Content uploaded by Nick Bryan-Kinns
Author content
All content in this area was uploaded by Nick Bryan-Kinns on Jun 19, 2020
Content may be subject to copyright.
On Digital Platforms and AI for Music in the UK and China
Nick Bryan-Kinns
Queen Mary University of
London
London, UK
n.bryan-kinns@qmul.ac.uk
Zijin Li
China Conservatory of Music
Anxiang Rd. Chaoyang
Beijing, China
lzj@ccmusic.edu.cn
Xiaohua Sun
Tongji University
Fuxin Road 281, Yangpu
Shanghai, China
xsun@tongji.edu.cn
ABSTRACT
Digital technologies play a fundamental role in New Inter-
faces for Musical Expression as well as music making and
consumption more widely. This paper reports on two work-
shops with music professionals and researchers who under-
took an initial exploration of the differences between digital
platforms (software and online services) for music in the
UK and China. Differences were found in primary target
user groups of digital platforms in the UK and China as
well as the stages of the culture creation cycle they were
developed for. Reasons for the divergence of digital plat-
forms include differences in culture, regulation, and infras-
tructure, as well as the inherent Western bias of software for
music making such as Digital Audio Workstations. Using
AI to bridge between Western and Chinese music traditions
is suggested as an opportunity to address aspects of the di-
vergent landscape of digital platforms for music inside and
outside China.
Author Keywords
Digital Platforms, AI, Music, UK, China, Cross-Cultural
CCS Concepts
•Applied computing →Sound and music computing;
Performing arts; •General and reference →Surveys and
overviews;
1. INTRODUCTION
Digital technologies play a fundamental role in New Inter-
faces for Musical Expression as well as music making and
consumption more widely. The unique cultural heritages,
economic, social, and political profiles, and divergent dig-
ital infrastructure and regulations around the world may
have an effect on how these digital technologies are created
and used.
In this paper we report on the differing landscape of dig-
ital platforms (software and online services and reposito-
ries) for music in the UK and China. The UK and China
make compelling comparative case studies in this area as
the UK is a leader in Music Industries and one of the few
countries which is a net exporter [1], yet there is very little
understanding of, or access to, the Chinese music market
by UK music-technology companies. The Chinese sector is
Licensed under a Creative Commons Attribution
4.0 International License (CC BY 4.0). Copyright
remains with the author(s).
NIME’20, July 21-25, 2020, Royal Birmingham Conservatoire,
Birmingham City University, Birmingham, United Kingdom.
growing rapidly, and the Chinese government is encouraging
the promotion of Chinese music performance and apprecia-
tion [7] including moves to popularize Chinese music in pri-
mary and secondary schools. Furthermore, the market for
music consumption is large in China and rapidly expand-
ing (in 2017, China’s digital music industry reached RMB
58bn with 523m digital music users, up 9.6 percent year-
on-year [12]). Moreover, research and development in digi-
tal music is well established in Europe and North America
with international conferences such as the Sound and Music
Computing conference being established in France in 2004,
whilst it is in early stages in China e.g. the first China
Sound and Music Computing Workshop was launched in
2013 by Fudan University and Tsinghua University1, now
the China Sound and Music Technology Conference. Sim-
ilarly, research on digital music is notably advanced in the
UK compared to China, for example using AI throughout
the creative pipeline from generative music composition e.g.
[4] and musical assistants e.g. Amper2, to automated mix-
ing and production e.g. LANDR3. Yet, at this time, Chinese
companies such as Tencent and Alibaba are investing heav-
ily in both AI and digital music in much the same way as
companies such as the BBC and Google are doing in Europe
and North America.
More broadly, there are deep rooted differences in con-
temporary culture and heritage between China and the UK
which would affect conventions and methods of music pro-
duction and consumption and digital platforms to support
them. For example, social media such as Twitter4and Face-
book5are ubiquitous in the UK in contrast to Chinese social
media platforms such as QQ6and WeChat7and there are
significant differences between the use of Twitter (outside
China) and the Chinese use of the somewhat equivalent Sina
Weibo8[11] which may also be reflected in the how music
is consumed through digital platforms.
