Statistical and Visual Analysis of Audio, Text, and Image Features for Multi-Modal Music Genre Recognition

We present a multi-modal genre recognition framework that considers the modalities audio, text, and image by features extracted from audio signals, album cover images, and lyrics of music tracks. In contrast to pure learning of features by a neural network as done in the related work, handcrafted features designed for a respective modality are also integrated, allowing for higher interpretability of created models and further theoretical analysis of the impact of individual features on genre prediction. Genre recognition is performed by binary classification of a music track with respect to each genre based on combinations of elementary features. For feature combination a two-level technique is used, which combines aggregation into fixed-length feature vectors with confidence-based fusion of classification results. Extensive experiments have been conducted for three classifier models (Naïve Bayes, Support Vector Machine, and Random Forest) and numerous feature combinations. The results are presented visually, with data reduction for improved perceptibility achieved by multi-objective analysis and restriction to non-dominated data. Feature- and classifier-related hypotheses are formulated based on the data, and their statistical significance is formally analyzed. The statistical analysis shows that the combination of two modalities almost always leads to a significant increase of performance and the combination of three modalities in several cases.

Three Problems in the Philosophy of Movie Music

This chapter examines three issues. First, the question of the nature of movie music. Second the issue of whether there is an implicit musical organizer with respect to the non-diegetic music track. And third, Peter Kivy’s hypothesis that the music track “fills-out” the otherwise “ghostly” visual track.

Approaching the Irreal: Realistic Sound Design in Andrei Tarkovsky’s Films

This chapter focuses on Tarkovsky’s sound design that underwent several transformations throughout his career. Whereas in his early features the director used to employ a regular music track composed by Viacheslav Ovchinnikov, in his later films made in in the 1970s he worked together with composer Eduard Artemiev who created original and highly experimental soundtracks. As the chapter demonstrates, Tarkovsky primarily strived for the realistic sound aesthetic, requesting a variety of noises and natural sounds, in conjunction with electronic arrangement. At the same time his sonic realism progressively tends towards irrealism in his later films when selected natural sounds manifest their otherworldly dimension. His preference for natural sounds sounding unnaturally is especially evident in his last film The Sacrifice (1986), when Tarkovsky decided to work with the sound mixer only, Owe Svensson, without a composer.

Sweet anticipation: Predictability of familiar music in autism

Autism has been characterised by different behavioural and cognitive profiles compared to typically developing (TD) individuals, and increasingly these differences have been associated with differences in structural and functional brain connectivity. It is currently unknown as to whether autistic and TD listeners process music in the same way: emotionally, mnemonically, and perceptually. The present study explores the brain’s dynamical landscape linked to music familiarity in an fMRI dataset from autistic and TD individuals. Group analysis using leading eigenvector dynamics analysis (LEiDA) revealed significantly higher probability of occurrence of a brain network in TD compared to autistic individuals during listening to familiar music. This network includes limbic and paralimbic areas (amygdala, hippocampus, parahippocampal gyrus, and temporal pole). No significant differences were found between autistic and TD individuals while listening to a scrambled, i.e. unfamiliar and more unpredictable, version of the same music track. These findings provide novel neuroimaging insights into how autistic prediction monitoring may shape brain networks during listening to familiar musical excerpts.

The Effect of music on maternal anxiety and maternal and fetal heart rate during cardiotocography

Introduction: It is well known that music acts as an analgesic. Facilitates muscle relaxation, reduction of physical pain and mental tension. Purpose of this original investigation: study the effect of music on maternal anxiety during cardiotocography as well as its effect on embryonic cardiac function in relation to mother’s heart rate. Method and materials: The survey, conducted in external midwifery medical services of Alexandra’s Hospital, involved 80 pregnant women who met the criteria for participation. They were divided in two groups. The 40 pregnant women were the Musical Group, and the other 40 were the Control Group. In both groups, the STAI scale was used. The control group underwent cardiotocography examination without listening to music. The Musical Group were selected to hear the music track ‘Kung Fu Piano: Cello Ascends’, a cover of the Piano Guys band. The hearing started 5 minutes before the end of the cardiotocography, with headset playing frequencies that are within the frequency spectrum of the music track and special music player, and 10 minutes before the end of the cardiotocography the pulse oximeter Beurer P080 was placed. In this music track we did music and frequency analysis with the following programs: SPAN of Voxeno and Reaper of Cocos, as well as Theory-Harmony of Music, to see if at the time certain changes in the track occur, there are corresponding changes in the heart rate of the mother and the fetus. Results: In our study we observed that music significantly decreased the level of anxiety of women subjected to non-stress test (NST) (Median anxiety score prior to the conduct of the non-stress test 53.38 (49-57) vs 25.20 (23-28) following the completion of the test. Moreover differences among women that heard music were significant compared to those that did not (25.20 (21-28) vs 56 (48-64)) despite the fact that baseline differences among the two groups were comparable (54.45 (59-67) vs 50.80 (53-58). Finally, following performance of music analysis we observed significant variations in the baseline heartbeat of pregnant women as well as in the cardiotocographic analysis of fetuses (number of accelerations, baseline rhythm); those patterns were directly related to musical characteristics of the track that women listened to. Conclusions: Music has a positive effect on pregnancy. It is a non-invasive way of anxiety relief, as well as a simple, non-time-consuming way of improving cardiotocography among low risk cases; thus, potentially diminishing false-positive results which may result in unnecessary deliveries.

