DJs and musical shaping

A DJ's act of performance has both similarities with and differences from that of a classical performer. A classical performer traditionally begins the performance preparation process by studying a score, though findings by Ginsborg and Prior (2011) suggest that many younger musicians use recordings as a starting point for learning a new piece. Whichever approach is adopted, the performance materials are specified by someone else, whether that person is a composer or another performer. Many of a DJ's materials are specified by others in as much as they are produced by others (unless the DJ is also a producer working with his or her own materials), but there is far greater freedom of choice in the performance itself as the DJ can decide which, and in what order to play music, and the audience is largely unaware of the selections they are going to make. DJs take a recording that is, in itself, a finished product, and transform it into a new performance. Thus whilst classical performing musicians and DJs are often realizing another musician's conception of a piece of music, this may vary from an exact reproduction of that piece to a radically different performance that may even be considered to be a new piece.

The cultural worlds of classical performers and DJs are very different, as are the ways each type of musician is trained. While a classical performer will usually receive performance training from teachers when they are younger, at school and then at a conservatoire or a university, it is more common for a DJ to learn his or her performance skills informally. 3

These similarities and differences between classical musicians and DJs inspired the current research, which is designed to assess the extent to which DJs use notions of musical shaping in their performances. The questionnaire study mentioned above (Prior, 2012c) received a single response from a practising DJ, whose responses to the more open-ended questions provided considerable food for thought. This DJ was female, aged between 25 and 34 years, and was born and lives in the UK. She completed the questionnaire on the 30 March 2010. She studied music at two UK Universities as an undergraduate and a postgraduate respectively. She described her main instrument as 'turntables (DJing)' but also reported playing the piano and violin. She had been playing the turntables for less than ten years, and indicated her musical status as being within the category of 'An amateur, intermediate-level performer (none, or a very low proportion of your income comes from performing activities; you are a capable player, but not of professional standards)'. She reported playing music that was within the categories of 'Unaccompanied solo music' and of 'Popular/non-classical music, Dance/electronic (house, techno, electronica etc.)', but also had interests in record production and musicology. Her responses to questions concerning her experience of music and shape are shown in Table 1.

She described a solo performance (making a mixtape) and a 'back-to-back' DJ set (in a nightclub) and both experiences had occurred in the four months prior to her completing the questionnaire. It is interesting that both episodes involved planning a performance and the choice of tracks involved, rather than the act of performance itself. This might be likened to a classical performer choosing a programme for a concert. Interestingly, this participant related the overall shaping of a set (on both the occasions described) to the energy level of the tracks played. There is a desire to consider the overall contour of a mix and a desire to maintain interest for the listener or audience (in the case of the nightclub performance). This too is a useful concept for classical musicians, who in the interview study described above (Prior, 2012b) often discussed the overall contour of a musical work with several movements, and gestured to indicate the energy levels implied within those movements. The DJ participant also discusses the manner in which she could make a transition between tracks as part of the musical shaping, an idea that emphasizes her role as a performer in creating a desirable set. Although this participant's original emphasis was on the planning of sets, this suggests that musical shape is also considered during the act of performance. Like the classical performers in the questionnaire study and interview study, this DJ uses a range of metaphors to describe the musical shape. When asked later about words with a similar meaning to 'shape' in a musical context, she suggested contour, mood, energy level and pattern.

