Brass Pedagogy · Paper

Exposure to Musical Models

Learning how it should sound

A blind spot in music education

Jaap Klijnstra · English translation

Version 1.7 | 6 September 2026

For teachers and researchers in instrumental music education

Abstract

Learning and maintaining brass playing involves developing an accurate sound concept, learning to hear relevant differences, controlling performance and learning the connections between these processes. A wind player must imagine, prepare, produce, perceive and assess a note. Being able to recognise a good performance does not mean that the necessary control is reliably available at the right moment. Without an accurate external and internal reference, performance may become more stable and efficient, but stable and efficient within a limited or undesirable solution.

Conventional brass pedagogy acknowledges listening and demonstration, but almost never operationalises model formation in hours of exposure. At the same time, it devotes considerable time to separate breathing exercises and buzzing. This reverses educational priorities. A healthy learner already has sufficient breathing capacity for a normal human activity such as brass playing. What the learner does not yet possess is a refined internal model of sound, note onset, articulation, timing, intonation and phrasing. That model develops through repeated exposure to good performances.

Buzzing is also not the same task as playing the complete instrument. The different acoustic load organises different coordination. After buzzing, the player must rediscover the coupling between lip oscillation, mouthpiece, air column, instrument resonance, timing and sound. Buzzing therefore incurs three costs: buzzing time itself, lost instrument-specific practice time and the subsequent work of recovery.

Ordinary media make it possible to organise this missing condition. Audio through Spotify or other sources makes good sound examples available on demand; video through YouTube or WindPlayer adds the visible performance; play-alongs connect perception with the learner's own actions. These media can both encourage more voluntary playing hours and substantially increase the proportion of playing hours with a model. Stereo can refine the auditory model, but remains secondary to frequent, relevant exposure to musical models.

Keywords: target model; model formation; exposure to musical models; brass pedagogy; audio; video; play-along; motor learning; WindPlayer

1. Comparison is impossible without a clear target

A player does not produce a note in a vacuum. Before each attempt, consciously or less consciously, there is an expectation of the result. The player begins, hears and feels what actually happens, and uses the difference to organise the next attempt. This simple cycle of target, action, perception, comparison and adjustment is the basis of perception-guided motor learning.

In brass playing, the target is much richer than the right note at the right time. It also includes timbre, note onset, articulation, dynamic development, intonation, timing and phrasing. Notation describes only part of this. The learner therefore needs to know not only what is played, but above all how the intended result sounds.

This requires a model: a good performance that can be perceived as a reference. Through repeated listening and watching, this external reference gradually becomes an internal expectation. That expectation does not prescribe individual bodily movements. It gives the motor system a result towards which it can organise performance.

Instrumental lessons, however, are scarce. A learner generally meets the teacher once a week and spends most playing time working independently. An example given during the lesson disappears when the learner goes home. During subsequent playing, the learner mainly hears their own developing performance. The question is therefore how the model can remain available during the much greater number of independent playing hours.

An ordinary musical note may last about 0.2 seconds. Pitch, timbre, dynamics and articulation must then be available at the onset; they cannot be assembled only during a note lasting several seconds. A longer note may make certain differences easier to hear, but does not replace the precise coordination of a normal, brief musical action.

This issue concerns more than beginners. An experienced wind player must also maintain perceptual discrimination, control and the connection between expectation and result. Less frequent exposure, a changed playing environment or prolonged repetition within a limited model may affect the performance available to the player. Exposure to models therefore serves both development and maintenance.

2. What the model does in learning

2.1 From external example to internal expectation

The word model may sound abstract, but its function is simple. A good player makes audible and visible a result that the learner cannot yet reliably produce. Repeated exposure makes that example recognisable and predictable. Before beginning a note, the learner starts to anticipate how the note or phrase should sound. This is the internal model.

That internal expectation is necessary, but insufficient. The wind player must also learn which preparation and control produce the intended result under the current conditions. Hearing, bodily sensation, learned motor coordination and the instrument's response work together in this process.

In this paper, control model is used solely as a working term for the learned relationships between preparation, breathing during playing, embouchure, tongue, instrument operation, anticipated consequences and experienced result. The term names a relationship relevant to teaching; it is not presented as a newly demonstrated neurological system.

2.2 The comparison cycle

The audible model does not prescribe actions for the lips, tongue, jaw or breathing, nor does it control performance from outside. It provides a reference through which the wind player learns to hear differences. The player attempts the task, perceives their own result and its bodily consequences, compares and organises the next attempt. Sound concept, discrimination, control and their connections thus develop together. Not every comparison needs to be consciously expressed in words.

MODEL → PLAY → LISTEN → COMPARE → ADJUST

2.3 Hearing, seeing and playing along

Listening supplies the necessary information about the musical result. A visible model adds information about how the performance unfolds: entries, breathing within phrases, timing, instrument movement and musical intention are perceived as one event. When playing along, the model and the learner's performance are also heard within the same musical time. The learner can therefore experience their relationship immediately.

2.4 Sound, image, notation and playing along

The application is not one new method. It combines ordinary media functions, each with its own contribution. An audio recording makes the musical result audible. Video shows the player performing. Synchronised notation shows what is played at each moment. In a play-along, the learner performs the same musical task simultaneously. The essential feature is not the screen, but the continuing availability of a perceptible target during playing.

A media environment for exposure to models can combine four layers of information:

• Auditory model: an expert performance with recognisable instrumental sound, timing and expression;
• Visual model: visible information about the player, instrument, entries, breathing within phrases and musical movement;
• Task representation: synchronised notation or another time-linked indication of what should be played;
• Simultaneous performance: the learner plays along and tries to hear the model and their own result as one musical whole.

