Vous l’avez sans doute déjà repéré : sur la plateforme OpenEdition Books, une nouvelle interface vient d’être mise en ligne.
En cas d’anomalies au cours de votre navigation, vous pouvez nous les signaler par mail à l’adresse feedback[at]openedition[point]org.

Précédent Suivant

Berlin-Dresden Traditions in Experimental Phonetics and Speech Communication

p. 191-208


Texte intégral

1The early development of experimental phonetics which occurred mainly in France had a strong influence on advances in Germany at the beginning of the twentieth century. It is well known (Köster 1979, 1992) that Giulio Panconcelli-Calzia (1878- 1966) studied and then did his doctorate (1904) in Paris under the supervision of the father of experimental phonetics, Jean Rousselot (1846-1924). Later, Panconcelli-Calzia directed the famous Phonetic Laboratory in Hamburg from 1910 until 1949. He also cooperated with Hermann Gutzmann (1865-1922) who, in turn, was one of the leading scientists in the development of experimental phonetics in Berlin.

2With Hamburg, Berlin was the other center for experimental phonetics in Germany, reflecting a second very strong traditional line emanating from the physicist and physiologist Hermann von Helmholtz. We describe the development of experimental phonetics in Berlin in Section 1 of this paper.

3When automatic speech communication emerged as an engineering discipline in the second half of the twentieth century, experimental phonetics proved to be an essential contributor to this movement towards interdisciplinary research. In Eastern Germany, the University of Dresden formed the center for research and development into speech technology. We describe the range of early speech synthesis devices which were developed in Dresden in Section 2.

4Owing to their geographical proximity, the cooperation necessary between phoneticians and engineers was especially close where Berlin and Dresden were concerned. In Section 3, we describe some aspects of this cooperation, which was mainly directed at prosodic problems. It is well known that speech prosody is still one of the controversial subjects in speech technology.

5Finally, Section 4 is devoted to the material traces. At Dresden University, a collection of historic phonetic and acoustic devices demonstrates the development of the hardware from the beginning of the last century up to the introduction of processor-controlled equipment. This includes objects from its history in Berlin and Dresden as well as exhibits from Hamburg University that date back to Panconcelli-Calzia, mentioned at the beginning of this short survey.

1 The historic development of the phonetic sciences at Berlin University.

Image

1. Experimental Phonetics in Berlin

1.1 Early Contributors

6There is evidence that in Berlin scholars of various disciplines had long been working on questions related to human language and had tried to answer these using various methods from the natural sciences (Lindner 1961). One such scholar was Hermann von Helmholtz (1821-1894) who was well-known at that time for his outstanding work on the acoustic structure of sounds, in particular the structure of vowels produced by the human voice. He found that formants are the most intense parts of the speech spectrum which describes the vowel sound. The book Lehre von den Tonempfindungen (Theory of Sound Reception), published in 1863, is a summary of his epoch ground-breaking findings at that time. Carl Stumpf (1848-1936) continued the work on these issues and went further, using interference examinations, first testing stationary speech sounds. Only the brothers Wilhelm (1877-1946) and Ferdinand (1896-1973) Trendelenburg were able to perform tests on the modification of the voice signal: they succeeded in the oscillographic recording of continual sounds of voice and music.

2 The phonograph used in fieldwork (from Grieger 1989).

Image

7At Berlin University, phonetics was established as an institution created from two disciplines: linguistics and medicine. This development is illustrated in Figure 1, which will be explained in the following subsections. For a more detailed description, see Mehnert 2005.

1.2 From the Phonographic Commission to the Institute of Phonetics

8Let us look at the linguistic root first (the left column in Figure 1). The Phonographic Commission, founded in 1915, could be regarded as the predecessor to an institution which dealt with phonetics in its widest sense. This commission came under the authority of the Prussian Ministry of Science, Culture and National Education. It was created to record the voices of speakers representing foreign language speakers on wax cylinder or record (see Figure 2). These voices belonged to individuals who had been in German prisoner-of-war camps during World War I. The chairman of the Phonographic Commission was the director of the Psychological Institute at Berlin University, the aforementioned Carl Stumpf. In 1918, in a memorandum to the Prussian Ministry, Doegen (1877-1967) who managed the affairs of the commission, suggested the foundation of a sound institution and a sound commission to recruit competent experts who would assist in the assessment of the voice recordings. For more details, see Mehnert 1996.

