2009年12月30日 星期三

Palatography

Palatography is a technique used to identify which parts of the mouth are used when making different sounds. This technique is often used by linguists doing field work on little-known natural languages. It involves painting a coloring agent, such as a dye or a mixture of charcoal and olive oil on the tongue or the roof of a person's mouth and having that person pronounce a specific letter or sound. A photograph is then made of the mouth roof and tongue in order to determine how the sound was articulated. The technique can also be performed electronically using a tool called a pseudo-palate, which consists of a retainer-like plate lined with electrods that is placed on the roof of the mouth while the speaker pronounces a sound. A record made through palatography is called a palatogram.
Electropalatography

Electropalatography (EPG) is a technique used to monitor contacts between the tongue and hard palate, particularly during articulation and speech.
A custom-made artificial palate is moulded to fit against a speaker's hard palate. The artificial palate contains electrodes exposed to the lingual surface. When contact occurs between the tongue surface and any of the electrodes, particularly between the lateral margins of the tongue and the borders of the hard palate, electronic signals are sent to an external processing unit. EPG provides dynamic real-time visual feedback of the location and timing of tongue contacts with the hard palate.This procedure can record details of tongue activity during speech. It can provide direct articulatory information

2009年12月19日 星期六

語音的音質
人類的語音大致上可以看作是一個由若干頻率不同的簡單聲波所構成的(F0),而與基礎頻率成整數比的頻率則稱為陪音(harmonics諧頻,或稱overtone泛音)。
  當我們發音時,聲帶振動的次數構成語音的基礎頻率,基礎率和陪音使(formants)。以下將語音分為母音及子音兩個部分討論之:
A.母音的聲學特性:
  母音的共振峰有五、六個。隨著發音腔道的大小及其形狀的改變,不同(F1,F2,F3)的不同分佈情形,決定了不同母音的聲響特性。一般說來,母音的共振峰結構都相當穩定,聲音的基礎頻率變化對母音,不論是任何人所發出的,其頻率大約都是固定的,不會有太懸殊的差異。
B.子音的聲學特性:
塞音(stops)在聲譜上可觀察到的特性有:送氣(aspiration),帶聲時間點(voice onset time),特強頻率變化形態(formant transition,這是辨別不同部位子音的最主要聲學特性)。
擦音(fricatives)因為發音時空汽油很窄的通道擠出,產生混亂的氣流,造成噪音(noice,或稱 white noice,類似收音機沒有調到電台播音頻道時發出沙沙聲的噪音)。
介音(glides 如 j, w)及流音(liquids 如 r, l)。這兩種聲音類似母音,有特別的特強頻率帶。
鼻音(nasal)也有特別的特強頻率帶的形態,與母音類似,但是鼻音因為共鳴的方式與口腔母音不同,formant的強度比較弱。
喉塞音(glottal stop)再聲譜與塞音相似,有一段靜止的時間以及帶升起點的輕微延緩,但喉塞音卻沒有特強頻率帶變化形態(formant transition)。
語音的聲學特性
語音的物理屬性
  語音如果做為一種物質材料,就跟自然界裡其他聲音有相同的一面,也具有物理屬性。一切聲音都是由物體振動而產生的,物體受外力的作用而振(邢公畹,1994,39-41)。因此,我們可以從這四個方面來分析和認識語音的物理特徵。
A.音高:
  就是聲音的高低。它主要決定於發音體振動的頻率,所以也有人把音高叫做音頻。
B.音強:
  音強就是聲音的強弱。它主要決定於振幅,而振幅指發音體振動的幅度。
C.音長:
  就是聲音長短。它主要決定於發音體振動時間的長短。振動持續的時間長,聲音就長,反之就短。
D.音色:
  指聲音的品質和特性而言。音色決定於聲波的形式。而音叉振動發出的聲波是頻率固定而有規則的,這種聲音叫純音。而其中頻率最低、振幅最大的那個純音叫做「基音」,其餘的叫「泛音」。
簡介
  在人類語言連鎖活動的過程中,最具體而可以直接觀察的部分是語音。語音可以聽得見,而且也可以記錄下來。從最具體的層次看,語音是聲波,聲波是空氣分子(particle)振動所導致的氣壓變化,是物理的現象,聲學語音學是研究語音的物理特性以及這些特性在語言系統中所具有的功能。簡單來說,聲學語音學就是一門研究「語音的物理基礎」的學問。
那麼我們為何要研究聲學語音學?因為那聽起來似乎是物理學家的工作,但是由人所發出的語音,其聲學上的特性也是語言學研究的一部份。現在,就讓我們簡單的說明幾點研究聲學語音學的理由吧!第一,我們對聲音的感覺取決於語音的聲學性質,因此,明瞭語音聲學性質的原理基本上是很重要的事。第二,語音不容易以發音動作來描述,因此以聲學特性為基礎比較容易解釋,比如母音可以以所謂的「特強頻率帶」(formant)的不同來顯示其差別。另外,聲學特性也方便我們去解釋一些容易混淆的語音。第三,說話的聲音是很短暫的,語音會隨著時間而消失。雖然語音可以模仿,甚至可以用符號記錄下來,但那與原本語音並不是同一回事,因此,獲得永久性的語音記錄是研究語音的一種助力。第四,由於近代聲學儀器的發明,我們不只可以使語音有重現上的可能,我們更可以使語音變成視覺上的記錄(如聲波圖、聲譜圖spectrograph等)。這麼一來,語音不只可以聽見,還可以“看見”。第五,在研究分析上,我們可以更進一步的克服語音瞬間消失的困難,使我們可以更詳盡的分析語音的聲學特質,以便和相關學科做結合,如:聽力學、特殊教育的聽障教學、語言治療等。
共振峰