Table 1. Questionnaire data concerning a DJ's experiences of music and shape

	Do you ever think about shape when THINKING about how to perform music?	Do you ever think about shape when TALKING with others about how to perform music?
Response	Yes	Yes
When this occurred (i.e. date)	December 2009	March 2010
The piece of music	A DJ mix I was preparing	Various dance tracks
The situation	I was planning which tracks should go in a mix (approx 50-60 minutes long) and in what order.	When preparing to DJ in a bar with a partner - 'back to back' DJing (alternating 2 records each)
How shape was used (please give as much detail as possible)	I conceive of certain tracks as being 'higher' than others. This has to do principally with their mood and energy level. More elevated, 'peaky' tracks are higher, while more relaxed tracks are 'lower'; certain tracks - normally called 'tracky' - are quite 'flat' - they keep the mood and energy level constant. So when thinking about how to put together a mix, you have a sense of overall contour. You take care to create a meaningful trajectory both on a large scale and a smaller scale. For example, on a small scale, you might alternate high and low tracks to create contrast and a sense of development (banging in 'peaky' tracks one after another quickly loses its impact unless carefully done). On a large scale (over a whole mix) you might combine tracks in such a way as to produce an arch shape, meaning a mix that gets more and more peaky towards the middle then tapers off towards the end (this would be a classic and very common contour). Or you can start off 'flat', keeping the mood rather level then gradually introduce peaky elements by stealth, as it were. There are many different possible strategies, but they all involve thinking about shape. Also, the manner in which you mix one track into the next also has shape: you can mix in a 'choppy' way (quickly, making sudden movements of the faders on four/eight/sixteen bar downbeats), or you can do a 'smoother fade', which is more gradual and linear. These produce very different effects: the choppy mixing style is more exciting as it foregrounds the introduction of the new elements, but the smooth style is more hypnotic as it emphasises their continuities.	We were discussing which track to open our set with, given the style of the DJ who would be playing before us. We wanted to find an opening track that would work well after the previous DJ and be significantly different but also energetic enough not to clear the dance floor. We discussed it in terms of energy level, often using contour metaphors, which are absolutely central to DJing (peaks and troughs, building it up, taking it down).

When asked to describe links between music and shape, she reiterated her relatively large-scale ideas concerning shape in relation to patterns of mood or energy level over different tracks in a set. She linked these moods and energy levels to 'a metaphorical scale of high and low arousal and accompanying notions of contour', which in turn were seen as related to various musical features including tempo, timbre, instrumentation, thickness of texture, and harmonic features. She commented that 'shape is a general concept that is inseparable from all these things'. The ways in which she described building sets is similar to the guidance provided by Broughton and Brewster (2002), in their book How to DJ Properly, for encouraging and controlling the energy on the dance floor (see Figure 1). The authors argue that a DJ's choice of what records to play, and in what order, is central to shaping a set, and that mood and tempo are integral to this. The specific 'arch' shape mix she refers to ('a mix that gets more and more peaky towards the middle and then tapers off towards the end') has been previously noted by Spring (2004) who argued that a 'typical' set "begins softly and builds to an intense peak and then mellows out to bring the dancers down and signal the set's end" (p. 50).

In addition, she related shape to space, discussing both the acoustics of a particular room and the appropriateness of a track for that space, and her manipulation of the equalisation (EQ) of tracks to make them blend, which as she commented, 'means thinking in terms of frequency space, often on an up/down or left/right scale'. With the exception of the specific technical discussion of equalisation, these comments were similar to those made by classical musicians in the interview studies, who discussed tempo, timbre, thickness of texture and harmonic features as influencing the musical shape. Harpsichordists in particular also discussed listening to the tone of the instrument they were playing and its effect in the particular acoustic of the room in which they were performing.

While the responses of this single participant of the questionnaire study are interesting and insightful, she represents just one view of the importance of shape in relation to music. In addition, her exposure to musicology and classical practices through her degree courses and performance on the violin and piano may also have influenced her ideas concerning music and shape, even when she relates her views in relation to her practice as a DJ. For these reasons, an interview study of the ways in which DJs use the idea of shape or shaping in relation to music was undertaken. It was hoped that a more in-depth study might allow insight into DJs' practices and a comparison with the findings concerning classical musicians outlined above.

Participants

Three participants were recruited through personal contacts of the lead researcher. Each participant varied in their area of expertise, and in the style(s) of music they most enjoyed performing. They also differed in the equipment they were using, ranging from vinyl decks, to vinyl decks coupled with vinyl emulation software, to digital media players with a vinyl simulation platter (see Figures 2-4 below).