9In 1920, a sound department was established at the Prussian State Library. Doegen became its director. Westermann (1875-1956), who was professor with a chair in African languages from 1925, made a proposal to develop the sound department into an Institute of Sound Research at Berlin University (Lindner 1976). On the institute’s program were research tasks in all fields of phonetics and related areas, teaching phonetics classes and special courses for teachers and others interested in the subject, the organization of lectures on ethnic studies and ethnic music and the continuation of scientific voice recordings. From 1935, the institute accommodated a linguistic department, a music department and a Phonetic Laboratory under the direction of Franz Wethlo (1866-1960). Thanks to Wethlo the study of experimental phonetics was introduced into the university’s curriculum.

10The Institute of Sound Research remained under the direction of Westermann after 1945 and was renamed the Institute of Comparative Phonetics. Westermann managed the affairs of the institute until he died in 1956. He focused his work on practical phonetic exercises from the perspective of language learning. In 1951, the institute was renamed the Institute of Phonetics. After Westermann’s death, the institute was directed by the African Studies scholar Ursula Feyer (1901-1989) until the erection of the Berlin Wall in 1961.

1.3 The Phonetic Laboratory

11The second root of phonetics at Berlin University (the right column of Figure 1) is represented by Hermann Gutzmann Senior (1865-1922), who worked as a voice and speech pathologist at the University. Gutzmann, who made Speech Therapy an integral part of the University’s curriculum, collected all the new instruments and research devices that had been used since 1900 in the new field of experimental phonetics. He continued the work of Helmholtz, Marey (pneumography), Zwardemaker (laryngography), Meyer (pitch measurement) and others and developed several instruments alone.

12It was on Gutzmann’s initiative that the first Phonetic Laboratory was founded in Berlin, making it possible to perform experimental research. In particular, phonetic research to examine voice and speech disorders was carried out in the laboratory. At the same time, the laboratory provided visual objects for lectures on the physiology of voice and speech which Gutzmann read to physicians as well as voice and speech therapists. According to the university prospectus, phonetics lectures had been offered from 1917 onwards in the categories of “Therapies” and “Linguistics”.

13Gutzmann’s Phonetic Laboratory was linked to the Outpatients’ Department for Voice and Speech Disorders in the ENT Clinic of the Charité. After an initial stage the development of the Phonetic Laboratory stagnated owing to World War I and Gutzmann Senior’s death in 1922. Moreover, the laboratory was later separated from the clinic. Under the direction of Eiken, the clinic accommodated both the clinical department for voice and speech disorders and the Phonetic Laboratory as a teaching and research institution, independent, however, of each other. In 1926, the Phonetic Laboratory became an independent institution under the direction of Franz Wethlo. As from 2 March 1926, Wethlo was assigned a teaching role in Experimental Phonetics, which gave him the opportunity of extending the laboratory and of purchasing new equipment. He developed numerous apparatuses, published abundantly and gave a great number of lectures (Wethlo 1954/55). His famous cushion pipe from 1913 represented an essential improvement in earlier larynx models by replacing the earlier membranes with air-pressurized cushions (see Figure 3).

3 Wethlo’s cushion pipe. a) Cross-section of the cushion pipe from the original publication (Wethlo 1913). b) A cushion pipe from the estate of Franz Wethlo, held in the Dresden University historic collection. c) Experimental setup for the reconstruction of the original experiment at the TU Dresden 2003.

Image

14After World War II and the re-opening of Berlin University in 1947, in 1950 the Phonetic Laboratory and the Outpatients’ Department for Voice and Speech Disorders became part of the Institute for Special Education which had just been founded together with the Faculty of Philosophy. The institute was directed by Reinhold Dahlmann (1883-1972). Dahlmann contributed to the Phonetic Laboratory by providing personnel and physical assistance. In addition to Wethlo, who in the beginning worked as a docent and later as a titular professor, the institute employed a research assistant, Gerhart Lindner (1925-2005, see Figure 4) and a mechanic. Besides his research tasks, Wethlo gave lectures for teachers of deaf-mute children and students, educators who worked to help individuals with impaired hearing and speech therapists and he maintained close contacts with expert groups and social institutions which dealt with phonetic issues.