Finally, comparing the context and digital platforms for
music in the UK and China sheds light on differences be-
tween China and other regions such as North America and
Europe more broadly given the regulatory, cultural, and
technological similarities between the UK and other Euro-
pean countries as well as North America. There are cur-
rently no surveys which explore the differences between dig-
ital platforms for music in the UK and China, nor between
China and other countries more widely.
1http://www.csmcw-csmt.cn/
2https://www.ampermusic.com/
3https://www.landr.com/
4https://twitter.com/
5https://www.facebook.com
6https://www.qq.com
7https://www.wechat.com/en
8http://weibo.com
Table 1: Workshop Participant Demographics
Shanghai London
Chinese 35 4
UK 9 12
Academia 18 11
Industry 22 4
Artists 4 1
Female 14 3
Male 30 13
2. METHOD AND RESULTS
Two three-day workshops on AI for Music9were held in
2019 including explorations of the landscape of digital plat-
forms for music in the UK and China and building networks
with participants from academia, music industry, and per-
forming artists. Participants were recruited through email
and social media networks and their participation in the
workshop funded by government research agencies in the
UK and China. The first workshop took place in Shanghai
(44 participants) hosted by the College of Design and Inno-
vation, Tongji University, China, the second took place in
London (16 participants) hosted by the Centre for Digital
Music, Queen Mary University of London, UK, see Table 1
for participant demographics. Both workshops were struc-
tured as two days of face-to-face workshop activities fol-
lowed by one day of visits to local digital music technology
centres. Workshops were facilitated by researchers from the
UK and China. Workshop structure and pace was designed
and maintained by the lead author with each day structured
into four group activity sessions of 90 minutes with group
report back twice a day. Data from workshops was collected
through written and oral summaries from each group activ-
ity and documentation of brainstorming activities such as
post-it note grouping.
Prior to the first workshop in Shanghai participants were
asked to complete an online survey to collect their thoughts
and opinions on current topics and trends on digital plat-
forms for music in China and the UK to inform the work-
shop discussion topics, particularly focusing on AI for music
as a key contemporary concern. This activity produced 33
topics of interest. Participants then grouped topics into
themes using an online tool10 resulting in five themes which
were used to as prompts in the workshops: AI for compo-
sition and music generation; Production workflow; Cata-
logue management and recommendation systems; IP man-
agement; Cross-cultural challenges and opportunities.
Two key areas of comparison between the UK and China
explored in the workshops are reported in this paper: i)
broad comparisons between the context of music making
and consumption; ii) comparisons of use of digital platforms
for music creation, production, and consumption.
2.1 Context of Music Making
On the first day of the workshops participants developed a
high level comparison between music making and consump-
tion in the UK and China by brainstorming key differences
between China and the UK’s music sector broadly split into
groups based on the themes identified prior to the work-
shop. Participants worked in four groups, each examining
one theme and the cross-cultural similarities and differences.
At the end of the day participants reported back their find-
ings. On the second day of the workshops participants un-
dertook a thematic grouping using post-it notes and large
9http://ai4music.eecs.qmul.ac.uk
10http://well-sorted.org
whiteboards to group their findings into key topics which
are illustrated in Table 2.
In group discussions and reflection at the end of the work-
shops participants identified that these differences lead to
barriers to: the adaptation of Western content for Chinese
audiences; the consumption of Chinese content by Western
audiences; and the production of traditional and experimen-
tal Chinese music using existing digital music workflows.
These barriers are largely due to the inherent bias of con-
temporary music workflows, and also practices, norms, and
technologies that favour Western musical traditions. This
makes it difficult to, for example: access and operate in the
music sector of the Chinese creative industries from outside
China; access, understand, and exploit Chinese music and
musical instrument collections; and to treasure, embrace,
and reimagine Chinese intangible cultural heritage in the
global digital age.
