Composer Research Articles

The analysis of Chinese National ballad composition education based on artificial intelligence and deep learning

This study focuses on generating national ballads with ethnic characteristics through algorithms, and exploring the application of artificial intelligence in music composition. As an important component of Chinese traditional culture, national ballads carry rich emotions and historical significance. However, due to the complexity and artistry of their creation process, traditional manual composition faces certain limitations. To address this, the study proposes a music composition model based on the combination of the Markov chain (MC) and Bidirectional Recurrent Neural Network (Bi-RNN). The model aims to generate melodies and emotional expressions that align with the style of ethnic national ballads. This method uses the MC to generate the basic framework of the melody, and adopts the Bi-RNN to further optimize rhythm and emotional expression. Experimental results show that, compared to traditional manual composition and the MC approach in music computing, the proposed method has significant advantages in melody creation and emotional expression. Besides, it can generate music that is consistent with ethnic styles. This study provides a new approach and method for the application of artificial intelligence in music composition, which has important implications for music education and cultural heritage preservation.

Algorithmic Orchestration: Deep Learning Techniques in Music Generation

Music generation using Deep Learning presents a comprehensive system for generating music in ABC notation using character based recurrent neural networks (RNNs), accompanied by a conversion pipeline that transforms the output music from ABC notation to a playable audio format. The integration of character-based RNNs allows for the creation of coherent and melodic musical compositions, while the ABC- to-MIDI-to WAV conversion enhances the usability and accessibility of the generated music. The work starts by preprocessing the ABC notation dataset and representing it at the character level. A character-based RNN, such as a long short-term memory (LSTM) network, is then employed to learn the sequential dependencies within the ABC notation and generate music that follows the learned patterns and structures. The RNN model is trained on a substantial corpus of ABC- encoded music, enabling it to capture the statistical regularities and nuances of the dataset. To make the generated music readily playable, the project incorporates a conversion pipeline that translates the output from ABC notation to MIDI format. MIDI files serve as a widely supported industry-standard representation of music, making them compatible with a variety of digital audio workstations (DAWs) and synthesizers. The MIDI files are subsequently converted to WAV format, a universally recognized audio format suitable for playback on diverse platforms and devices. This project contributes to the field of AI-generated music by providing an integrated system for music generation in ABC notation, accompanied by a seamless conversion pipeline for playback in the universally supported WAV audio format. The combination of character- based RNNs, ABC notation, and the ABC-to-MIDI-to-WAV conversion offers a valuable tool for musicians, composers, and music enthusiasts, facilitating the generation, sharing, and playback of musical compositions. Future research may explore advanced synthesis techniques, incorporate additional musical features, or investigate other music notation systems to further enhance the capabilities and usability of AI generated music. DOI:https://doi.org/10.52783/pst.1623

Melody AI – An AI Music Generator

MELODY AI is a music generation system based on AI that converts textual descriptions into novel musical pieces. Built with the Hugging Face model, the system uses deep learning algorithms to create music from user-input prompts. The process entails processing user-provided prompts, representation of these prompts in musical form, and synthesis of related audio outputs. The model has been optimized to preserve rich musical structures, allowing for coherent and expressive music to be generated across various genres. The model is successful in preserving the spirit of user inputs, generating expressive and relevant melodies. The outputs reflect the ability of the model to generate contextually relevant and beautiful music, making it an important asset for filmmakers, content creators, and artists. This study identifies developments in AI-augmented music composition and its potential to transform the creative sector. Keywords: AI music generation, deep learning, music synthesis, Hugging Face, text-to-music, MusicGen.

Strategy for Depicting the Character of Water in “Tabuh Kreasi Tirta Pemutih”

This study aims to provide a reference regarding the methods or techniques used by a Balinese Gamelan music composer in depicting a water character. Qualitative research is used because it is an appropriate type of research where the data obtained is a description of the condition of an object being studied. The data collection techniques used in this research are literature study, observation, interviews and documentation. There are several strategies that can be used to depict the character of water, which the composer previously used in the work “Tabuh Kreasi Tirta Pemutih”, such as processing the battens found in Balinese Gamelan as a medium for speaking, ornamenting the playing patterns of the instruments, and utilizing various song genres that can be played by Gamelan Semarandana. This study contributes to providing thoughts or points of view regarding strategies or techniques that can be used in depicting the character of water in a Balinese karawitan musical composition using gamelan as a medium for expression.

