Chapter 4
Escalator
The concept for Escalator, for Yamaha Disklavier and field recording, is the discovery of music in everyday and often mechanical surroundings, such as an escalator in The Hague Central Station. In this transcription my method is filtering to zoom in on elements of the multi-layered and noisy field recording, and also to visualize the rhythm of the audio. This recording was also made in September 2014 (some days after the train recording) with a Zoom H4N hand-held recorder.
My first encounter with this instrument was in a concert for Yamaha Disklavier (and the Peter Ablinger automatic piano) at the Darmstadt Internationale Ferienkurse 2014. There I saw and heard a variety of uses for this automated instrument in combination with electronics, speaker feedback, harp pedals, etc. Intrigued by this memorable concert, I applied to an opportunity to work with the Disklavier owned by the Conlon Foundation in the Netherlands a couple of months later. After an introductory workshop on the basic mechanics, electronic connections, and example works by Robert van Heumen (Conlon Foundation board member), I was set free to begin exploring. At the time I had been transcribing the audio of the escalator recording into music notation, and thought that an automated instrument would be an ideal match. A Disklavier is a mechanical machine that plays and moves on its own, much like an escalator, and the piano keys have a visual resemblance to stairs. The foundation liked my idea and challenged me to compose the piece using its instrument. The transcription that I had made up to that point was mainly a rhythmic one, analyzing different layers, which I broke down into drone, ostinato, and pitched events such as OV-chipcards beeping. (These are the electronic check-in of the Dutch transportation system.) This recording also seemes to have a drone, although the layers are more rhythmic and semi-pitched than in Train. So this made it difficult to give to the ensemble of pitched instruments, and more applicable to the percussive Disklavier, which also creates mechanical sounds. Even more essential, the Disklavier offers an entertaining and meaningful visual layer that is very direct, so I explored this first. The opening Section A features pulsations of the left and right pedals on the beat (quarter and half notes). This highlights the groove of the escalator and creates mechanical pedal noises, squeaks and such. The next Section B features the addition of a silent ascent of the piano keys, like an escalator always traveling in one direction. The Disklavier can depress the keys without the hammer striking the strings; only small mechanical noises result. Thus, the audience can focus on hearing the field recording with a focus on the pulse. At this point the Disklavier visually translates what is happening in the recording and adds to the pulse in a blended way. Section B continues this combination with more waves of ascending keys and some OV-Chipcard noises, eventually having the piano sound. Over time, more and more scales fill the piano, and “random” or grouped notes begin to sound. Section C is the denouement – return to visuals only and as the tape fades out we hear solely the mechanics of the Disklavier.
The difficult part for me in composing the piece was designing the form with the tape as the passacaglia ground. Gradually introducing each element is rather predictable, but I think the way I set the 'machine' of the piece in motion, it took on this form. So I designed three phases to introduce the three major materials: 'silent' mechanical: pedal ostinato, silent rising scales, gradually sounding scales (from sporadic notes to fully sounding), and low drone. The piece also got a lot longer in this design, as we eventually hear notes resulting from the many scales set in motion. The first Train version of four minutes was too compressed in terms of the information I was giving the audience, so I wanted to take more time with this Escalator piece as well. Learning about timing and especially application and introduction of materials is a central part of this research. As I am very deeply familiar with the field recording it is difficult to relate to a listener's 'virgin ears' so I must remember to give more time to the audience. The first sketches of transcription was for the sameinstrumentation as Train, but it became clear that it would be difficult to execute in an ensemble because of timing.
The escalator recording has a regular pulse around 94 beats per minute with polyrhythms layered on top. The pulse is deceptive, though, as it is not square or as steady as it immediately seems. I had to make a click track with specific beat mapping in Logic Pro X to adjust for the slight irregularities to keep the transcription as synchronized as possible. (And it can always be improved upon.) In terms of pitch, this field recording has a harmonic drone field of B-C-E-flat to my ears, and surrounding incidental sounds relate to this mode, including OV-chipcard check-in beeps pitched with B natural's.
This is the first time in my transcription series that I edited the field recording through filtering. The train piece was a pure loop of the raw recording. The filtering in escalator focuses on bands of frequencies. Sometimes there is the original, at times there are only highs and lows, and some have a band-width filtering cross-fading back into the original unfiltered tape. One section includes a re-synthesis through SPEAR. How I did the audio manipulation for the tape part was intuitive as I aimed to allow space for the transcriptions, and to focus the listener's ear on details of certain registers or layers. I did keep the simple loop consistent in that the 90 seconds were not cut or spliced, merely filtered differently for the six iterations. (The exception is a small coda at the end.)