共振峰是用來描述聲學共振現象的一種概念,[1]在語音科學及語音學中,描述的是人類聲道中的共振情形。常用的量測方法是由頻譜分析或聲譜圖(spectrogram,見右圖)中,尋找頻譜中的峰值。但假如說話者,用比較高的基頻發出母音,例如小孩或女性的聲音,則頻譜上看起來比較像是寬帶狀,比較無法看出明顯的峰值。在聲學中,共振峰是用來描述聲源內部的共振,特別是對樂器而言,指的是共嗚箱內的共振。

人類說話或唱歌產生的聲音包含許多不同的頻率,共振峰是這些頻率中較有意義的部份。定義上,人類若想分辨幾個不同的母音,我們所需要的資訊是完全可以被量化的。共振峰是使聽者能夠區分母音的關鍵泛音。大部份的這些共振峰是由管內或腔體的共振產生,

頻率最低的共振峰頻率稱為f1,第二低的是f2,而第三低的是f3。絕大多部分的情形是,前兩個共振峰,f1 和 f2就足以劃分不同母音。這兩個共振峰可以描述母音的開/閉、前/後兩個維度(過去傳統上把這和舌頭的位置聯結在一起,不過這不是完全精確)。因此開母音(例如[a])有比較高的第一共振峰頻率f1,而閉母音(例如 [i] 或 [u])的則比較低;前母音(例如[i])的第二共振峰頻率f2較高,後母音(例如[u])的則比較低。[2][3]母音幾乎都有四個以上的共振峰,有時還會超過六個。然而,前兩個共振峰還是最關鍵的。通常我們會用第一共振峰對第二共振峰的 關係圖描述不同母音的性質。[4] 但這不足以描述某些母音的性質,例如圓唇與否。[5]

2009年12月17日 星期四

Spectrogram

A spectrogram is an image that shows how the spectral density of a signal varies with time. Also known as spectral waterfalls, sonograms, voiceprints, or voicegrams, spectrograms are used to identify phonetic sounds. The instrument that generates a spectrogram is called a spectrograph or sonograph.

The most common format is a graph with two geometric dimensions: the horizontal axis represents time, the vertical axis is frequency; a third dimension indicating the amplitude of a particular frequency at a particular time is represented by the intensity or colour of each point in the image.

2009年12月16日 星期三

overview of Acoustic phonetics

Acoustic phonetics is a subfield of phonetics which deals with acoustic aspects of speech sounds. Acoustic phonetics investigates properties like the mean squared amplitude of a waveform, its duration, its fundamental frequency, or other properties of its frequency spectrum.

The study of acoustic phonetics was greatly enhanced in the late 19th century by the invention of the Edison phonograph. The phonograph allowed the speech signal to be recorded and then later processed and analyzed. By replaying the same speech signal from the phonograph several times, filtering it each time with a different band-pass filter, a spectrogram of the speech utterance could be built up. A series of papers by Ludimar Hermann published in Pflüger's Archiv in the last two decades of the 19th century investigated the spectral properties of vowels and consonants using the Edison phonograph, and it was in these papers that the term formant was first introduced. Hermann also played back vowel recordings made with the Edison phonograph at different speeds to distinguish between Willis' and Wheatstone's theories of vowel production.