Not every application needs all four layers. Listening alone, demonstration video, notation video and play-along are possible partial forms. The pedagogical distinction arises because the learning environment is deliberately designed around the connection between perception and playing.

The function of the media is to keep the target audible and visible at the moment when the player attempts, compares and acts again.

2.5 Maintaining model formation and control

An acquired performance does not automatically remain equally available under all conditions. Register, dynamics, repertoire, ensemble composition, venue, fatigue and the audibility of other players change the information on which a wind player acts. Repeated comparison with a relevant model helps check whether the intended sound, articulation, intonation and phrasing remain immediately available.

Hours without an external model are therefore not worthless. They reveal what is independently available. They should, however, alternate with a return to a reliable reference, so that gradual deviation is not assessed exclusively against the player's own, now familiar performance.

3. The pedagogical blind spot is exposure

Brass pedagogy recognises the idea of a sound concept. Listening to excellent players, inner hearing, demonstration and imitation are considered important in various traditions. In broader research on instrumental teaching, aural modelling has also been described as listening to expert performances to develop internal auditory representations and expressive performance approaches (Meissner, 2017; Meissner, 2021).

The blind spot is therefore not that the model is entirely unknown. It is that exposure is rarely treated as a curriculum component to be planned, quantified and evaluated in its own right. Methods schedule exercises, scales, pieces and technical points. The number of minutes per day in which a learner hears and sees a high-quality, instrument-specific reference usually remains unspecified.

This is a fundamental distinction. Advice such as “listen to good players” does not yet specify:

• which type of model is relevant to which task;
• how often and for how long the model should be available;
• whether the learner listens before, after or during their own playing;
• how hearing, seeing, predicting, performing and comparing are connected;
• how the teacher establishes that the internal model is becoming more accurate.

Audio, video and play-alongs make it possible to stop treating exposure as an incidental consequence of the lesson. It can be designed as a dose: number of sessions, minutes per session, type of model, alternation between observing and playing, and degree of simultaneity.

3.1 Why has this not previously been described in this way

If model formation is so essential to learning to play, one question immediately arises: why has conventional brass pedagogy not long treated exposure to models as a separate, measurable curriculum component? The answer is not that listening, demonstration or imitation are unknown. Good players have always demonstrated, learners have always been directed towards concerts and recordings, and concepts such as sound concept have existed for a long time. What is missing is a coherent development of these familiar elements into an organised learning condition.

The missing argument consists of five successive steps:

1. the learner must learn to recognise and predict how the intended playing sounds;
2. that internal model develops through repeated exposure to good performances;
3. the amount of exposure can be expressed in minutes and hours;
4. model hours must be distinguished from lesson hours and playing hours without an immediately available model;
5. ordinary audio, video and play-alongs can keep the model available during a large proportion of independent playing hours.

The individual components are known. The pedagogical connection and the accounting of time are missing.

3.2 The model depended on the teacher's presence

The traditional instrumental lesson developed in an environment where a high-quality performance could not be accessed at any desired moment. The model was present when the teacher played, when the learner played with a good player or ensemble, at a concert or through an available recording. Once the lesson ended, the direct instrument-specific reference usually disappeared from the practice environment.

Under those conditions, it was understandable to consider occasional demonstration sufficient. Prescribing twenty or forty minutes of daily exposure to a suitable player on the same instrument was scarcely practicable. Today's availability of streaming audio, video, phones, laptops, televisions and earphones has removed that material limitation. The pedagogical structure, however, has largely remained the same.

3.3 Demonstration was described as a teacher's action

In methods and teacher training, demonstration is mainly described as something the teacher does. The teacher demonstrates a note, phrase, articulation or technical possibility. This does not describe what the learner receives throughout the week. The relevant curriculum questions remain unanswered: how many minutes does the learner actually hear a good performance, how specific is that example to the task and instrument, and how many of those minutes coincide with the learner's own playing?

The distinction between provision and received dose is fundamental. One demonstration may be pedagogically excellent, but supplies only a small amount of exposure. Only when time is counted from the learner's perspective does it become clear that advice to listen carefully does not yet constitute an organised learning environment.

3.4 Visible exercises displace invisible model formation

Breathing exercises, buzzing assignments, long notes, scales and studies can be precisely notated, prescribed and ticked off. Model formation is less visible. The learner hears increasingly fine differences, develops expectations and begins to predict how the result should sound before an entry. This perceptual development cannot readily be identified in one bodily movement or one page of a method book.

This produces a distortion in the curriculum. What can easily be prescribed receives much attention; what is less visible but gives direction remains implicit. A method then specifies in detail which action should be repeated, but not how much high-quality musical information must be available to guide that repetition.

3.5 The body-centred tradition places the means before the target

Much brass pedagogy organises learning around breathing, embouchure, tongue, jaw, mouth corners, mouthpiece pressure and buzzing. The problem is thereby framed as separately developing physical prerequisites and component skills. The musical model then appears as inspiration, illustration or final result, rather than as the continuing reference against which the motor system selects its solutions.

In a perception-guided approach, the order is different. The audible musical result is the target from the beginning. The body organises breathing, lip oscillation, articulation and movement within the complete task. Without a sufficiently accurate target model, the selection criterion is missing: the system can become more stable and efficient without approaching the musically intended result.

3.6 Separate research and teaching traditions

The necessary building blocks are distributed across different fields. Motor-learning research studies observational learning and the relationship between target, perception, action and feedback. Music education research addresses aural modelling, imitation, expression and learning by ear. Brass acoustics describes the coupled interaction between player, lips, mouthpiece and instrument. Digital music education examines audio, video, online instruction and interactive applications. Meanwhile, brass teaching uses demonstration, joint playing and recordings.