4 Franz Wethlo (right) and Gerhart Lindner in the Phonetic Laboratory at Berlin Humboldt University at the end of the 1950s (private photograph).

Image

1.4 The Institute of Phonetics and Communication Sciences

15In 1962, the Institute of Phonetics and the Institute of Rhetoric were united into the Institute of Phonetics and Communication Science under the direction of the linguist Georg Friederich Meier (1919-1992). Gerhart Lindner (1925-2005), who had been Wethlo’s assistant for many years, was appointed professor for phonetics. This brought together experimental phonetics and the Phonetic Laboratory in the new institute (see lower part of Figure 1).

16As Meier was a linguist and Lindner a phonetician, this was the first time that humanities and the natural sciences had been united in one institute at the university. The interdisciplinary approach to phonetics was also reflected in the selection of the special fields represented by the co-workers. The fifteen members of the institute represented the fields of Communication Science, Linguistics, Phonetics, Psychology, Speech Acoustics and the Science of Speech. This interdisciplinary composition of the faculty paved the way for phonetics as an arts subject and as a natural science in both teaching and research. Lindner concentrated on perception tests for synthetic vowels and the recording and interpretation of the speech motion processes of the German language, and later also of the Russian language. He worked closely with the speech acoustician Dieter Mehnert (*1935), which was demonstrated in many joint projects. In addition to fundamental research, the results of the work were also transferred to practical phonetics (German as a foreign language, education of the deaf, see Figure 5).

5 At the Institute of Phonetics and Communication Science in Berlin: development of early solutions using animation software for rehabilitation purposes and foreign language education.

Image

17In the course of the Third University Reform in the GDR in 1969, the Institute of Phonetics and Communication Science was restructured and together with the Institute of Special Education formed the new section for Rehabilitation Pedagogics and Communication Science, in which the discipline of phonetics was accommodated in a unit for Phonetics/Speech Training. At the same time, the section had a technical unit which dealt mainly with Experimental Phonetics and Language Acoustics, under the direction of Dieter Mehnert.

18The year 1990 saw the reunification of the two German states and brought about fundamental changes in the structure of Berlin University. The section for Rehabilitation Pedagogics and Communication Science became the School of Rehabilitation Sciences, which, after the foundation of a Philosophical Faculty IV in 1994, became the Institute for Rehabilitation Sciences. In 1990, Mehnert was appointed Professor of Phonetics. He focused his work on experimental phonetics (intonation research) and speech analysis and synthesis, for their application to modern speech technologies (man-machine communication) and the rehabilitation of hearing and speech-language impairment. We will come back to this work in Section 3.

19As a result of the higher education reform at the three Berlin universities – Humboldt University, Free University, and the University of Technology – and the restructuring of the School of Rehabilitation Sciences in 1991/92, enrolment for the course of “Science of Speech/Specialization Voice and Speech Therapy” was terminated by decree in the autumn semester of 1993. This involved the cessation of the subject of Phonetics by the end of the year 1996 and the demise of a long tradition in this successful scientific domain at the Humboldt University Berlin following 80 years of scientific work.

20Although phonetics could not be re-established as a special field during the years after political breakdown, work in this field did continue, however. After a break of four years, Bernd Pompino-Marschall (*1950) was appointed to the Institute for German Language and Linguistics in the specialist area of Linguistics of the German Language/Phonetics/Phonology in the summer semester of 2000.

6 Principle of the vocoder. Figure from the first vocoder-like patent.

Image

2. Speech Acoustics at Dresden University of Technology

2.1 Early Development

21Dresden University was founded in 1828 as the Royal Saxon School of Polytechnics. From the 1870s Electrical Engineering developed in the Department of Mechanics. The development of communication engineering is closely connected to Heinrich Barkhausen (1881-1956), the “father of the electron valve”. He taught in Dresden from 1911 to 1953. He was very interested in acoustics, especially psychoacoustic problems (e.g., he invented the first measuring device for loudness). In 1938 he organized the founding of an Institute for Telecommunication and Technical Acoustics, directed by W. Wolman (1901-2003).

22In 1945 the University was destroyed by war. Among the institutes which were re-established step by step were:

  • Institut für Elektro- und Bauakustik (Electrical and Architectural Acoustics), founded by Walter Reichardt (1903-1985), and
  • Institut für Fernmeldetechnik (Telecommunication), founded by Kurt Freitag (1901-1977, see Figure 7).