Matthew a.k.a. DJ Kanza is a semi-professional DJ who has been mixing for ten years (2003-2013) and specialises in drum and bass music. 6 Aged 30, he regularly performs in and around the Midlands, UK, and more recently internationally, and hosts three Internet radio shows per week. Although he has recently learnt to perform on CD decks and Traktor he uses vinyl decks in the majority of his live performances. 7 He is self-taught, using the Broughton and Brewster (2002) book and through the experience and guidance of peers. In the interview, he used 2 x Technics 1210 mkII turntables, Pioneer DJM600 mixer, Ortofon Pro S carts and needles, Sennheiser HD25 headphones, and a desktop PC (Windows) with the software programmes Cool Edit Pro 3 for recording mixes and Edcast (an audio encoder) for broadcasting radio shows (see Figure 2).

Richard a.k.a. C@ in the H@ [sic] a.k.a. The Doctor is a semi-professional DJ and producer who has been mixing for 15 years (1998-2013) and producing for six (2007-2013). 8 Aged 32, and also self-taught, Rich performs electro swing and glitch hop under the alias C@ in the H@, and hip hop, reggae, bass music, and party mash-up genres under the alias The Doctor, and performs across the UK and internationally. He began his DJing career using only vinyl, but now uses the vinyl emulation software Serato which allows the user to perform with digital music whilst retaining the original feel of vinyl for scratching techniques. 9 He is a member of the Alternative Dubstep Orchestra, a live dubstep orchestra; the Brotherhood of Filth, a three-piece scratch crew; and The Silverbacks, a live hip-hop and funk band consisting of trumpet, drums, guitar, bass, keyboard, synth, scratch DJ and live MCs. He also runs an electro swing-based record label, Ragtime Records, launched in November 2012. 8 In the interview, he used 2 x Technics 1210 mkII, Vestax PMC 05 Pro 3, Ortofon Scratch carts and needles, Sennheiser HD220 headphones, and a Mac Book Pro, Mac OS X Version 10.6.8, with Serato SL-3 hardware and Scratch Live software Version 2.4.2 (see Figure 3). Note the turntables are in 'battle' position to facilitate scratching techniques (cf. Hansen, 2010).

Anthony a.k.a DJ Switch is a professional DJ specialising in turntablism. 10 Aged 23, he started mixing at the age of 11, and was self-taught until completing a course at the DJ Academy at the age of 15. Starting out with styles such as hip-hop infused rock (he mentions the collaboration between Linkin Park and The Executioners called It's Goin' Down as having been particularly inspiring) and UK garage, he has learnt mixing, scratching and turntablism techniques and has gone on to become one of the most highly regarded multi-genre DJs of his time. He has won the World DMC Championship three times (2008-2010) and has regular guest appearances on national radio stations such as 1Xtra, Radio 1 and the BBC Asian Network. In 2010, he performed closing sets at Glastonbury and Bestival, and in 2011 was the first DJ ever to play at the BBC Proms performing Gabriel Prokofiev's Concerto for Turntables. He is currently involved in a range of projects including the Nottingham-based band Hey Zeus, a four-piece band consisting of DJ, keyboard player, drummer and rapper; Boom Operators which is a six-piece live hip-hop, electronica and dub act featuring beats, bass guitar, beatbox/vox, turntablism and scratching, percussion and electric guitar; Brotherhood of Filth which is dubstep scratch; 8 and Tony is currently one of the forefront DJs of the electro swing genre. He is resident tutor at the DJ Academy, where he teaches turntablism and digital technology courses and also one-to-one tuition in beat juggling. 11 Tony stated that it had taken him approximately ten years of practising (at every available opportunity) in order to reach the standard he was currently performing at, which echoes the findings of previous research by Ericsson, Krampe, and Tesch-Römer (1993) on the role of practice in acquiring expert performance.