Moreover, participants identified that the level of techno-
logical and design innovation in the music industry in China
is low compared to the UK, focusing on consumer consump-
tion such as streaming services, rather than the design of
new products, services, and music production workflows.
This also reduces the contribution of the digital music sec-
tor in China to other Creative Economy sectors such as film
making and gaming, for example there are fewer Chinese
systems for generative music in computer games.
2.2 Landscape of Digital Platforms for Music
On the second day of workshops, participants identified 104
digital platforms used in the UK and China for music mak-
ing and consumption based on their own use and knowledge
of such platforms. Two key differences between China and
the UK were found in this review: i) the intended user
of the software - whether amateur, semi-professional, or
professionals; and ii) which stage(s) of the UNESCO cul-
ture cycle [9] of creation, production, dissemination, exhibi-
tion/ reception/ transmission, consumption/ participation
the digital platform was used. The culture cycle was used
as a framework to structure the analysis of the use of dig-
ital platforms given the wide range of platforms identified.
These differences are summarised in Figure 1 which shows
the proportion of kinds of intended user (amateur, semi-pro,
pro) in each of the culture stages for China and the UK with
key differences summarised in Table 3 for ease of reading.
Classification of which platforms were used in which stages
of the culture cycle was undertaken by groups who reported
back their assessment and merged their classification with
other groups until a mutually agreeable and stable classifi-
cation was established as reported in Figure 1. It should be
noted that the second day in both workshops included net-
working and collaborative project ideation which for brevity
is not reported in this paper.
Participants found that aside from core professional mu-
sic software such as Logic Pro11 and Pro Tools12, which as
noted above are inherently biased toward Western musical
traditions and norms, the digital platforms used for music
making and consumption in the UK and China were largely
divergent. For example spotify13 is a digital platform for
streaming music used extensively in the UK, but it is not
used in China. An equivalent streaming platform in China
is NetEase Cloud14 which is not used in the UK despite
being available for download there. No doubt much of this
divergence is due to licensing and copyright differences be-
11https://www.apple.com/uk/logic-pro/
12https://www.avid.com/pro-tools
13https://www.spotify.com/uk/
14https://music.163.com
Table 2: Outline comparison between music context in the UK and China
Topic UK China
Automatic mixing Predominantly used in professional production
systems Predominantly used in karaoke apps
Digital Audio Worksta-
tions
Tailored to Western music production and
styles Lack of Chinese tools for audio production
Audio libraries Extensive, high quality libraries of audio from
Western instruments
Lack of high quality an accessible recordings
of Chinese instruments
Copyright infringement
detection Advanced audio fingerprinting technologies Advanced audio fingerprinting technologies
Royalty payments (e.g.
for streaming services) Per track payments Per label payments
Music performance the-
ory and practice Well documented Learnt by praxis
Tools for music educa-
tion
Extensive tools for learning and practicing in
Western musical styles
Lack of tools for learning Chinese musical in-
struments and performance style
Audio consumption pref-
erence HiFi systems Mobile phone
Table 3: Comparison of music software produced in
the UK and China
Stage China vs. UK
Creation China: more emphasis on amateur;
UK more focus on semi-pro
Production China: more focus on amateur;
UK: more focus on professionals
Dissemination China: more focus on semi-pro;
UK: more focus on amateur
Consumption China: more focus on semi-pro;
UK: more focus amateur
0%
25%
50%
75%
100%
Creation (UK)
Creation (CN)
Production (UK )
Production (CN)
Dissemination (UK)
Dissemination (CN)
Consumption (UK)
Consumption (CN)
Amateur Semi-pro Professional
Figure 1: Landscape of Music Software produced in
the UK and China
tween the UK and China. One notable exception that may
cross over from China to the UK is TikTok15 which was
initially launched in China and has since become popular
across the globe.
In China, there is a large music consumption market
given the large population, which may explain why the
biggest three music companies in China (Tencent, Alibaba
and NetEase) primarily focus on playing songs on their
digital platforms rather than supporting semi-professionals
and processionals in music making and production. For in-
stance, the Chinese company Tencent Music recently claimed
to have 800 million users in 2019, much the same as the
Chinese company NetEase Cloud. In contrast, independent
musician users on these platforms are only 100,00016, ap-
proximately just 0.001 percent of the users.