Because these fields have different concepts, research questions and professional groups, their simple shared implication has remained largely unspoken: if brass playing is perception-guided motor learning, the availability of information about the intended result must be deliberately designed and counted.

3.7 Successful learners conceal the missing curriculum

Learners who listen extensively to good players, join strong ensembles, attend concerts and study recordings on their own initiative organise a rich model environment for themselves. Their exposure hours are not recorded. When they develop well, the result is readily attributed to talent, discipline, breathing technique, embouchure or the teaching method.

The other side remains hidden too. A learner who mainly hears themselves, other beginners and the overall ensemble sound builds a model from that input. When timbre, articulation or phrasing remains limited, attention usually turns to the learner's body or technique. The poverty of the auditory environment is rarely investigated as a cause in its own right.

3.8 Media were treated as aids rather than distribution of the model

Audio and video in instrumental teaching are usually treated as supplements: an inspiring recording, an explanatory video, an accompaniment track or a way to listen back to one's own performance. The existing lesson structure remains the starting point. Media add something, but do not change how the model is distributed across the week.

The approach in this paper is more fundamental. Audio distributes the sound model. Video adds the visible performance. Synchronised notation connects the passage of time with the part. Playing along immediately links the perceived performance to the learner's own motor task. The devices are not the pedagogical subject; they make a previously scarce learning resource available during independent playing hours.

3.9 Where the innovation lies

Without a systematic literature review, it cannot be claimed that this idea has never been formulated before. The wider literature demonstrably contains related concepts and applications. The defensible observation is narrower and at the same time more important: in conventional brass pedagogy, exposure to models does not function as a separate, quantified and evaluated curriculum component.

The innovation therefore lies neither in advising learners to listen to good players nor in the existence of audio, video or play-alongs. It lies in the complete pedagogical formulation: model formation is a necessary learning process; exposure to relevant performances feeds that process; exposure can be operationalised in time; and media can dramatically increase the number of hours during which the model is available while the learner plays.

The medium is not the innovation. The innovation is to organise, count and investigate model availability as a curriculum variable.

4. Hours of learning, practice and exposure

In language learning, it is taken for granted that development consists of more than lessons. The learner receives instruction, studies under guidance and also encounters the language in conversations, media, work, school and everyday life. These hours must not be lumped together. Cambridge English uses guided learning hours for lessons and supervised study. Independent use and exposure outside that supervised context are excluded, although Cambridge explicitly names the amount of external study and exposure as a factor affecting the learning time required (Cambridge English, n.d.).

4.1 Cumulative language-learning hours from Cambridge English

Cambridge presents indicative figures starting from a complete beginner. They are cumulative: 350–400 hours at B1 does not mean 350–400 hours in addition to A2, but the total guided learning hours from the start. As a broad rule of thumb, Cambridge gives approximately 200 guided hours for moving from one CEFR level to the next. It stresses that age, previous language experience, study intensity and exposure outside lessons can produce large differences.

Table 1. Indicative figures from Cambridge English. The fourth column is calculated from the cumulative ranges and shows that an exact hourly increment cannot be derived from these broad ranges.

Their significance for this paper lies not in directly comparing language and musical levels, but in the way of thinking. In language learning, expressing guided learning time in hundreds of hours and recognising exposure outside those hours as a separate influence is considered entirely normal. There is, incidentally, no universal total of hours of language input actually heard during first-language acquisition. Researchers use methods including daylong recordings and sampling, because the amount of child-directed speech and the outcome vary substantially with the measurement method (Ferjan Ramírez & Hippe, 2024).

4.2 The HaFaBra phases specify years and outcomes but not hours

The Dutch national Framework Curriculum for Wind Instruments organises amateur music education into four phases. A is the foundation phase, B development, C further depth and D specialisation, potentially preparing for professional study. The framework gives broad durations and detailed learning outcomes. It does not, however, specify how many lessons, guided study hours, independent playing hours or hours with an expert model are required for each diploma. A HaFaBra diploma therefore describes an attained level, not a fixed total of hours.

Table 2. Phases according to the Dutch national Framework Curriculum for Wind Instruments. The times are broad educational durations; the framework states that each phase requires at least approximately two years, and the complete programme through D eight to nine years (LKCA, 2014/2016).

4.3 From HaFaBra years to hours through a transparent calculation

A worked example is needed to reveal the hidden structure of hours. The following table assumes forty thirty-minute lessons and 150 independent playing hours per year. It also assumes that the teacher plays for five minutes in each lesson as a separately audible model. These are neither national averages nor examination requirements. They are explicit assumptions that reveal what the HaFaBra framework itself does not record.

Table 3. An illustrative scenario, not taken from the framework: 40 thirty-minute lessons per year, 150 independent playing hours per year and 5 minutes of separately audible teacher demonstration per lesson. Ensemble rehearsals and informal listening hours are excluded.

At diploma D, this scenario yields 160–180 contact hours and 1,200–1,350 independent playing hours. The directly and separately audible teacher model accounts for only 26.7–30 hours, approximately 2.2 per cent of independent playing time. Playing together with the teacher may also supply model information, but is not automatically equivalent to attentive listening to a separate model and should therefore be recorded separately.

4.4 What can and cannot be compared between Cambridge and HaFaBra

Equating the levels in content is impossible. CEFR levels describe language proficiency; HaFaBra A–D describe instrumental and musical learning outcomes. C1 is therefore not equivalent to C, nor C2 to D. Nor are Cambridge hours directly comparable with the weekly HaFaBra lesson alone: Cambridge counts lessons and guided study, whereas the HaFaBra framework does not prescribe corresponding time records.