23Both institutes worked on speech-related problems, but only the latter had a longterm impact, beginning with vocoder development.

7 At the Dresden Institute of Telecommunications in the 1960s. Left: Kurt Freitag (1901-1977), Head of the Institute. Right: Walter Tscheschner (1927-2004), (private photograph).

Image

24The vocoder is a device which aims to transmit only relevant parts of the speech signal. For this purpose, a combination of an analyzer and a synthesizer is required at the ends of the transmission line. We consider here the so-called channel vocoders which perform spectral separation of the signal (see Figure 6). The first working vocoder was developed by Homer Dudley (Bell) in 1936-39. The first German vocoder was constructed by the company Siemens & Halske. Parts of it are now in the Deutsches Museum Munich. The details and origin of the first device are unknown (probably military). Later, it was used for early electronic music.

25At Dresden University, a vocoder was developed by Eberhard Krocker (see Krocker 1957). According to a proposal from the famous psychoacoustician Eberhard Zwicker (Stuttgart/Munich), a channel scheme contrived according to the frequency groups (mel or bark scale) was adapted in this vocoder. The device was mainly used as an experimental base for investigating the speech signal. A large quantity of very profound experimental work was performed by Walter Tscheschner (e.g., Tscheschner 1961). He had had a position as an assistant since 1954 and finished his Dr.-Ing. thesis on the analysis of German speech with special reference to in stationary parts in 1961. In 1968, his habilitation thesis entitled Speech and its Recognition followed. He was awarded the first chair in speech communication at TU Dresden.

2.2 Research into Speech Synthesis in the Chair of Speech Communication

26The third University Reform in the GDR, which we mentioned above, also led to changed structures in Dresden, which now considered in particular the growing role of electronic computing in engineering. According to Figure 8, these parts of different former institutes were united into a new unit which contributed to the interface between the computer on the one hand and either the human or a process on the other. It is remarkable that in this way an institution targeting what we now call Man-Machine or Human-Computer interaction was created very early (1969). This unit for Computing and Measurement is the root of our current Institute of Acoustics and Speech Communication. A chair for Speech Communication was planned for this scientific unit, and Walter Tscheschner (1926-2004, see Figure 7) was the appropriate person for the post. He held this chair until his retirement in 1992 (see Hoffmann 2004).

8 The historic development of Speech Communication at Dresden University.

Image

27Based on his experience in speech analysis and synthesis, mainly with the above-mentioned vocoder, Walter Tscheschner initiated a broad research program that included Psychoacoustics (in cooperation with the Institute of Musical Instruments in Zwota), Speech Recognition and Speech Synthesis (see Tscheschner 1990). In this short presentation, we will consider only the speech synthesis branch. A more extended survey (but also restricted to speech synthesis) can be found in Hoffmann 2005.

28In the 1970s, parametric synthesis was the only reasonable way of operating synthesis terminals because the core memories of the computers were too small for storing speech samples. Walter Tscheschner’s group developed a series of formant synthesizers. These are summarized in Table 1. For most of these historic developments audio examples are available from the “Museum/Sprachsynthese” site of our web presentation at www.ias.et.tu-dresden.de/sprache. The technological progress from the electronic valve to the ASIC can be demonstrated by means of these synthesizer terminals, as can the controllers developed from the human controller via mainframe computers to DSPs.

29Formant synthesizers have been used mainly in rehabilitation engineering (aids for the blind). With the introduction of powerful personal computers, speech technology has been able to access

  • essentially more memory,
  • increased computing power.

30This has led to a general paradigm shift offering the following new options:

  • Capacity for storage of small speech segments, enabling concatenative speech synthesis in time-domain,
  • Computing power sufficient to include linguistic preprocessing and grapheme-to-phoneme conversion (Text-to-Speech, TTS).

9 The Dresden Speech Synthesis System DRESS links back to this hardware-supported version of a diphone based TTS system.

Image

31The group at the TU Dresden began the work on TTS with a hardware-supported solution: the Voice Card (Figure 9), this in cooperation with the Fraunhofer Institute (IMS 2) in Dresden.

32With the growing capabilities of processors, the hardware support offered by the Voice Card was required only in very rare cases. A purely software solution formed our baseline TTS system DRESS (DREsden Speech Synthesis) which is still undergoing further improvement (Hoffmann 1999). Inventories of different languages have been created. A special task was the development of a Mandarin syllable-based synthesis which procured experience in synthesizing tone languages.