STRATEGI KREATIF PRODUSER DALAM MEMPERTAHANKAN EKSISTENSI PROGRAM MISSION X TRANSTV

Mission X merupakan salah satu program game show yang ditayangkan oleh TRANS TV, tantangan bagi Produser yaitu penonton atau masyarakat yang kian hari semakin cerdas dalam memilih sebuah tayangan. Hal ini menarik peneliti untuk mengetahui bagaimana strategi kreatif produser dalam mempertahankan eksistensi program Mission X pada stasiun televisi TRANS TV. Tujuan penelitian ini untuk mengetahui bagaimana strategi kreatif produser dalam mempertahankan eksistensi program Mission X pada stasiun televisi TRANS TV. Metode yang digunakan dalam penelitian ini adalah metode deskriptif kualitatif dengan teknik pengumpulan data melalui observasi, wawancara, dan penelusuran dokumen dan diakhiri dengan analisis. Hasil penelitian yang diperoleh yaitu bagi Produser dan tim, komunikasi merupakan hal yang mempengaruhi keberhasilan dalam tim. Seorang Produser dalam menentukan strategi kreatif program Mission X berlandaskan pada elemen-elemen sebagai berikut: target penonton, bahasa naskah, format acara, punching line, gimmick and funfare, clip hanger, time and bumper, penata artistik, music and fashion, ritme dan birama acara, andago dan music track untuk ID tune, general rehearsel (GR), dan interactive program. Eksistensi program Mission X dapat bertahan dikarenakan adanya strategi kreatif yang selalu berinovasi dan komunikasi yang sudah terjalin dengan baik.

Musiikki ja kerronnan tasot Kirill Serebrennikovin elokuvassa Kesä

Venäläisen ohjaajan Kirill Serebrennikovin elokuva Kesä (Leto, Venäjä 2018) on tositapahtumiin perustuva, mutta silti voimakkaan fiktiivinen kuvaus Neuvostoliiton underground-rockin noususta valtavirtaan 1980-luvun alkupuolella, päähenkilöinä kaksi neuvostorockin tosielämän suurnimeä Majk Naumenko (1955–1991) ja Viktor Tsoi (1962–1990). Elokuvassa kuullaan luonnollisesti paljon musiikkia: venäläisiä ja länsimaisia 1980-luvun alun rock-hittejä sekä alkuperäisinä että uusina tulkintoina, mutta myös elokuvaa varten sävellettyä originaalimusiikkia. Kesä voitti parhaan soundtrackin palkinnon Cannesissa 2018, mutta sai Venäjällä ristiriitaisen vastaanoton.Artikkelissa keskitytään tarkastelemaan Kesän musiikinkäyttöä formalistisen lähiluvun ja rakenneanalyysin keinoin. Tarkastelun keskiössä on erityisesti elokuvamusiikin narratologia ja Claudia Gorbmanin (1987), Rick Altmanin (1987) ja Guido Heldtin (2013) määritelmät diegeettisyydestä, ei-diegeettisyydestä, supradiegeettisyydestä sekä ekstrafiktiivisyydestä. Tutkimuskysymyksenä on se, miten Kesä-elokuva käyttää neuvostoajan rockia elokuvamusiikkina ja miten tämä musiikinkäyttö ankkuroituu neuvostoelokuvien rock-musiikkikonventioihin.Kesän musiikkistrategiassa muodostuu selkeä vastakkainasettelu kotimaisen ja ulkomaisen, elävän ja nauhoitetun, sekä alkuperäisen ja uudelleentulkitun musiikin välille. Eri musiikkityylien ja kerronnan tasojen välille muodostuu elokuvassa tietynlainen säännönmukaisuus ja tehtävänjako. Se miltä kerronnan tasolta musiikin katsotaan milloinkin tulevan vaikuttaa vahvasti siihen, miten realistisena minkäkin kohtauksen voi lukea, ja tämä puolestaan sitoo elokuvan eri genreperinteisiin: realistisempi, diegeettinen musiikkiesitys ankkuroi elokuvan elämäkertaelokuvien genreen, kun taas täysin epärealistiset, supradiegeettiset musiikkiesitykset viittaavat enemmän musikaaliperinteeseen. Music and Levels of Narration in Kirill Serebrennikov’s film LetoRussian director Kirill Serebrennikov’s Leto is a film about the rise of the underground rock scene into the mainstream in early 1980s Soviet Union. The film is based on a true story and the lives of two rock legends Majk Naumenko (1955–1991) and Viktor Tsoi (1962–1990), even though it also contains plenty of highly fictitious elements. The music track features Russian and Western rock songs of the era both as original performances as well as cover versions, additionally there are also excerpts of original score. Leto won the best soundtrack award in Cannes in 2018, but it received mixed reviews in Russia.This article analyses the use of music in Leto through formalist close reading and structural analysis. The analysis relies heavily on film music narratology and in particular on Claudia Gorbman’s (1987), Rick Altman’s (1987), and Guido Heldt’s (2013) definitions of diegeticity, non-diegeticity, supradiegeticity, and extrafictivity. The research question concerns how is Soviet era rock used as film music in Leto, and how does this use relate to the rock music conventions in Soviet cinema.In the musical strategy of Leto, there arises a juxtaposition between domestic and foreign music, as well as live and recorded, and original and re-interpreted music. Furthermore, there is a structure and logic in the way the different types of music relate to the levels of narration throughout the film. The narrative level that the music is anchored to has an effect on the interpretation of individual scenes and events as realistic or unrealistic. This also anchors the film in different film genres or traditions: a more realistic, diegetic music performance is connected with traditional biopic dramas, whereas the unrealistic, supradiegetic musical performances are more closely connected with the tradition of Hollywood musicals.