Looking only at orders of magnitude nevertheless reveals something important. Cambridge considers 700–800 guided learning hours plausible for C1 and 1,000–1,200 for C2. According to the calculation, a traditional HaFaBra route may contain only 160–180 individual lesson hours over eight to nine years. Actual instrumental development must then largely occur during independent practice, ensemble rehearsals and other informal learning environments. It is precisely within this large body of hours that the model the learner hears is generally not recorded.

The central comparison is therefore not C1 against C or C2 against D. It concerns the accounting of learning: Cambridge makes guided learning hours visible and recognises external exposure as relevant; HaFaBra describes years and outcomes, but does not count how many hours the learner plays, listens or works with an expert instrumental model.

Here the blind spot becomes measurable. A learner may play more than a thousand hours towards level D, while a separate good player on the same instrument is directly audible for only a few dozen hours. Media can change exactly that ratio: by keeping a relevant model available during a much larger share of independent playing hours, rather than replacing a HaFaBra diploma.

4.5 Three kinds of time around an external reference

Useful learning records must distinguish at least three kinds of time: hours when a suitable external model is audible or visible, hours when the wind player plays without an external model, and recurring checks when independent performance is compared with the model again. The first feed perception and alignment; the second reveal what is independently available; the third show whether performance still corresponds or has gradually shifted.

These categories do not form a hierarchy in which playing alone without a model is inferior. Independent playing is necessary. The problem arises when direct checking is absent for hundreds of hours and the player compares performance exclusively with an internal model maintained by that same performance.

5. Brass playing as perception-guided motor learning

Brass playing is not a separate category of human learning. It is a sensorimotor skill in which intention, perception, action and feedback are continually connected. The player does not produce movement for its own sake, but organises actions to achieve an acoustic and musical result.

Motor-learning research treats observational practice as a relevant form of learning. Reviews describe effects of observing models on skill acquisition and indicate that observational learning can be combined with physical practice (Wulf et al., 2010; Ste-Marie et al., 2020; Han et al., 2022). Krigolson and colleagues (2021), for example, incorporated observational learning into a ten-session practice programme in which participants also physically performed a motor stacking task. This literature is not specifically about brass playing, but is relevant because brass playing likewise requires goal-directed motor coordination.

Music research generally approaches modelling through hearing. Aural modelling uses an expert's performance as an external reference for an internal representation. Studies of expressive performance show that accurate imitation requires not only motor production, but also perception and mental representation of tempo, dynamic and expressive contours (Woody, 2003).

This does not imply that every video model is automatically effective. It does imply that a curriculum making the target model available only occasionally leaves an important learning resource unused. The relevant design question is not whether modelling belongs within motor learning, but how the model is selected, dosed and connected to the learner's own actions.

5.1 Where the first learning gains are likely to occur

Exposure to models is expected to affect most quickly those features that the learner can compare directly and almost immediately with the example. Rhythm, tempo, melodic contour, note duration, articulation, dynamics, timing and phrasing unfold over time and are relatively clear in a good recording. During simultaneous playing, temporal deviations also become immediately apparent. Research on learning by ear mainly supports gains in pitch, rhythm, contour, tempo and musical closure; expressive-performance research describes modelling as useful for timing, dynamics, articulation and phrasing, among other features (Baker & Green, 2013; Meissner, 2021).

This makes a general effect of model use plausible, but does not demonstrate a linear dose–response relationship. For brass players, it has not been established how much faster a learner develops with, for example, twenty, one hundred or five hundred additional model hours. Nor has it been demonstrated that all aspects of playing benefit equally. This paper therefore treats more model hours as a reasoned curriculum variable, rather than an already quantified acceleration factor.

5.2 Why timbre is a special and contested case

For timbre, the relationship between model and performance is less direct than for rhythm or note duration. A learner can immediately locate a timing difference, but must first learn to distinguish subtle differences in harmonic composition, noise, resonance, note onset and the development of a note before these can function as targets for the next attempt. Better sound through exposure to models is therefore plausible, but probably slower, more indirect and harder to measure reliably.

Physical production is also strongly constrained. A brass tone arises from a nonlinearly coupled system of player, lip valve, mouthpiece and instrument. Mouth pressure, lip opening, airflow, lip mechanics and acoustic feedback from the instrument must together allow stable periodic oscillation. Research on minimum blowing pressure shows that stable oscillation regimes arise only under particular combinations of parameters. As a first approximation, the minimum pressure at which such a regime can arise may indicate playing ease and required physical effort, but is not the same as musical sound quality (Mattéoli et al., 2021).

Through repetition, the motor system generally organises performance more stably and with less unnecessary effort. It is too strong, however, to claim that the body always seeks a universal maximum of energy efficiency. It stabilises a solution that accomplishes the perceived task within the existing constraints. If the task consists only of “producing the correct audible note”, a limited sound can nevertheless become a successful and efficient solution. If refined timbre forms part of the perceptible target, the selection criterion for retaining a performance as successful also changes.

The relationship between mouth pressure and aperture thus imposes narrow limits on playability, but does not determine timbre alone. The shape and time course of the lip opening, resulting airflow pulses, lip tension and vibrating mass, dynamic level, mouthpiece and instrument, nonlinear wave propagation and radiation into the room also contribute. Numerical models show, for example, that nonlinear wave propagation can substantially affect the spectrum and therefore timbre (Berjamin et al., 2015). There is no single optimal pressure–aperture combination for “the best sound”, but task-dependent regions within which different stable sound outcomes are possible.