The First Viennese Music Composer of Vocal Chamber Music

In international musicology works, the mention of the first Viennese vocal chamber music composers usually gives a comprehensive introduction to the creative heritage of this genre. At this stage of the science of musicology, without the role of songs in the works of outstanding masters of the classical period, the conclusions of their musical heritage may not only be narrowed one-sided. In this context, this paper studies the first composer of vocal music through the analysis of the origin of the first Viennese school art songs, the rise of Vienna classical music school art songs, the common expression of Vienna classical music school songs, the artistic tension of Vienna classical music style songs, the analysis of different composers Vienna style songs, etc. Finally, it is necessary to fully consider the role of the works of the outstanding masters in the classical era, bring the style of the works into the composition, turn the boring notes into lifelike musical images, and fully express the true meaning of vocal music chamber music, so as to make the works endowed with strong vitality and appeal.

Integration of biomechanics and AI in music therapy: Exploring the impact of personalized music composition on psychosocial rehabilitation and data support

The purpose of this study is to explore the application of the integration of biomechanics and artificial intelligence (AI) technology in the field of music therapy and its impact on psychological rehabilitation. The results of the study show that the personalized music composition method based on biomechanics and AI technology can effectively improve the relevance and effectiveness of music therapy, and significantly promote the rehabilitation of patients with anxiety, depression and other psychological disorders. By analyzing the relationship between patients’ physiological and psychological data and music parameters, the superiority of personalized music therapy in terms of psychological recovery indicators is confirmed. The results of the study provide theoretical basis and practical guidance for the innovative development of the music therapy field, pointing out the future research direction of optimizing the AI model, expanding the scope of application and exploring the therapeutic mechanism in depth. The integration of biomechanics and AI in music therapy presents a novel approach to enhancing psychological rehabilitation. This addition will discuss the potential of this interdisciplinary approach to offer more precise, tailored treatments that can adapt to the individual needs of each patient in real-time. It will also address the ethical considerations and potential challenges associated with the use of advanced technologies in therapeutic settings, such as data privacy and algorithmic transparency.

Melody prediction of vocal performance using LSTM and attention mechanism and its application in folk music innovation

This paper combines Long Short-Term Memory (LSTM) and Self-Attention mechanisms to predict melody direction in vocal performances, exploring its application in folk music innovation to enhance expressiveness and creative diversity. Using the MAESTRO dataset, which includes audio and MIDI data, the continuous note sequence is divided into fixed-length segments (20 notes each) via a sliding window. The LSTM model captures temporal dependencies in the sequence, while Self-Attention assigns varying weights to inputs across time steps to better capture global context. To overcome the limitations of traditional pitch prediction methods, 10-fold cross-validation is employed to evaluate model performance. Experimental results show that with a window size of 20, the Mean Squared Error (MSE) is 0.023, and training takes 86 minutes, yielding the most balanced results across all configurations. Compared to Bi-LSTM and MusicTransformer, the proposed model excels in pitch prediction accuracy, achieving an average Mean Absolute Error (MAE) of 0.022, R2 of 0.895, and Pearson Correlation Coefficient (PCC) of 0.925. The model is also tested in the creation of six types of folk music, demonstrating that while it reduces creation time, the harmony consistency slightly lags behind manually created melodies. This model has significant potential in pitch prediction and folk music creation, offering a valuable tool for music composition with high practical application value.

Research on the Application of AI Technology in Music Composition Teaching

With the rapid development of artificial intelligence technology, AI applications in education have become increasingly widespread, bringing new opportunities and challenges to music composition teaching. Based on constructivist learning theory, this study deeply explores the application value and implementation paths of AI technology in music composition teaching. Through analyzing the functional positioning of AI technology in music composition teaching, the dual role of artificial intelligence as an auxiliary tool and teaching interaction facilitator is elaborated. The research finds that AI technology can effectively support the construction of personalized music composition learning models and promote the optimization and upgrade of teaching evaluation systems. On this basis, optimization strategies for improving teacher professional literacy and integrating teaching resources are proposed, providing theoretical reference for promoting the deep integration of AI technology and music composition teaching. The study shows that the introduction of AI technology can not only enhance students' music composition ability and learning interest but also promote the overall improvement of teaching quality, which has important significance for advancing the modernization of music education.