In the interview, Tony used 2 x Denon SC3900 Digital Media Players, a Rane 61 digital mixer, Denon DN-HP1000 headphones and a 13 inch MacBook Pro, Mac OS X Version 10.7.5, with Serato version 2.4.3 (see Figure 4). Given his DMC titles, Tony has been asked by Denon to promote their new equipment and the set-up presented here is at the forefront of digital media players. 10 The Denon SC3900 player is a CD turntable with a nine inch platter that moves in real time and is also a digital controller for most DJ software.

Procedure

The interview comprised a mixture of musical demonstrations and discussion, focused around unfamiliar records provided by the researcher as well as pieces from the DJs' own collections. Identification of the unfamiliar tracks was a joint effort between the participant and researcher. The interview began with the researcher asking the participant to demonstrate the way in which they would perform with unfamiliar records for approximately five minutes, which was followed by a discussion of the process. Five minutes was chosen because it is approximately the time that would be taken to mix two or three drum and bass records together. The researcher then asked the participant to perform with the records once more, but with a particular focus on the musical shape, or the way in which they were shaping the music. Again, this was followed by a discussion of the process. The participant was then asked to perform the records in the same order without any shaping, and this was also discussed. A more general discussion of their general practice in relation to musical shaping followed, with the participants using familiar examples from their own collections.

The interviews were recorded using an Olympus WS-321M voice recorder and a Panasonic NV-GS230 video recorder. The procedure was approved by the ethics committees of King's College London (REP-H/09/10-6) and the University of Leeds (LTMUSC-014).

Method of analysis

The interviews were analysed using Interpretative Phenomenological Analysis (IPA). Influenced by phenomenology and hermeneutics, this approach is particularly suited to small sample sizes as it focuses on an idiographic understanding of the data in question (Smith, Flowers & Larkin, 2009). The analysis followed the procedure described by the method's pioneers (Smith & Osborn, 2003; Smith, Flowers, & Larkin, 2009), beginning with verbatim transcription of an individual participant's data and subsequent familiarisation with its content before an initial summary was written. Following this, two distinct stages of coding were undertaken: the phenomenological coding focused specifically on understanding the participant's conception of the music they were performing and the ideas they expressed verbally; the interpretative coding allowed patterns to be identified in the data and links to be made with other ideas and the lead researcher's own knowledge. Following this, themes were identified in the data and clustered together before the coding process was reviewed by the second researcher. Each interview was analysed individually up to this point, at which time the themes from each interview were compared and an integrative analysis of all the interviews was performed. Finally, a narrative was constructed to describe the findings.

The analysis of the transcribed verbal material was aided by the use of the audio and video recordings during each stage of the analysis. Where participants referred specifically to musical shapes they tried to create—and demonstrated these shapes using gestures or other visual means—video clips were isolated and examined closely in relation to the surrounding interview content. This approach was influenced by the methods demonstrated in Cienki and Müller (2008).

A draft copy of the final paper was provided for participants for reflection and feedback. Matt and Tony felt that their interviews and practices were accurately represented; Rich made a few minor changes and additions, which will be indicated in the results.

Matt

Matt's extensive vinyl collection (over 3,000 drum and bass records) meant that choosing unfamiliar tracks for him to perform with took some consideration. He was asked to perform with these tracks for five minutes as he would if he were starting a live DJ performance. 12