EFFICIENTLY PLAYABLE ≠ MUSICALLY DESIRED TIMBRE

5.3 The precise function of the model in sound development

The model does not prescribe mouth pressure, aperture size or lip movement. It supplies an auditory selection criterion. Repeated listening develops increasingly detailed discrimination and an internal expectation. During playing, the coupled system inevitably varies from attempt to attempt. Among the playable solutions, those whose audible results approach the target can then be preferentially stabilised.

HEAR THE MODEL → PERCEIVE THE SOUND DIFFERENCE → PREDICT THE RESULT → PLAY AND VARY → COMPARE → STABILISE A SUITABLE SOLUTION

Motor efficiency and sound quality are therefore neither opposites nor synonyms. Narrow physical constraints help the system find stable and relatively economical production. The model helps determine which available solutions are musically desired. This distinction may be particularly important for beginners and amateurs: without an accurate sound target, an audible, stable and locally efficient beginner's solution can be reinforced for years.

5.4 Expected sensitivity of different playing features

Table 4. The ordering is a reasoned expectation, not a measured ranking of effects. It makes testable predictions explicit and prevents “more model” from being presented as an undifferentiated miracle.

6. The reversed priorities of conventional brass pedagogy

A learning programme should devote most of its time to what the learner still needs to acquire. In much brass teaching, the reverse occurs. Extensive attention goes to breathing as a separate skill and buzzing without the complete instrument. Developing an accurate internal playing model, by contrast, is usually limited to a brief demonstration, a general encouragement to listen to good players or a chance encounter with a good recording. An already available human capacity is thus taught, a different task is practised, and the actual missing learning target is scarcely organised.

The curriculum extensively teaches what a healthy learner can already do—breathe—and scarcely organises what the learner actually still needs to learn: how good brass playing sounds.

6.1 Breathing is not a preliminary learning task

Brass playing is a normal human activity. Healthy people's existing breathing capacities are sufficient for it. They do not first need to be developed through separate breathing exercises. A healthy learner does not need breathing instruction to learn a brass instrument. The learner needs to learn to play.

During playing, breathing does not occur separately. Inhalation, exhalation, lip oscillation, mouthpiece loading, instrument resonance, note onset, note duration, dynamics and musical timing constitute one task. The body organises breathing within that complete task, as it does in speaking, singing, walking and other normal human activities. Breathing in isolation does not practise those couplings and therefore does not teach brass playing.

The minimum consequence of general breathing exercises is lost time. Every minute of isolated breathing is a minute in which the learner hears no sound model, experiences no instrument resonance and performs no musical action. The same time can be spent directly on the actual task: playing music with a clearly audible model. For a healthy learner, this is no detail, but a choice between nonspecific activity and instrument-specific learning.

6.2 Buzzing is a different motor task

Buzzing is often treated as though it were a simplified version of brass playing. This is incorrect. When playing the complete instrument, lips, mouthpiece and resonating air column respond as one coupled system. The instrument's acoustic impedance and resonances support and constrain lip oscillation. In free buzzing or mouthpiece buzzing, the complete coupled system is absent or substantially altered. The required sound, resistance, feedback and stable solutions are consequently different.

The motor system learns the task actually repeated. During buzzing, it organises pressure, airflow, lip tension and aperture for buzzing. This organisation is not instrument coordination, because the instrument's determining resonances and normal acoustic feedback are absent. Repetition makes this different solution more accessible and stable. That is precisely where the harm lies: the player practises coordination that does not fit the complete task on returning to the instrument.

6.3 Recovery work as an overlooked time cost

The costs do not stop when buzzing stops. Recently activated coordination does not immediately disappear when the instrument returns to the mouth. It affects the first attempts on the instrument. The player must then rebuild the normal coupling. This recovery work is almost never named in methods and is never entered as a time cost in practice schedules.

In concrete terms, recovery means rediscovering:

• the pressure at which the lips readily begin oscillating within the instrument's resonance;
• the aperture size that balances air consumption, response and timbre;
• the distribution of lip tension and mouthpiece pressure allowing free oscillation;
• the note onset in which tongue action, airflow and instrument response coincide;
• the intonation centres and resonance slots of the complete instrument;
• the normal relationship between effort, dynamics and sound;
• the auditory–motor coupling through which the player once again directs performance towards music rather than an isolated lip sensation.

During this return, entries may be less reliable, the tone may respond too tightly or too airily, the player may use excessive pressure and timbre may temporarily shift. Every corrective attempt consumes practice time, and every repeated deviant attempt may become ingrained again. Recovery is therefore not a neutral transition, but additional motor work made necessary solely by first practising a different task.

Beginners and amateurs are affected most strongly. Their instrument coordination is not yet robust and is therefore more readily displaced by recently practised buzzing coordination. An advanced player may return more quickly to established instrument coordination, but buzzing does not provide that player with additional minutes of practising the complete task either.

Buzzing costs time three times over: buzzing time + lost instrument-specific practice time + recovery work.

6.4 The full accounting of time

Conventional time accounting records only how long someone has buzzed. For learning, this is incomplete. The actual burden of a buzzing session consists of three items. The first is direct time on the different task. The second is opportunity cost: during those minutes, the learner could have played the instrument, listened and compared. The third is the time needed for instrument performance to return to its pre-buzzing level.

Table 5. The full accounting of buzzing consists of direct buzzing time, lost instrument-specific playing time and subsequent recovery time.