33Our recent endeavors in speech synthesis are moving towards embedded applications. For example, our microDRESS system – a special version of DRESS – was developed for an industrial partner and was the first commercial TTS system with a total footprint of less than one Megabyte (Hoffmann et al. 2003).

3. Long-term Cooperation in Prosody Research

3.1 Historic Pitch Analysis prior to Franz Wethlo

34“Pitch determination is one of the most important but also most delicate problems in speech analysis.” This statement from the standard book on the use of electronics in this field (Hess 1983) describes a scientific problem which was well-known for a considerable time before the computer found its way into phonetic laboratories. Phoneticians became aware of the importance of pitch measurement approximately100 years ago (Panconcelli-Calzia 1907). How did they perform this without electronics?

10 In 1916, Calzia and Schneider developed a “throat sound recorder” which enabled the use of a kymograph for pitch measurements. Left: The original arrangement from Panconcelli-Calzia (1922). Right: Reconstruction of the equipment in the historic collection of Dresden University.

Image

35For a considerable time the “kymographion” was the basic device for this purpose (Panconcelli-Calzia 1936). It follows the principle of recording a waveform on a registration area which is moved by a driving force. This idea was applied for first time in 1734 to register wind parameters (anemograph by Ons-en-Bray). The later standard configuration was introduced by Ludwig in 1847 and further developed over the ensuing decades. It consists of the following main components: a revolving drum, the clock mechanism for driving the drum, and paper which is blackened with soot as a registration medium. This standard device was produced industrially and used worldwide.

36Depending on the application, different recording tools (writers) had to be added to the kymograph to produce a curve on the sooted paper. To record a sound, a writer from Krüger-Wirth was used from 1905, among other devices. From the recorded waveform, the pitch contour had to be calculated. Ernst Alfred Meyer developed a tool which supported this process. This was improved in 1913 by Stilke and Schneider.

37The writers were also improved. In 1916, Calzia and Schneider published the description of a writer ( “Throat sound recorder”) able to register vibrations up to 900 Hz, which is sufficient for pitch measurements (Schneider 1916). This was a very delicate construction with a blade of straw combined with a pig bristle as writing pen. A kymograph equipped with this writer is shown in Figure 10. Later, the writer was further improved to work in a frequency range of 200 Hz-2000 Hz (Schneider 1935).

38The kymograph was used over a long period. It must be mentioned that the Berlin phonetician Franz Wethlo improved the kymographic pitch measurement until 1954! He also introduced a measuring projector (Wethlo 1954, see Figure 11). This device makes the interpretation of the recorded waveforms easier by using magnification (10 x) and illumination. The length of a period of the magnified waveform is measured by the special vernier gauge, also from Wethlo’s estate.

11 Franz Wethlo improved kymographic pitch measurement until 1954. Left: The original aids from Franz Wethlo, to enable easier interpretation of the recorded waveforms. This equipment is now in the Dresden University historic collection. Right: Example for a measured “speech melody” from an original publication by Franz Wethlo (see Wethlo 1954).

Image

3.2 Steps in the Cooperation between Berlin and Dresden

39In the 1960s, the emerging speech technology in Dresden attracted phoneticians from Berlin. Close cooperation developed, summarized in Table 2. We will make some comments on the different steps.

40In 1963, two scientists from the German Academy of Sciences in Berlin, Alexander Vasilevič Isačenko (1910-1977, well-known as an important Russicist) and Hans-JoachimSchädlich (*1935, now a well-known novelist), demonstrated that “binary” intonation contours as shown in Figure 12 are sufficient to communicate the correct grammatical structure of a sentence to the receiver. The model describes a kind of minimal requirement for the intonation and forms an important base for the more recent quantitative models of intonation. We mention this model because the authors describe (Isačenko and Schädlich 1963) that the acoustic test material was prepared together with Walter Tscheschner using the vocoder in Dresden. The test sentences were monotonized with the vocoder. Then, the intonation contour was added under manual control.