After his first performance he talked about the mechanics of mixing. He reported that the most essential skill was listening through headphones to ensure that both tracks are in time with each other by adjusting the pitch fader on the vinyl decks. Maintaining consistency in volume and equalisation were his next concerns ("as you move from one tune to the next, you reduce elements of the out-going tune, increase elements of the in-coming tune, so you get a smooth transition"). He emphasised the role of familiarity in enabling a smooth mixing style ("normally when you're playing with records you know, that information is already in your head from previous use…but you're never entirely sure with a new record what the pitch is and what the volume adjustment is going to be"). He described the transition period as a creative practice, using EQs as improvisational tools; he was 'creating his own tune' from the existing tunes by altering the low, mid and high range elements. He also described the use of special effects (e.g., echo, flange, filters) on the mixer which allowed the emphasis of specific preferred elements (e.g., vocal melody, drum rhythm). He argued these added "a little something extra to the tunes" but maintained that they were easy to over-use ("I'd rather under-use them than over-use them"). He re-emphasised the role of familiarity, and the importance of listening and responding quickly ("particularly when you are mixing two unfamiliar tunes together, it's being aware that things can change at a moment's notice, and being able to react, ear to eye to hand coordination to work out what's starting to sound wrong"). He argued that it is more difficult to be creative with unfamiliar tracks because of the need for a higher level of concentration on the structural features than is required with familiar tracks.

When asked to perform a second time thinking about the shape of music, or musical shaping, his initial response was "I'm not entirely sure music has a shape" but began his performance anyway. When asked to describe his approach, he reported that he visualised each track as a graph ("in my mind's eye I visualise each individual track as a graph with time along the X axis obviously, and I suppose sound intensity as the Y axis") suggesting that the shape of music is represented by its waveform. Despite his initial comment, he went on to describe his 'tried and tested' mixing style with respect to shape, describing the mix on a large scale (over the course of an hour) and small scale (during the transition between records).

On a large scale, his main aim was to mix records in quick succession using a smooth (rather than 'choppy') mixing style, building the overall sound intensity/energy (he used these terms interchangeably) gradually over the course of an hour whilst allowing for dips in energy at certain points to enable him to build the energy level back up again. We have depicted the shape of the mix he described in a graph which plots sound intensity/energy against time (see Figure 5). This preferred shape represents a combination of the 'Up, up and away' and 'Riding the waves' mixing styles outlined by Broughton and Brewster (2002). The sharp incline in sound intensity during the first five to ten minutes correlates with Matt's use of the opening section of the mix to show off DJ skills and versatility by mixing a diverse selection of tracks quickly. He reported he would build the energy for ten minutes, and then after four or five records allow one of the breakdowns (section in a tune where the bass is typically stripped out) to come through to give the audience a break from dancing and to allow him to rebuild the energy. He stressed the importance of variation in tunes over time, and also the role of playing familiar older tunes to please the crowd and add an element of surprise (cf. Huron, 2006).

On a small scale, Matt's aim was to select tracks which fitted together well ("choosing a tune which complements the existing tune through either similar or opposing elements"). In the transition period between two tracks, he reported that his first step is to identify the track's pattern, which can be done visually by inspecting the grooves on the record. He then focuses on the first entry of the bass line (the 'drop') and on mixing the bass line from the in-coming record before the bass line finishes on the out-going record to maintain consistency in energy level. He then intensifies the mix with creative use of EQs to add extra elements to the tune which "goes beyond what people know from that pre-recorded music". This, he noted, gives the performance a feeling of acceleration without increasing the tempo. Matt also referred to the way in which he used the shape of tracks to vary the intensity, such as bringing the music to a halt unexpectedly ("sometimes for dramatic effect you can use the shape of the record, to have a dramatic pause") or leaving the tune to breakdown so that the overall shape descends into a trough (see Figure 5).

When Matt was asked to perform without any shape or shaping of the music, he described how he had changed his behaviour:

"Basically, I tried to unlearn everything I know about mixing, with the exception of how to beat match cause that would have just sounded horrible. So I did not use the existing shape of the records in any way to complement the shape of the adjacent record; intentionally did not attempt to match the peaks and troughs; intentionally did not use the EQs to add definition to any specific elements of either record. There is existing shape within the music itself so it isn't possible to completely remove shape, but what I set out to do was to not enhance the existing shapes in any way, in the ways that I would normally do when I mix, so basically stripped back all of the elements that I add, or that the DJ would add to a mix by choosing specific points to cue a tune, to drop a tune in, to take the next tune out. I intentionally brought a tune in out of sequence at one point as well, out of phrase, to just make the overall shape more, more muddy, so less defined as a whole so it would be less pleasing and has less direction to the listener."