The size of the third item can be measured empirically: record stable baseline performance before buzzing, then count how many instrument attempts or minutes are needed before response, note onset, intonation, timbre and perceived playing ease again fall within that baseline range. Exact recovery time will vary by player and type of buzzing. The existence of recovery work follows from the task difference; its size is a measurable research variable.

6.5 What the curriculum should organise instead

Scarce practice time should go to the complete task: listening to a relevant model, internally anticipating it, playing the instrument, perceiving the result, comparing and playing again. Target formation and motor organisation are thereby kept together. Every attempt contains the instrument, resonance, musical timing and sound result for which the coordination is intended.

This shifts the main pedagogical question. Rather than asking which preparatory bodily or lip exercise should come first, it asks which musical model should now be available, which passage the learner plays with it, and which audible difference constitutes the next learning step.

7. The white road line and hours with and without a reference

An accurate instrumental model functions like the white line on a road. The line does not steer the cyclist. It provides an external reference through which the cyclist perceives their course and learns the relationship between looking, steering and the actual result. Similarly, an audible musical example does not control the wind player. It enables comparison and helps the player learn to perceive differences in note onset, timbre, intonation, articulation, dynamics, timing and phrasing.

During playing, comparison and renewed attempts, perception and control develop together. The player learns not only how the intended performance sounds, but also which organisation produces that result under current conditions. The white line therefore represents the availability of an external reference while that coupling develops and is maintained.

Without an external model, internal feedback still exists. The player continues to hear and feel their own performance. What is missing is direct checking against an independent, relevant reference. A familiar deviation can consequently begin to sound normal, while the motor system performs it ever more stably and efficiently.

Hours with a model, hours without a model and returning to the model each have their own function. With the model, differences become perceptible and adjustments possible. Without it, independently acquired learning becomes apparent. Returning to it checks whether independent performance still corresponds. A curriculum must therefore record not only total playing hours, but also when the white line was available and when performance was recalibrated against it.

8. How media keep the model available throughout the week

8.1 From occasional to continuously available

During a lesson, the teacher can demonstrate, play together and give feedback. Outside the lesson, this source disappears. An audio or video recording makes the same or another expert player immediately and repeatedly available without additional teacher time. “Continuously” here means accessible at any desired practice moment, not that the learner must listen or watch without interruption.

8.2 From explanatory video to playing environment

Much digital instrumental teaching consists of spoken instruction followed by independent practice. Here, by contrast, the model's performance remains present during the learner's playing. The learner does not have to reconstruct a brief demonstration from memory, but can listen, watch, play, stop, repeat and compare within one task.

8.3 From general ensemble sound to an instrument-specific model

Listening to good ensembles develops style, balance and musical expectation, but does not always supply an isolated reference for the learner's own instrumental production. For developing timbre, onset and articulation, a model on the same or a closely related instrument is probably more relevant. How specific the model should be—in instrument type, register, dynamics or difficulty—is a research question.

8.4 From passive listening to coupled action

A wind player can listen to a recording without applying the perceived information in motor action during the same session. Simultaneous playing links perception and action in time. The player hears the relationship between model and personal performance as the music continues. Timbre and intonation must be separately assessable: a note can be approximately the same pitch yet have a different timbre or onset. An overly loud model can mask differences; a good balance and regular moments without the model therefore remain necessary.

For learning to adapt to a sounding musical example, playing along with a good model recording is superior to practising with a drone. This is an explicit teaching position for that specific learning target, not the outcome of a cited comparative study. A drone's note remains sustained. With a model recording, the wind player must keep following the part through the music and continually relate pitch, timbre, note onset and phrasing to the sounding example.

The central action is simultaneous one-to-one playing until the intonation audibly corresponds, as is possible with chorale recordings by René Fennema. First the model sounds alone, then model and player perform together, then the player performs alone, and finally they check together again. The recording does not respond to the player. This is also its limit: it provides a stable reference, but does not yet practise reciprocal adjustment between ensemble members.

8.5 Drone, tuner and isolated tuning note

A drone and a tuner can offer a fixed pitch reference, but do not automatically address adjustment within a musical passage. In ordinary ensemble practice, correction often proceeds as follows: the conductor hears poor intonation; a player plays an isolated G4 or C5; the tuner shows whether it is sharp or flat; the tuning slide moves in or out; the note is played again; the section then plays together and rehearsal continues.

This procedure corrects tuning at that moment. It does not describe a systematic learning route through which a wind player learns to recognise differences, select a reliable reference within a passage and then decide whether to adjust or hold steady. When the same problem returns, tuning again is not the same as learning to intonate independently within sounding music.

9. Listening and model environments

Exposure to models is not a uniform category. A learner may listen to a soloist, play beside a teacher, follow a video, play along with a recording, sit in a strong section or perform within a full ensemble. A model is present in all these situations, but the same information is not equally clear. The pedagogical value of an environment therefore depends on more than its general musical quality. It depends on correspondence with the task to be learned, audibility of the relevant instrument, visibility of performance, opportunities to compare model and personal result, and the hours during which those conditions actually exist.

Not every good musical environment provides an equally clear model for every learning target.

9.1 The model must fit what is being learned

The model must fit the aspect being learned or maintained. For note onset, timbre, intonation, articulation, dynamics, timing and phrasing, different combinations of instrument, passage, audibility and visibility may be most informative. A good player on the same or a closely related instrument is central to instrument-specific sound development, but an ensemble recording may be better for balance and overall musical coherence. These features do not exist independently and cannot be perceived separately equally well in every environment.

A general ensemble model, for example, may demonstrate very well how a phrase moves within the whole, but much less precisely how a single cornet note begins. An isolated cornet recording can clearly present onset and timbre, but reveals less about how the part relates to woodwind, percussion or lower parts. The question is therefore not simply whether a model environment is good or bad. The proper question is: for which aspect is this environment informative, how specific is the model, and which information does the environment strengthen or mask?