Katha AI : A Story Generator with image, music, and voice synthesis capabilities

t—This paper presents an AI-driven story generator that integrates text generation, image creation, music composition, and text-to-speech (TTS) capabilities into a single system. The objective of the project is to provide users with an immersive experience, allowing them to generate interactive stories enhanced with multimedia content. The system leverages natural language processing (NLP) for story generation, generative adversarial networks (GANs) for image creation, deep learning-based models for music composition, and TTS models for vocal narration. This paper outlines the system architecture, methods, and potential applications of the AI-based tool. Keywords—AI, story generation, image generation, music composition, TTS, natural language processing, generative adversarial networks, multimedia.

“She Must Be a Pure Vessel”: An Examination of a Spirit Medium Persona

Rosemary Brown (1916-2001) is certainly a highly unusual case in Music History. In the 1960s, she started to notate hundreds of musical pieces that she attributed to the spirits of several great concert music composers, with whom she claimed to be in touch as a spirit medium. Brown also furnishes a promising persona case study. In order to convince the public that her music had a spiritual origin, she described herself as a simple housewife and mother with no profound musical knowledge, therefore hardly capable of writing original musical pieces in the styles of acclaimed composers. The purpose of this paper is first, to provide an examination of Rosemary Brown’s public persona; second, to relate it to spiritualist tradition, in order to demonstrate that Rosemary Brown’s persona features were available in the spiritualist cultural repertoire; third, to discuss the implications of gender for the understanding of mediumship among spiritualists.

Exploring the Realm of Automated Music Composition with Models based on Transformer and its Variants

The Transformer architecture and its modelling variants have great potential in the music domain. Traditional approaches, such as rule-based systems and RNNs, have limitations in capturing the complex temporal dependencies and hierarchical structures inherent in music. With the introduction of its self-concern mechanism, the Transformer can effectively address this problem by capturing remote dependencies and allowing parallel processing of sequences. Transformer-based model variants such as Music Transformers, MuseNet and Jukebox have demonstrated that they can generate high-quality, varied, and stylistically rich compositions. Despite the success of these models, some challenges still need to be solved, such as high computational requirements and limited control over musical style and emotion. Possible future research directions include optimizing computational efficiency, enhancing stylistic and emotional control, and developing hybrid models that combine other models with Transformer. This article provides a comprehensive overview of the current state of research and potential and future applications of Transformer for music generation.

Preservice Teachers Use of Digital and Non-Digital Music Composition Systems as Resources for Teaching Elementary Mathematics

Preservice Teachers Use of Digital and Non-Digital Music Composition Systems as Resources for Teaching Elementary Mathematics

Unveiling the Use of Networked Ontologies to Develop a Supporting Tool for UX Evaluation in an Immersive Context

Immersive technologies have emerged as a new type of interactive system that aims to provide users with immersive experiences. They have been adopted in various fields and are gradually becoming part of our lives. UX is a key quality attribute to evaluate or model such experiences. However, when it comes to immersive experiences, evaluating UX is particularly challenging because the user should not be interrupted to provide feedback. Aiming at giving a step to address this issue, we have explored using ontologies from an ontology network to support evaluating immersive experiences. In this work, we adopted the Human-Computer Interaction Ontology Network (HCI-ON) and used an extract containing concepts from some of its networked ontologies to develop the User eXperience evaluation based on Ontology Network (UXON), an ontology-based tool that supports UX experts evaluating immersive experiences based on data recorded in interaction logs. HCI-ON is a framework for organizing knowledge of the HCI domain, offering a general understanding of the field, regardless of specific solutions. UXON was used to evaluate the UX of Compomus, an immersive application that supports collaborative music composition. UXON extracts data from the application interaction logs, calculates UX metrics, and provides consolidated data and information in graphs and tables. We conducted a study and collected feedback from the UXON developer and three UX experts who used the tool. Results showed that using networked ontologies to develop a tool to support UX evaluation is feasible and valuable. In summary, the ontologies helped at the conceptual level by offering a basis to define the system's structural model and at the implementation level by assigning semantics to data to make inferences about UX. Based on the UX experts' perceptions, UXON was considered a promising system, beneficial, helpful, and easy to use. The conceptualization used to develop UXON was evaluated by HCI experts and it was considered adequate and understandable, having the potential to be used by other people to solve HCI evaluation problems.

