83900 -2〇> 1263928 = inputting the input text (in this case, the number), inputting the text itself instead of inputting the candidate text, and in this case, similarly to inputting the candidate text Enter the text sound to synthesize its pronunciation. As described in detail above, according to the present invention, when the input candidate character corresponding to the specific operation of the input character is displayed, the sound synthesizing means outputs the input vocabulary and the initial vocal synthesis. The user who uses the portable type end device such as a mobile phone can confirm the input candidate character by listening to the pronunciation of the input candidate document of the voice synthesis, so that it is not necessary to visually confirm the display of the candidate word as before. In addition, even if the user is a visually impaired person, it is easy to lose the word. In addition, the mobile phone <6-shaped, the sound-synthesis method is used for the sound source of the sound synthesis and is used in the aforementioned mobile phone. The sound source generated by the caller tone is shared, and it is not necessary to add a new device for the sound synthesizing means, and the increase in the manufacturing cost can be suppressed. [Simplified description of the drawings] FIG. 1 is a view showing the structure of a mobile phone according to an embodiment of the present invention. 2 is a block diagram showing the structure of the sound source device of the same embodiment. FIG. 3 is an operation of an FM sound source device not included in the same embodiment. Block diagram of the clock structure. α Fig. 4(a) and Fig. 4(b) are diagrams showing an example of the combination of the operators of the FM sound source. Fig. 5 is a diagram showing the source structure of the call melody synthesis by "sound synthesis". Figure 0 is a flow chart of the operation in the call waiting mode. Fig. 7 is a flow chart of the operation in the text input mode. 83900 1263928 Fig. 8(a) to Fig. 8(e) are input candidates for displaying kana characters input. Fig. 9(a) to Fig. 9(c) are diagrams showing an example of display of input candidate characters and a pronunciation example thereof when an English character is input. Fig. 10 is a view showing a general mobile phone. Figure 11 is a diagram showing the appearance of a mobile phone according to an embodiment of the present invention. [Description of Symbols] 1 Mobile Phone (Portable Terminal Device) la CPU (Sound Synthesis) Part of the means) lb ROM (one part of the sound synthesis) lc RAM Id communication device I e input device If the call device lg ear speaker lh microphone li source device (part of the sound synthesis means) lj back speaker lk display device II antenna 10 exchange Row
21 輸出入I/F21 output I/F
22 FIFO 83900 排序器 FM音源 WT音源 波形記憶體 加法器 運算器 SIN波形表 相位產生器（PG) 加法器 色絡產生器（EG) 乘法器 加法器22 FIFO 83900 Sequencer FM Source WT Source Wave Memory Adder Operator SIN Waveform Phase Generator (PG) Adder Color Maker (EG) Multiplier Adder