41In the 1970s, developing automatic speech communication led to first experiences with speech synthesis systems, in our case with the Dresden SYNI system. It became clear that naturalness requires powerful models of prosodic parameters. In a parallel project in Berlin Dieter Mehnert analyzed the pitch contour of sound transitions (Mehnert 1975). Figure 13 shows as an example some classes of sound transitions. The results were applied in a microcontour generator for the Dresden formant synthesis system SYNI. Experiments gave first results on the perceptive relevance of pitch contours in synthetic speech.

12 Binary intonation patterns according to Isačenko and Schädlich’s intonation model, 1963. The acoustic patterns were produced using the vocoder at Dresden University.

Image

13 Example of microcontour research in the 1970s (from Mehnert 1975).

Image

42The following investigations focused on F0 macrocontours. In the 1980s Dieter Mehnert developed a basic model and some variants of suprasegmental F0 contours (Mehnert 1985, see Figure 14a). A contour generator was developed using hardware (Figure 14b). It was applied in two cases:

  1. Intonation control of synthetic speech, especially related to the Dresden speech synthesis terminal ROSY,
  2. Improvement in the intelligibility of electrolarynx speech.

43In the 1990s, Berlin-Dresden cooperation in prosody research was continued by Hansjörg Mixdorff, a scholar of Hiroya Fujisaki (Tokyo). He prepared and defended his PhD and Habilitation theses at Dresden University. In a first step (finished 1997), a quantitative model of German intonation based on the Fujisaki model was developed. An integrated model that noted the intonation contour as well as the rhythm (sound duration) constituted the second step (Mixdorff 2002). The models were developed using the Dresden TTS system DRESS. Further development of the prosodic component in DRESS is now being continued by Oliver Jokisch (see Jokisch et al. 2002, Mixdorff and Jokisch 2003).

44Finally we would like to mention the latest step in this cooperation in the field of prosody research. The third international conference on Speech Prosody was held in Dresden in May 2006 and was co-chaired by Hansjörg Mixdorff (Berlin) and Rüdiger Hoffmann (Dresden).

14 Example of macrocontour research in the 1980s. a) Typical contours (from Mehnert 1985). b) Intonation control of synthetic speech by combining a contour generator (top) with the Dresden formant synthesizer terminal ROSY.

Image

4. The Historic Collection

45Dresden University of Technology owns several historic collections. Among these, the acoustic-phonetic collection is located in the Department of Electrical Engineering and Information Technology.

46This collection expanded considerably after 1996. In that year, the Chair of Phonetics at the Humboldt University in Berlin was discontinued (see Figure 1). A number of scientific instruments which were related to the Berlin-Dresden cooperation described above, were added to the Dresden collection. Further exhibits came from other institutions working in phonetics or the rehabilitation sciences in other parts of Germany. We are now able to demonstrate the development of the hardware base from the beginning of the twentieth century until the introduction of computer controlling. Figures 3, 10, 11, and 15 show pieces from our collection.

47During the last few years, photographic documentation of the collection has been completed. A commented version is available on the web at www.ias.et.tu-dresden.de/sprache/museum/phon/index.html. We are working on reconstructing and publishing the most interesting parts of the collections (Mehnert and Hoffmann 2003, Hoffmann et al. 2004, Mehnert and Hoffmann 2004). Items from the collection have been used in various exhibitions and in lectures (as an example, see Figure 15).

48Finally, we must return to the chair of phonetics at Hamburg University, referred to at the beginning of this paper. This tradition-rich institution, which is linked to the names of Giulio Panconcelli-Calzia, Otto von Essen and other powerful scientists, was dissolved during the years 2005/2006. This institute owned a large historic collection which was described in an ownership catalogue (Grieger 1989). Because of the closure of the institute, the complete collection has been transferred to Dresden University as an unlimited loan. It is now part of a common exhibition in Dresden, called “Historic acoustic-phonetic collection” (HAPS) and described in Hoffmann and Mahnert 2007.

15 The mechanical Fourier analyzer from Mader (left, from the historic collection of Dresden University) is explained in our lectures on Signal Processing (right) in order to improve basic understanding of the algorithms.

Image

5. Acknowledgements

49The installation of the historic collection in Dresden was kindly supported by many members of the scientific community and also by the administration. Special thanks to Dr. Rolf Dietzel who patiently took photographs of all the items of our collection, among which those selected for this paper.