Does the Tempo of Music Impact Human Behavior Behind the Wheel?

Objective: Assess the influence of background music tempo on driving performance. Background: Music with a fast tempo is known to increase the level of arousal, whereas the reverse is observed for slow music. The relationship between driving performance and level of arousal was expected to take the form of an inverted U-curve. Method: Three experiments were undertaken to manipulate the musical background during driving. In Experiment 1, the driver’s preferred music track played at its original and modified (plus or minus 30%) tempo were used together with the simple ticking of a metronome. In Experiment 2, music tracks of different tempos were played during driving. In Experiment 3, music tracks were categorized as arousing or relaxing based on the associated perceived level of arousal. Results: Listening to music tended to influence drivers’ performances in a car-following task by improving coherence and gain adjustments relative to the followed vehicle but simultaneously shortened the intervehicular time. Although the tempo of the music per se did not directly affect driving behavior, arousing music tracks improved drivers’ adjustments to the followed vehicle (Experiment 3). Conclusion: The tempo of the music listened to behind the wheel was not found to influence driving behaviors. However, arousing music improved drivers’ responsiveness to changes in the speed of the followed vehicle. However, this benefit was canceled out by a reduction in the drivers’ intervehicle safety margin. Application: Listening to arousing music while driving cannot be considered to improve road safety, at least in a car-following task without attentional impairments.

Striking the Right Chord: A Theoretical Approach to Balancing Artists' Intellectual Property Rights on Remix Audio-Sharing Platforms

102 Cornell Law Review 211 (2016)If you search for the popular electronic dance music track “Turn Down for What” on SoundCloud, the online audio-sharing platform, over 500 tracks will surface on your computer screen. Most of these tracks are unofficial remixes from a variety of music genres, everything ranging from rap to reggae, blues to bhangra. Even though many of these unofficial remixes are available as free downloads, they were never authorized by the original artists, their publishers, or their record labels. Instead, these tracks are homegrown creations, born out of the basements of aspiring, yet amateur, remixers. SoundCloud, the Swedish brainchild of two entrepreneurs who launched the site from a nightclub dance floor in 2008, prides itself on allowing anyone to create, upload, share, and download music. The platform’s self-proclaimed mantra celebrates the fact that “anyone can create sounds and share them everywhere.” Their mission resonated with the masses; SoundCloud boasts approximately 175 million monthly listeners—a base which more than doubles the size of Pandora’s active monthly users and roughly doubles the size of Spotify’s free user base.

The Effects of Diegetic and Nondiegetic Music on Viewers’ Interpretations of a Film Scene

Previous studies have shown that pairing a film excerpt with different musical soundtracks can change the audience’s interpretation of the scene. This study examined the effects of mixing the same piece of music at different levels of loudness in a film soundtrack to suggest diegeticmusic (“source music,” presented as if arising from within the fictional world of the film characters) or to suggest nondiegetic music (a “dramatic score” accompanying the scene but not originating from within the fictional world). Adjusting the level of loudness significantly altered viewers’ perceptions of many elements that are fundamental to the storyline, including inferences about the relationship, intentions, and emotions of the film characters, their romantic interest toward each other, and the overall perceived tension of the scene. Surprisingly, varying the loudness (and resulting timbre) of the same piece of music produced greater differences in viewers’ interpretations of the film scene and characters than switching to a different music track. This finding is of theoretical and practical interest as changes in loudness and timbre are among the primary post-production modifications sound editors make to differentiate “source music” from “dramatic score” in motion pictures, and the effects on viewers have rarely been empirically investigated.