The discourse of 'unlearning everything' suggests that Matt sees shape as integral to his DJ performances. When asked whether he found it a strange thing to be asked, he argued that he had not really thought about it until the interview, and that is was an abstract concept to try to articulate into words due to its subjective nature:

"Well it's not a discussion I've ever had before, the shape of music is something that only really existed in my head, up until this point, although I'm sure that everybody defines the shape of music in some way to themselves but because these things are so very subjective, it wouldn't be necessarily something that you would, discuss with someone else, or you couldn't really draw the shape of music beyond, to me, what I've just explained about it being a graph of intensity against time. There are other people that maybe would say, this music is a triangle, this music is circular, this music is a zig-zag, and if you had synaethesia, obviously music has a lot of other shape, texture and colour elements to it, so, music anyway is quite a difficult thing to put into words, particularly non-vocal music, because it's about the emotions that it stirs up within you so discussing the shape of music is quite an abstract concept, different people would use different language to discuss that."

When asked about words with a similar meaning to shape in a musical context, he suggested 'vibe' or the 'feeling of the music'.

Matt stated that he would be more likely to talk about the overall shape of the music if he was performing with another DJ. He said that this would normally take the form of a discussion at a broad level about the selection of tracks they were going to perform, and that a mix with a co-performer tended to be more undulating overall (see Figure 6) because no one individual had complete control. He drew on a recent example of performing with someone for the first time, maintaining that they would not need to have such a detailed conversation the following time they performed together: "that pre-existing discussion regarding shape would be implicit for future reference."

Rich

Rich performed with several drum and bass records from the first author's personal collection which required him to perform with vinyl (and not Serato). 13 He was unfamiliar with all tracks except for one track 'Movin' Fast' from the album EP Stonehead by the artist Enei so did not perform with this track. 14

When he talked through his approach to the first performance, he emphasised the importance of familiarity with tunes. He chose vinyl from the Enei album for the first two mixes, referring to this as a "safe bet" because he had some awareness of the artist, and because tracks from the same EP often mix well together. He then talked about mixing straight to the 'drop' because he was not familiar with the introductions of the tracks and wanted to avoid mistakes. He stressed the advantage of being able to see the waveform in Serato as you can ascertain more quickly where structural landmarks are (e.g., breakdowns, samples) to facilitate quick mixing and allow time for scratching. The importance of identifying a sample (e.g., vocal snippet) to scratch with as early on as possible was clearly highlighted in his reflection on the performance.

When Rich was asked to perform with the same tracks again, this time thinking about the shape of music or musical shaping, his initial reaction was jovial (he said "oblong", laughing) but then slightly perplexed ("I don't really know what you mean by that, sorry"). When prompted to take his own concept of what 'musical shaping' was (or was not), he said that he did not think he would do anything differently to the first time round. This suggests that his notion of musical shape (whatever this constitutes or represents) is inherent in his DJ performances already. At the end of the second performance, rather than describing the specific techniques he used to shape the performance, he emphasised that he had not been able to mix in the way he wanted to because it was not possible to see where the samples were. Ideally, he noted, he would have identified a sample to scratch with, started the mix with a slow introduction without a beat so he could scratch over the top, and then scratched the second tune straight into a drop. He pointed out that the sample he had wanted to scratch with was on the first tune so he was not able to do this on vinyl format (but this would have been possible on DVS as you can load the same track to play on both decks). Describing hypothetically how he would have performed with his usual set-up (Serato) provided insights into what was important for him in shaping a mix, namely, familiarity with records, a real-time graphical representation of musical materials, and identifying samples to scratch with.