3. REFLECTIONS
Through two workshops with professionals and researchers
in the music sectors of China and the UK we found that
both the context of music making and consumption, and
the digital platforms used were divergent between the UK
and China. We found more focus on amateur music cre-
ation in Chinese digital platforms (e.g. pitch correction for
karaoke) compared to the UK, and more emphasis on pro-
fessional music production technologies in the UK. Two key
reasons for these differences were identified: i) the inherent
bias of professional music production platforms to Western
musical styles, and ii) the preference of mobile phones as
the platform of choice for music consumption in China.
3.1 Opportunities: AI for Music
Current trends in AI research were identified by participants
as possible solutions to some of the barriers to the use of con-
temporary music production tools for Chinese music making
and production. It should be noted that the identification of
these opportunities comes in part from participants’ knowl-
edge, experience, and interest in AI for music. The work-
shops highlighted that the rapid growth and acceptance of
AI technologies in music making and consumption offer op-
portunities to explore support for the creation, access, and
consumption of large data sets of Chinese and Western mu-
sic in radical yet culturally sensitive and responsive ways.
Firstly, AI may help to reduce the imbalance between mu-
sic production and consumption between China and the UK
15https://www.tiktok.com/en/
16http://www.chinambn.com/show-6123.html
where most professional music tools have inherent Western
biases. For example, participants noted that a practical
way to address this may be to use AI-based style-transfer
techniques (cf. [3]) to support cross-cultural content cre-
ation, production and consumption, allowing us to bridge
the gap between contemporary music technologies and Chi-
nese music traditions. Indeed, questions of how to create
and tailor creative content across cultures are currently hot
topics in both current academic research (e.g. [2]) and in-
dustrial research and design. This may bring new life to
Chinese traditional music through new technology, and fos-
ter greater access to Chinese music both in the UK and in
China, especially for younger generations.
Secondly, workshop participants identified that we may
be able use adaptive music techniques to reduce the barri-
ers to creating Chinese music with contemporary music pro-
duction tools. Again, this would help to open professional
Chinese music to the wider use of digital music platforms
going someway to redressing the imbalance of professional
music production in China. Adaptive music techniques also
offer the opportunity to create new products and services
at the intersection of Chinese and Western music. For ex-
ample, variPlay [8] was discussed in the workshop as a UK
produced adaptive-music production tool and player which
could be refined to support cross-cultural music production
and consumption. Similarly, Composer4Everyone [6], a Chi-
nese social media app that uses AI based rearrangement
to automatically generate music in Western styles of pop,
electronic, and classical from an audio snippet is an early
exemplar of the potential of AI supporting the intersection
of Chinese and Western music.
Thirdly, the workshops identified opportunities for us-
ing machine learning techniques to facilitate greater ac-
cess to Chinese music datasets which were noted to be
largely inaccessible outside China due to language and cul-
tural barriers. This again would help with incorporating
richer and more nuanced Chinese musical element in pro-
fessional music production tools and at the same time pro-
vide opportunities for Western music producers to access
existing Chinese datasets. For example, some workshop
participants are already exploring musical instrument play-
ing technique detection based on Fully Convolutional Net-
works(FCN) and applying this to analysing Chinese Bowed-
Stringed Instruments in the Dataset of the Chinese Music
Instrument(DCMI) [10], starting with the Erhu, a popu-
lar Chinese bowed-stringed instrument. Based on the same
DCMI, researchers are undertaking analysis and modeling
of timbre features of the sound of Chinese musical instru-
ments using support vector machines(SVM) [5], both of
which have the potential to open these datasets to researchers
and musicians outside China.
Finally, it was noted in the workshops that the largest
Chinese technology companies have already realized that
they need to be developing a whole digital music ecology
to retain their market advantage and rebalance the Chinese
music production ecosystem e.g. addressing the fact that
there are no Digital Audio Workstations made in China.