The use of the Kodály method in the creation of modern Chinese music / El uso del método Kodály en la creación de música china moderna

The primary objective of this article is to analyse the application of the Kodály method for the creation of contemporary Chinese music, which depends on creative skills, emotional expression and musical hearing. In pursuit of this inquiry, a cohort of 118 third-year students specializing in composition was enlisted from diverse academic institutions, namely Ningbo University, Beijing International Art School and the Shanghai Conservatory of Music. The fulfilment of the article’s intended objective was made feasible through the employment of an analytical methodology, the utilization of the Thurstone scale, as well as the application of Mann-Whitney coefficients to ascertain pertinent statistical significance. Additionally, the evaluation of the resultant musical compositions was undertaken to appraise their artistic quality. The obtained findings substantiate the efficacy of the employed Kodály method in the genesis of musical compositions. In light of these results, the authors proffer a recommendation for the implementation of a comprehensive pedagogical approach, one that encompasses a multifaceted acquisition of musical knowledge, diverse melodic orientations and variations in compositional elements, as well as strategies for the fusion of textual and musical components.

AI-based Chinese-style music generation from video content: a study on cross-modal analysis and generation methods

In recent years, Artificial Intelligence Generated Content (AIGC) technologies have advanced rapidly, with models such as Stable Diffusion and GPT garnering significant attention across various domains. Against this backdrop, AI-driven music composition techniques have also produced significant progress. However, no existing model has yet demonstrated the capability to generate Chinese-style music corresponding to Chinese-style videos. To address this gap, this study proposes a novel Chinese-style video music generation model based on the Latent Diffusion Model (LDM) and Diffusion Transformers (DiT). Experimental results demonstrate that the proposed model generates Chinese-style music from Chinese-style videos and achieves performance comparable to the baseline models in audio quality, distribution fitting, musicality, rhythmic stability, and audio-visual synchronization. These findings indicate that the model captures the stylistic features of Chinese music. This research not only demonstrates the feasibility applications of artificial intelligence in music creation but also provides a new technological approach to preserve and innovate the traditional Chinese music culture in the digital era. Furthermore, it explores new possibilities for the dissemination and innovation of Chinese cultural arts in the digital age.

MAICO: A Visualization Design Study on AI-Assisted Music Composition.

We contribute a design study on using visual analytics for AI-assisted music composition. The main result is the interface MAICO (Music AI Co-creativity), which allows composers and other music creators to interactively generate, explore, select, edit, and compare samples from generative music models. MAICO is based on the idea of visual parameter space analysis and supports the simultaneous analysis of hundreds of short samples of symbolic music from multiple models, displaying them in different metric- and similarity-based layouts. We developed and evaluated MAICO together with a professional composer who actively used it for five months to create, among other things, a composition for the Biennale Arte 2024in Venice, which was recorded by the Munich Symphonic Orchestra. We discuss our design choices and lessons learned from this endeavor to support Human-AI co-creativity with visual analytics.

Mimetic Possibilities: Collaboration through Movement in Multimedia Opera

Abstract In the opera The Ghost, artists with backgrounds in music composition, Greek tragedy, creative coding, and opera performance collaborated to create an artwork that engages with philosophy and technology. A light sculpture with a novel Internet of Things (IoT) architecture was adapted for the opera’s stage, where both the sculpture and its controlling interface were programmed to visually represent musical processes within the composer’s approach. During filming, the light sculpture contributed to the dramatic context through movement and gestural interaction between the sculpture and the soprano. Drawing upon human-computer interaction (HCI) research and its inquiry into movement for engaging with new technology expressively, movement and gesture transpiring from mimetic techniques in the opera were central in this interdisciplinary collaboration.