After the second performance, Rich was asked to perform without any shape or shaping of the music. His immediate response was "so you're asking me to play badly?" which implied that a good performance needs shaping. When he talked through his approach afterwards, he said he had intentionally disregarded structural features of the music, by bringing a track in out of phase, and not using the EQs. He went on to say that "trying not to shape seems a bit unnatural" suggesting its implicit nature, which was further supported by his response to the question of whether it was an odd thing to be asked "it [shaping] is a bit of a strange thing to be asked but actually it's something you do without really thinking about it".

He then discussed his conceptions of shape in relation to mixes and tracks, explaining how different mixes can take different shapes, indicating through gesture and the use of contour metaphors (e.g., build it up, take it down). He said a mix might be one 'vibe' all the way through, gesturing a horizontal line, that it might be gradually building throughout, gesturing an upwards line, or that it might be fluctuating up and down, gesturing a sine wave (see Figure 7). The use of a wave gesture can be compared with the 'riding the waves' mixing style outlined by Broughton and Brewster (2002). Rich also emphasised the importance of varying musical styles during his mixes. He reported that if he plays a jungle set, he typically starts with a reggae track to warm up the audience, subsequently moves into jungle tracks, and then returns to reggae later on. This shape can be compared to the 'rollercoaster' mixing style outlined by Broughton and Brewster (2002) in which a DJ builds the tempo and then dramatically reduces it at specific points. He argued that a high level of shaping depends on technical ability and requires taking risks, such as performing reggae and electro swing styles together, or performing drum and bass and then introducing soul music unexpectedly ("all of a sudden there's a massive change of energy, but as long as you keep the shape going").

In addition to mixes, Rich described how the shape of tracks ultimately determined the mixing style that he would employ and drew on two examples to illustrate this (see Figure 8). The first track was The 900 Number by the 45 King. He explained that because the track has "no shape at all", as it is two repeated bars on continuous loop, he has to build the track into a shape by adding a cappella, beat juggling two copies (i.e. of the same track), layering, or equalisation. He also emphasised the need to mix this track quickly (within half a minute) to keep the audience engaged. In contrast, Rich described DJ Shadow's Stem/Long Stem as "extreme shaping" of a track. He described its distinct sections ("starts off really chilled, then Gabba craziness, then chilled again, and then really heavy again") and explained that he would not mix another track or a cappella over the top because the tune has "enough shape of its own".

Rich also talked about the shape of tracks in relation to his own musical productions. He talked of going through different processes depending on the genre he was producing. For example, when producing electro swing, he identifies samples he wants to use first, then builds a beat, loops the beat and layers the samples. When he is producing dubstep, he builds a beat, then works on the bassline ("to me the bass is the shape") and then adds in additional elements. He emphasised that when producing you have to think about listening and performing ("you need to be able to listen to the track on its own and it needs to be listened to in a DJing capacity, because if DJs can't mix it, then they're not going to play it, unless they're skilled"). 15 He illustrated further by describing a current track he was working on, outlining processes he had been through and explaining that it had been necessary to remove some of the elements he had embedded in the track to make it easier to mix. A final point, made in response to the draft paper, was his use of a microphone when performing to emphasise specific 'peak' structural points in his mixes. 16

Despite initial assertions that he was not sure what was meant by the shape of music, and musical shaping, his accounts suggest that notions of shape are deeply ingrained in his performance practice and music production.

Concluding remarks

DJ practice is underrepresented in musicological enquiry. This may be because musicology has only started studying popular music relatively recently; because of the lack of formal training courses (rendering the art of mixing a somewhat hidden practice); because of the improvisatory nature of the art, or a combination of these factors; whichever of these, it is clear that more research needs to be carried out. The current study provides a springboard not only for further studies on musical shaping, but for further investigation of the many facets of DJ performance. IPA is an appropriate method for this kind of exploratory in-depth study of expert musicians' experiences because it is designed for idiographic investigations. The procedure and analytical technique developed by Prior (2012b) and employed in this research could be used as a template for future investigations.