Bibliographie

Des DOI sont automatiquement ajoutés aux références bibliographiques par Bilbo, l’outil d’annotation bibliographique d’OpenEdition. Ces références bibliographiques peuvent être téléchargées dans les formats APA, Chicago et MLA.

References

GRIEGER Wingolf, 1989, Führer durch die Schausammlung, Phonetisches Institut, Hamburg, Hans Christians Verlag.

10.1007/978-3-642-81926-1 :

HESS Wolfgang, 1983, Pitch Determination of Speech Signals, Springer Verlag, Berlin.

HELMHOLTZ Hermann von, 1863, Die Lehre von den Tonempfindungen, Braunschweig, Vieweg & Sohn.

HOFFMANN Rüdiger, 1999, “A Multilingual Text-to-Speech System”, The Phonetician, n° 80, p. 5-10.

— 2004, “Obituary on Walter Tscheschner”, The Phonetician, n° 90, p. 28-29.

— 2005, “Sprachsynthese an der TU Dresden:Wurzeln und Entwicklung”, Beiträge zur Geschichte und neueren Entwicklung der Sprachakustik und Informationsverarbeitung, D. Wolf ed., Studientexte zur Sprachkommunikation, vol. 35, Dresden, w.e.b. Universitätsverlag, p. 55-77.

HOFFMANN Rüdiger, JOKISCH Oliver, HIRSCHFELD Diane, STRECHA Guntram, KRUSCHKE Hans, KORDON Ulrich, 2003, “A Multilingual TTS System with Less than 1 Megabyte Footprint for Embedded Applications”, Proceedings IEEE ICASSP1, Hong Kong, vol. 1, p. 532-535.

HOFFMANN Rüdiger, MEHNERT Dieter, DIETZEL Rolf, KORDON Ulrich, 2004, “Acoustic Experiments with Wethlo’s Larynx Model”, To the Memory of Wolfgang von Kempelen, Grazer Linguistische Studien, vol. 62, p. 51-60.

HOFFMANN Rüdiger, MEHNERT Dieter, 2007, “Early experimental phonetics in Germany – Historic traces in the collection of the TU Dresden”, Proceedings of the 16th International Congress of Phonetic Sciences, Saarbrücken, p. 881-884.

ISAČENKO Alexander V., SCHÄDLICH Hans-Joachim, 1963, “Erzeugung künstlicher deutscher Satzintonationen mit zwei kontrastierenden Tonstufen”, Monatsberichte der Deutschen Akademie der Wissenschaften zu Berlin, vol. 5, n° 6, p. 365-372.

10.1109/ICASSP.2002.1005766 :

JOKISCH Oliver, DING Hongwei, KRUSCHKE Hans, 2002, “Towards a Multilingual Prosody Model for Text-to-Speech”, Proceedings IEEE ICASSP, Orlando, Florida, p. 421-424.

KORDON Ulrich, 1998, “Sprachsynthesesystem im Geldkartenformat”, Dresdner Transferbrief, n° 6, p. 4.

KÖSTER Jens-Peter, 1979, “Giulio Panconcelli-Calzias geschichtlich-phonetisches Werk”, Hamburger Phonetische Beiträge, vol. 30, p. 71-85.

— 1992, “Giulio Panconcelli-Calzia”, The Phonetician, CL-61, p. 3-10.

KROCKER Eberhard, 1957, “Aufbau und Untersuchung eines Übertragungssystems für synthetische Sprache”, Wissenschaftliche Zeitschrift der TH Dresden, vol. 6, n° 4, p. 757-776.

LINDNER Gerhart, 1961, “Die Mitarbeit von Gelehrten der Berliner Universität an der Erforschung der akustischen Struktur der Sprache”, Wiss. Ztschr. der Humboldt-Universität zu Berlin, Ges.- und Sprachwiss. Reihe, vol. X, p. 113-116.

— 1976, “Diedrich Westermann und die Phonetik an der Berliner Universität”, Wiss. Zeitschr. der Humboldt-Universität zu Berlin, Ges.- und Sprachwiss. Reihe, vol. XXV, p. 189-190.

MEHNERT Dieter, 1975, Grundfrequenzanalyse und –synthese der stimmhaften Anregungsfunktion, ein Beitrag zur Erzeugung und Verarbeitung sprachlicher Signale, Dissertation, TU Dresden.

