“Seeing Sonic Styles” focuses on developing comparative visuals that
show the playing styles of classical music pianists. This research
examines and compares duo performances of Chopin Ballade No. 1 to
identify the playing styles of the pianists. Chopin Ballade No.1 is a
classical music piece composed by Chopin and is one of the most
popular pieces by him. The intention is to bring a comparative aspect
into music visualization. By analyzing the differences between piano
play styles, it will benefit the audience to identified various
interpretations among pianist and raise the question:
What’s the standard / correct interpretation of this piece?
This study will retrieve pianists’ live recordings using music
information retrieval, which is then analyzed as a sound spectrum. The
comparative visualization formed to describe the inner and external
differences between pianists using sound data.
The intended audience for this piece is music students and
professionals since this analysis of comparative performances will
help them to recognize the various approaches taken by famous pianists
to the same piece.
I personally have a great interest in the classical music genre, as I
have played piano myself. When I listen to different performances by
various pianists, I can identify the pianist by only using my ear, but
I cannot visually see the differences or talk clearly about them
without additional evidence. The complicated nature behind pianists'
playing styles is that they employ two layers of musical thoughts. The
first layer of thought is from the composer. The composer builds the
piece of music to reflect personal interests, cultural ideas or even
social issues. The performers study the composer’s intention,
notation, instructions and purpose to construct their own
interpretations and naturally express those in personal styles. In
this respect, the musical message they convey not only has unique
perspectives of the composer, sometimes from long ago, but also from
the moment of performance by a particular pianist. These are merged in
each performance to convey a sound that reflects all the influences on
the music. Through the dynamic markings on a manuscript or score and
the liberty and creativity of each pianist’s playing they can even
resonate with an audience hundreds of years after the original
composition was created.
Lang Lang’s Climax Part in Ballade No.1
Zimerman’s Climax Part in Ballade No.1
These visualizations were created using TouchDesigner. The same audio
file is used as the input, and the same FFT method is used to analyze
the sound. When the audio is being played in parallel, each sound
fragment is plotted to rotate. The keyboard's outer circles are the
lower portions of the keyboards, while the inner circles indicate the
higher frequency and higher notes on the keyboard. The whiteness of
the hue represents intensity; a stronger whiteness in color indicates
a more powerful attack on the keyboard. When comparing this to
librosa's earlier research, the general analysis is the same but
presented in a different format. In this distinct style, the material
has been condensed and the focus is clearer due to the use of white
visuals against a black background, which also makes for a more
succinct illustration idea. For its dynamic and lively visual formats,
it is a more appealing format for viewers to easily perceive how the
sound is evolving, it also enables audiences who have no prior
experience with music to experience the music. The animation also will
create an atmosphere to narrate emotions.
The final design prototype for this study incorporates a more complete
pianists’ performance to raise the question:
Is there a standard interpretation on a music piece?
It is apparent that piano pieces composed and performed over a century
ago did not sound the same as with these modern pianists for reasons
of factors like piano type and structure explained in the previous
piano section.
Here shown in the figure is the static version of the comparative
animated visualization. Each circle is drawn in the same time span and
the size and color refers to the amplitude and timbre at that specific
time. The timbre or the brightness of sound is calculated by finding
the centroid of frequency as the weighted mean of the frequencies
present in the signal, determined using a Fourier transform, with
their magnitudes as the weights.