This offers an opportunity for global collaboration on devel-
opment of Digital Audio Workstations (DAWs) and DAW
plugins designed for the Chinese music making market.
4. SUMMARY
In workshops of music researchers and practitioners held in
the UK and China we found differences between the context,
use, and landscape of digital platforms for music in China
and the UK. There are opportunities for using AI to bridge
across Chinese and Western musical traditions and digital
platforms. There are also opportunities for developing ded-
icated software and plugins for Chinese music makers which
would increase the access to, and consumption of, Chinese
music in China and across the world.
5. ACKNOWLEDGMENTS
Many thanks to Jan Dornig, Gyorgy Fazekas, Johan Pauwels,
and to all the participants of the workshops including Queen
Mary University of London, Tongji University, ACRCloud,
BaroxTech, ByteDance, CBI China Bridge, Central Conser-
vatory of Music, China Conservatory of Music, Crust Music,
Dogma Studio, Hangzhou Alibaba Music Technology Co.,
Ltd., NetEase Cloud, New York University Shanghai, 1618
Digital, PingAn Technology, QQ Music and WeSing at Ten-
cent Music Entertainment Group, Shanghai Conservatory
of Music, Sunny Media, Tido Music, the University of Not-
tingham, the University of West London, Xiami Music of
Alibaba Group, Zhejiang University, Zera Culture Group,
and major European digital music software companies who
prefer to remain anonymous.
This research was funded by AHRC grant AH/T001259/1
and Chinese funding sources.
6. REFERENCES
[1] H. Bakhshi, I. Hargreaves, and J. Mateos-Garcia. A
manifesto for the creative economy. NESTA, 2013.
[2] W. Benetos, S. Dixon, A. Duan, and S. Ewert.
Automatic music transcription: An overview. IEEE
Signal Processing Magazine, 36(1):20–30, 2019.
[3] S. Dai, Z. Zhang, and G. Xia. Music style transfer
issues: A position paper. CoRR, abs/1803.06841,
2018.
[4] G. Hadjeres, F. Pachet, and F. Nielsen. Deepbach:
asteerable model for bach chorales generation. In
Proceedings of the 34th International Conference on
Machine Learning, pages 1362–1371, 2017.
[5] W. Jiang, J. Liu, Z. Li, J. Zhu, X. Zhang, and
S. Wang. Analysis and modeling of timbre perception
features of chinese musical instruments. In Proceedings
of 18th International Conference on Computer and
Information Science (ICIS), pages 191–195, 2019.
[6] A. Liu, J. Wang, J. Peng, Y. Wang, Y. Mei, X. Liang,
Z. Xia, and J. Xiao. Composer4everyone: Automatic
music generation with audio motif. In Proceedings of
Conference on Multimedia Information Processing
and Retrieval (MIPR), pages 502–503, 2019.
[7] MoE. Ministry of education of the people’s republic of
china. www. moe. gov. cn/ srcsite/ A17/ moe_ 794/
moe_ 628/ 201906/ t20190614_ 385818. html , 2019.
[8] J. Paterson, R. Toulson, and R. Hepworth-Sawyer.
User-influenced/machine-controlled playback: The
variplay music app format for interactive recorded
music. Arts, 8(112), 2019.
[9] UNESCO-UIS. Unesco framework for cultural
statistics. ISBN 978-92-9189-075-0, 2009.
[10] Z. Wang, J. Li, X. Chen, Z. Li, and B. Han. Musical
instrument playing technique detection based on fcn:
Using chinese bowed-stringed instrument as an
example. arXiv:1910.09021, 2019.
[11] L. Yu, S. Asur, and B. A. Huberman. What trends in
chinese social media. In Proceedings of the
5thSNA-KDD Workshop’11 (SNA-KDD’11), San
DiegoCA USA, August 2011.
[12] Z.Zhao. China music industry development report.
baijiahao. baidu. com/ s? id=
1620435994855362418& wfr= spider& for= pc , 2018.