— 1985, Analyse und Synthese suprasegmentaler Intonationskonturen des Deutschen, ein Beitrag zur Optimierung technischer Sprachkommunikationssysteme, Habilitationsschrift, TU Dresden.

— 1996, “Historische Schallaufnahmen – Das Lautarchiv der Humboldt-Universität zu Berlin”, Elektronische Sprachsignalverarbeitung, D. Mehnert ed., Studientexte zur Sprachkommunikation, vol. 13, p. 28-45.

— 2005, “Phonetik an der Berliner Universität – ein Rückblick auf ihre Geschichte und auf Forschungsarbeiten der letzten Jahre”, Beiträge zur Geschichte und neueren Entwicklung der Sprachakustik und Informationsverarbeitung, D. Wolf ed., Studientexte zur Sprachkommunikation, vol. 35, Dresden, w.e.b. Universitätsverlag, p. 33-54.

MEHNERT Dieter, HOFFMANN Rüdiger, 2003, “Schallerzeugung, Schallaufzeichnung und Schallanalyse mit historischen phonetischen Instrumenten”, Signaltheorie und Signalverarbeitung, D. Wolf ed., Studientexte zur Sprachkommunikation, vol. 29, Dresden, w.e.b. Universitätsverlag, p. 77-91.

— 2004, “Measuring Pitch with Historic Phonetic Devices”, Proceedings of the Joint Congress CFA / DAGA2, Strasbourg, France, p. 1147-1148.

MIXDORFF Hansjörg, 2002, An Integrated Approach to Modeling German Prosody, Studientexte zur Sprachkommunikation, vol. 25, w.e.b. Universitätsverlag, Dresden.

MIXDORFF Hansjörg, JOKISCH Oliver, 2003, “Evaluating the Quality of an Integrated Model of German Prosody”, International Journal of Speech Technology, vol. 6, n° 1, p. 45-55.

PANCONCELLI-CALZIA Giulio, 1907, “Instrumentalphonetik oder Experimentalphonetik?”, Die neueren Sprachen, vol. XV, p. 121-197.

— 1922. Das Hamburger experimentalphonetische Praktikum, I. Teil, Hamburg, Verlag Otto Meissner.

— 1936, “Zur Geschichte des Kymographions”, Ztschr. f. Laryngologie, Rhinologie u. ihre Grenzgebiete, vol. 26, n° 3, p. 196-207.

SCHMIDT Karl-Otto, 1932, Verfahren zur besseren Ausnutzung des Übertragungsweges, D.R.P. 594976.

SCHNEIDER C., 1916, “Beschreibung eines neuen Kehlton-Schreibers nach Calzia-Schneider”, VOX – Internat. Zentralblatt für experimentelle Phonetik, vol. 26, n° 1, p. 1-3.

— 1935, “Verbesserungen an einem Kehltonschreiber”, VOX – Internat. Zentralblatt für experimentelle Phonetik, vol. 45, n° 1-6, p. 36-40.

TSCHESCHNER Walter, 1961, “Verfahren und Ergebnisse der Analyse deutscher Vokale und Konsonanten”, Proeedings 4th International Congress Phonetic Sciences in Helsinki, Mouton & Co., The Hague.

— 1990, “Programm und Ergebnisse der automatischen Sprachverarbeitung an der Technischen Universität Dresden”, Elektronische Sprachsignalverarbeitung, K. Fellbaum ed., Technische Universität Berlin, p. 1-10.

WETHLO Franz, 1913, “Versuche mit Polsterpfeifen”, Passow-Schaefers Beiträge für die gesamte Physiologie, vol. 6, n° 2, p. 268-280.

— 1954, “Phonetische Tonhöhenmessung mittels Schablonen”, Arch. Ohr- usw. Heilkunde u. Z. Hals- usw. Heilkunde, vol. 165, p. 563-567.

— 1954-1955, “Die Entwicklung des Laboratoriums für experimentelle Phonetik und Akustik”, Wiss. Ztschr. der Humboldt-Universität zu Berlin, Ges.- und Sprachwiss. Reihe, vol. IV, p. 89-90.

Notes de bas de page

1 IEEE: Institute of Electrical and Electronics Engineers; ICASP: International Conference on Acoustics, Speech, and Signal Processing.

2 CFA: Congrès français d’acoustique; DAGA: Deutsche Arbeitsgemeinshaft für Akustik

Précédent Suivant

Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.