9.2 The richest instrument-specific environment is together in the same room

For detailed development of an instrument-specific model, a good player on the same instrument, visible and audible in the same room, provides the richest reference. The learner receives direct sound and the room's contribution. Onset, core, harmonic structure, dynamic development and projection have not first been transformed by microphones, recording processing, data compression and loudspeakers. The learner also sees when the player begins, breathes, moves, articulates, rests and responds to fellow players.

The greatest pedagogical value arises when the learner does more than listen, playing the same passage alternately and jointly with the model. First the model can sound alone. Then the learner plays. Next both play simultaneously. Perception, prediction, performance and comparison are thereby connected within seconds. The teacher can immediately adjust tempo, register, dynamics and phrase length, and establish whether the learner is actually listening to the result.

This environment is optimal in information richness, but scarce in time. In a weekly lesson, the good player is present only briefly. Furthermore, not all lesson time is spent playing, and only part of joint playing consists of clearly perceptible model situations. The highest quality per minute therefore does not solve the quantitative exposure problem.

9.3 A recorded visible model is less complete but available much longer

A high-quality audiovisual recording cannot fully reproduce the live environment, but radically changes availability. The same performance can be started, stopped, repeated and immediately played along with every day. A large screen keeps instrument, posture, onset, movement and musical timing clearly visible. Good loudspeakers make the model audible in the room without requiring the learner to listen exclusively to a small device.

The strength of this environment is not that it is richer per minute than live joint playing. Its strength is that it can recur hundreds of times outside lesson hours. A somewhat less completely reproduced model available for one hundred hours may become more important for overall model formation than an acoustically perfect example actually heard and played along with for only a few minutes across an entire learning year. Quality and duration must therefore be assessed together.

A large image and good sound are not a pedagogical method in themselves. They improve the perceptibility of the model. A large image makes details and musical movement easier to follow; good reproduction reduces distortion and masking. Stereo can add spatial information and clarity, but remains secondary to instrument specificity, audibility, repeatability and sufficient exposure hours.

9.4 The minimal personal environment

A phone with earphones forms the smallest practically complete media environment: sound, image, notation and controls are immediately available. It offers neither life-size dimensions nor a real acoustic field, but lowers the access threshold to almost zero. Precisely for that reason, it can add many daily minutes. A laptop or tablet enlarges the image and notation. External speakers allow a more spatial presentation. The devices form a scale of accessibility and reproduction quality; they do not change the underlying learning principle.

With earphones or headphones, the learner's own performance must remain sufficiently audible. An overly loud model can prevent auditory comparison. The player then mainly experiences how good the recording sounds and too little of how their own tone differs. A safe playback level, workable balance and moments when the learner reproduces the passage without the model are therefore part of the design, not merely technical details.

9.5 The section as a social and instrument-specific model

A section of good players offers a different combination of information. The learner hears several performances on the same or related instruments, experiences shared breathing and phrasing moments, and learns to align note onset, articulation, intonation, balance and dynamics with others. The model is no longer one ideal voice, but a shared result to which the learner contributes.

The section can be particularly powerful when a strong player sits nearby and remains separately audible. Instrument-specific and social calibration then come together. In a weak, uneven or continuously loud section, however, the opposite can occur. Deviant onsets, poor intonation and limited sound development become the most frequently repeated environment. Several people playing together therefore does not guarantee that a clear or desirable model is available.

9.6 The ensemble model is essential for coherence but limited for individual sound

A good ensemble supplies information no isolated soloist can provide: pulse, shared timing, balance between parts, harmonic function, intonation within chords, dynamic relationships, phrasing across sections and stylistic coherence. The player learns when their part should lead, colour, support, blend or recede. For these ensemble functions, the complete ensemble is itself the relevant model.

For developing precise instrumental timbre or note onset, the full ensemble is less specific. Other instruments, acoustics and overall sound level mask individual details. Someone playing exclusively in a loud ensemble may gain extensive experience producing sufficiently audible sound without ever perceiving a separate good player on the same instrument for a sustained period. Ensemble experience and instrument-specific exposure therefore complement each other; they must not be counted as the same exposure.

9.7 Why a teacher can substantially shorten learning time

A beginner does not yet hear all the differences an experienced player or teacher hears. Perception is itself part of learning. Terms such as sound core, resonance, note onset, carrying power, balance and phrase ending refer to audible qualities, but the learner must first distinguish them in actual sound. As long as a difference is not perceived, it cannot be deliberately reduced.

This explains why someone can teach themselves a brass instrument, function in an ensemble for years and still retain a typical beginner's sound. The ensemble primarily requires the correct notes to be produced at the right time and with sufficient audibility. When this succeeds, the performance is functional for participation. The player can read parts, follow entries and support the ensemble while individual sound development lags behind.

The ensemble is not a clear listening environment for that individual sound development. The player's tone mixes with those of others. Other instruments, hall acoustics and overall sound level mask part of the information. Particularly when playing is predominantly loud, differences in timbre, note onset, resonance and articulation become hard to perceive separately. The player mainly hears the shared result and much less precisely what happens within their own tone.

A beginner's sound can therefore persist for years. The player receives insufficient information about the difference between their sound and that of an advanced player on the same instrument. They notice whether a note speaks, is approximately the right pitch and is audible within the ensemble. They may not yet notice that it is broad, airy, poorly centred or lacking a core. What they cannot yet distinguish, they cannot deliberately change.