Further advances in acoustic phonetics were made possible by the development of the telephone industry. (Incidentally, Alexander Graham Bell's father, Alexander Melville Bell, was a phonetician.) During World War II, work at the Bell Telephone Laboratories (which invented the spectrograph) greatly facilitated the systematic study of the spectral properties of periodic and aperiodic speech sounds, vocal tract resonances and vowel formants, voice quality, prosody, etc.

2009年11月9日 星期一

2009年10月20日 星期二

Speech synthesis

Speech synthesis is the artificial production of human speech. A computer system used for this purpose is called a speech synthesizer, and can be implemented in software or hardware. A text-to-speech (TTS) system converts normal language text into speech;

Synthesized speech can be created by concatenating pieces of recorded speech that are stored in a database. Systems differ in the size of the stored speech units; a system that stores phones or diphones provides the largest output range, but may lack clarity. For specific usage domains, the storage of entire words or sentences allows for high-quality output. Alternatively, a synthesizer can incorporate a model of the vocal tract and other human voice characteristics to create a completely "synthetic" voice output.

The quality of a speech synthesizer is judged by its similarity to the human voice and by its ability to be understood. An intelligible text-to-speech program allows people with visual impairments or reading disabilities to listen to written works on a home computer.

from http://en.wikipedia.org/wiki/Speech_synthesis

2009年10月6日 星期二

Alveolar trill

The alveolar trill is a type of consonantal sound, The symbol in the International Phonetic Alphabet that represents dental, alveolar, and postalveolar trills is [r].

In the majority of Indo-European languages, this sound is at least occasionally allophonic with an alveolar tap [ɾ], particularly in unstressed positions. Exceptions to this include Spanish, and Portuguese which treat them as separate phonemes.

Features of the alveolar trill:

Its manner of articulation is trill, which means it is produced by vibrations of the tongue against the place of articulation.

在一些語言, 例如捷克語, 齒齦顫音可以作為成節輔音 , 例如捷克語 krk (頸).

多數漢語中的方言都沒有此音。但湖北的中北部的一部分中原官話區和西南官話區里,即當陽江陵鍾祥京山一帶,直至神農架北部地區,存在有顫音/r/。而且此種顫音/r/,多半是由詞尾「子」演變而來。另外,某些普通話官話方言的民歌如鳳陽花鼓裡,歌詞「得兒叮噹飄一飄」裡,如果讀時舌肌放鬆,就會讀出清齒齦顫音。

from http://en.wikipedia.org/wiki/Alveolar_trill

2009年9月28日 星期一

Uvular trill

The uvular trill is a type of consonantal sound, used in some spoken languages. The symbol in the International Phonetic Alphabet that represents this sound is ʀ, a small capital R. Within Europe, the uvular trill seems to have originated in French, from where it spread to modern Standard German, most German dialects, some Dutch dialects, some northern Italian dialects, and the southern dialects of Swedish and Norwegian.

在這些語言的主要使用區域,濁小舌擦音成為小舌顫音的音位變體. 小舌顫音也在其它語言出現,尤其是患有語言障礙的人在無法或很難發出原來語種中標準的齒齦顫音.

少數語言使用不止一個顫音。例如,說奧克語的長者會區分齒齦和小舌顫音,如/ɡari/ 已治癒 和 /ɡaʀi/ 橡樹。

ps.老師是用類似早上刷牙漱口ㄉ方式來發小舌音.

from:http://en.wikipedia.org/wiki/Uvular_trill
Bilabial trill

The bilabial trill is a type of consonantal sound, used in some spoken languages. The symbol in the International Phonetic Alphabet that represents this sound is ʙ.

In many of the languages where the bilabial trill occurs, it occurs only as part of a prenasalised bilabial stop with trilled release, [mbʙ]. This developed historically from a prenasalized stop before a relatively high back vowel, such as [mbu]. In such instances, these sounds are usually still limited to the environment of a following [u].

(只有少數語言將此音視為獨立音素。在絕大部分情況,它只是作為前鼻化雙唇塞音的顫動除阻,讀音是 [mbʙ]。在歷史上,這是由較高的後元音緊接前鼻化輔音演變出來,例如[mbu]。這讀音僅出現在數種語言,而且往往限于這類情況才出現。例子有在馬努斯島東邊使用的Kele語中的[mbʙuen]「它的果實」。)

PS.老師說小時候有心理創傷ㄉ人~沒辦法發這ㄍ音.
這ㄍ論點是否成立還是會被推翻~~有機會ㄉ話找ㄍ人做做實驗吧!!!

from:http://en.wikipedia.org/wiki/Bilabial_trill
新春第一砲~~~~~~~~~~喔耶!!!!!!!!!