Development of Orchestration Course to Enhance Music Creative for Piano Major Student

This develops the orchestration course for piano major students and evaluates the orchestration course for piano major students. This study combines the principles of quasi-experimental and survey methods. Samples are 20 students majoring in piano at the School of Music Education of Xinghai Conservatory of Music. Research instruments are pretest and posttest, questionnaire, and interview. Data were analyzed by statistics and content analysis. The findings revealed that the orchestration course for piano major students consists of Music Composition Theory Foundation (α=0.9333), Instrument combination and part arrangement (α=0.898), Harmony processing and form structure (α=0.857), and Dynamics and effect processing (α=0.871), all of which are greater than 0.8, indicating that the pretest data has high reliability and the reliability of the research data is excellent and can be used for further analysis. The orchestration courses improve students' comprehensive musical abilities and promote their all-round development in many aspects. However, there are also some problems in the course, such as insufficient resources, limited technical support, and insufficient cooperation opportunities, which must be addressed and improved. In order to solve these problems and further optimize orchestration courses, it is recommended that musical instrument and orchestration resources be increased and more robust technical support provided. By continuously combining students' feedback to improve orchestration courses, we can better meet students' needs and enhance teaching effects and learning experiences. Such efforts will help elevate students' learning experience to a new level and achieve better teaching outcomes

The connoisseur’s emancipation of dissonance: how changing the experience of musical tension can inform individual responses to organizational paradoxes

ABSTRACT Although literature identifies a repertoire of constructive and defensive responses to paradox, the problem of overcoming dissonance remains. While a both-and response at the individual or collective level is desirable, individuals encountering paradoxes will likely respond defensively, separating tensions into either-or poles. Insight into how individuals can work with tension created by dissonance is limited. Musical composition may have something to share. Since the Renaissance, composers have exposed listeners to constructively responding to tension through increasingly dissonant music. Arnold Schönberg’s dodecaphonic (twelve-tone) system emancipates the constraints of dissonance, reducing listeners’ resistance and developing a connoisseur’s appreciation of structural incompatibilities.

중국 작곡가 시광남(施光南)의 예술가곡에 나타난 민족 음악적 특징

The Chinese composer Shi Guangnan's musical life, especially his art songs, contains profound musical characteristics of the Chinese nation. He has an important position in the development of modern Chinese art songs. From the content of this study, we find that Shi Guangnan's unafraid of difficulties and unyielding will and his love for the Chinese nation are the main driving forces for his success. In the aspect of music creation, he perfectly combined the folk songs of China's ethnic minorities with traditional music, and made flexible use of them in his art songs. At the same time, he also naturally combines artistic elements with popular elements. This is the main reason why his art songs are called “the music of the people”. So his art songs can probably be divided into two kinds of ethnic characteristics: The first is the specific ethnic music elements. This is called a “National characteristics in rhythmic patterns”. His art songs are characterized by the use of musical styles, culture, living habits of ethnic minorities, etc., By combining with Shi Guangnan's personal musical style. The second is the “Musical melodies with different national characteristics”, This model is the combination of the tone, rhythm and melody of southern China and northern China. “Musical melodies with different national characteristics” is to bind all the musical characteristics of the 56 ethnic groups in China in one framework，and he used western composition skills and the music characteristics of Chinese ethnic minorities to make the style of his musical works colorful. Finally, the perfect fusion of these two modes, he created the “Xi's pattern”. This pattern has a lot to do with his accumulation of his social experience and music experience. Because his art songs have beautiful melody, outstanding musical personality, and the characteristics of traditional folk music. Therefore, his art songs are used as vocal music teaching materials by Chinese higher art colleges and universities. His music works span from the 1950s to the early 21st century. He has integrated the musical composition methods and features of these periods. Therefore, his nationalized musical features have had a great influence on the musical creative thoughts and methods of other musicians. I hope this research can help people better understand and perform Shi Guangnan's art songs，and help the singers and students singing Shi Guangnan's art songs to better sing the feeling of the national cultural connotation and characteristics in the songs.