Meanwhile, the motor system does not stop learning. Each repetition stabilises the solution actually used. When the player produces notes with a broad, airy or weakly resonant tone, they gradually become more reliable at producing precisely that tone. Performance may become easier, more automatic and economical without approaching the intended sound. Efficiency therefore does not guarantee quality. The body organises a workable solution for the task as actually experienced.

An excessively large effective lip opening may play a role in a typical beginner's sound. The combination of mouthpiece pressure, air pressure, lip tension and aperture size then lies outside the narrow region in which a stable brass tone with a strong core arises. Relatively much air flows through an opening insufficiently guided by the instrument's resonance. The result may sound broad, airy and poorly centred.

The precise bodily organisation cannot, however, be read directly from sound. The teacher hears the acoustic result and can use it to investigate the playing organisation the learner employs. Sound remains the pedagogical point of entry. The target is not for the learner consciously to produce a prescribed aperture size. It is to learn to hear what sound they produce, which sound is sought and whether the next attempt approaches it.

Here lies the particular value of one-to-one teaching. The teacher hears more than whether the correct note is played. The teacher also hears differences the learner cannot yet distinguish and can temporarily direct attention to one aspect: the tone's core, the first instant of onset, the development of resonance, the transition between two notes or the way a phrase ends.

The teacher can then demonstrate a note or short phrase and have the learner answer immediately. Teacher and learner can subsequently play the same passage together. Model and learner can also alternate rapidly. Perception, performance and comparison are then only seconds apart. The difference need not be reconstructed from a distant memory, but is immediately audible.

The teacher can also reduce the task until the relevant difference does not disappear among other problems, adjusting register, dynamics, tempo and phrase length. If sound cannot be assessed adequately in a complete passage, one note, one onset or one short connection can be selected. Once the difference is audible and reproducible, the same solution can return to a larger musical context.

The teacher's main function is therefore not to prescribe bodily movements. The teacher shortens learning by selecting perceptible differences and directing the learner's attention towards them, making audible what the learner did not yet hear independently. The learner gains a more accurate target, changing the information on which the motor system selects its next solutions.

This intervention may save a great deal of time. Without a teacher, a player may repeat the same provisional solution for hundreds of hours. If the result suffices for functioning in the ensemble, that environment provides little incentive to change the solution fundamentally. One precise intervention can make audible a difference the learner might otherwise discover only months or years later.

From that moment, independent playing hours also change. The learner no longer practises exclusively with their existing beginner's production as the reference. They now know which difference to listen for and which result to approach. The teacher has therefore done more than correct something within the lesson: the direction of all subsequent practice hours has changed.

Early adjustment also prevents recovery work. A beginner's solution repeated for hundreds of hours has become not merely familiar but motorically stabilised. To change it later, the player must first recognise that the familiar solution is limited. They must then make it less automatic, find another solution and repeat the new organisation sufficiently often to make it reliable. The costs therefore consist not only of the original undirected practice hours, but also of the later time needed to replace an ingrained performance.

A relatively small number of good lesson hours can therefore substantially shorten total learning time. The teacher's importance is determined not by minutes spent speaking or demonstrating, but by the influence of perception and intervention on the much larger quantity of playing hours outside lessons. A teacher who makes one essential difference audible in time can prevent a learner from spending years further perfecting the wrong solution.

Audio and video do not fully take over this diagnostic function. They can make a high-quality performance available daily and repeatedly. The learner can listen, watch, stop, repeat and play along. A recording, however, does not independently notice which difference this particular learner cannot yet hear. It can offer the model, but cannot establish where attention needs to be directed at that moment.

Teacher and media therefore have different, complementary functions. The teacher helps the learner discover what to attend to and checks that the distinction is actually heard. Media then keep the refined model available between lessons. The learner can listen again to the same sound, onset, articulation or phrase every day and play along with it.

The greatest time saving arises when the two environments are connected. During the lesson, the relevant difference is selected, demonstrated and made audible. Outside the lesson, the same model returns during many independent playing hours. In the next lesson, the teacher can assess whether the internal model has become more accurate and motor performance has developed in the intended direction.

The teacher teaches the learner what to listen for. Media ensure that the refined model does not disappear when the lesson ends.

9.8 Comparing model environments

The following table organises environments by their main learning function. Its order is not a general quality ranking. It shows which environment supplies the clearest model for which target and where the main limitation lies.

Table 6. Each learning or maintenance target requires different model environments, audibility and materials. The table presents teaching choices, not an empirically measured universal ranking.

9.9 From one instrument to section and ensemble

The model environment changes as the scale increases, from one instrument through section and learning ensemble to top-level ensemble. This is not a fixed quality ladder for every aspect. A larger ensemble may be richer for balance and coherence, while one clear instrumental model remains more precise for note onset, timbre and direct intonation comparison. The possibility of playing along with a recording can make a separate reference available at every level.

Table 7. Moving from one instrument to a top-level ensemble changes which information is available and audible. Its value remains dependent on task and seating position.

A curriculum should therefore deliberately combine several environments. The teacher supplies a live model and teaches the learner to perceive differences. Media extend that model through the week. The section teaches shared organisation. The ensemble teaches placing individual performance within a larger musical system. Finally, independent playing without an external model checks what has actually been incorporated into the internal model and motor performance.

9.10 A learning ensemble with strong players

From the perspective of exposure to models, a learning ensemble with strong players is strongly preferred. Experienced players can offer directly audible models for timbre and intonation, as well as articulation, timing and phrasing. Shared age or an equal skill level may be socially pleasant, but does not guarantee a reliable musical model. Social similarity and model quality are different qualities.

Beginning wind players who mainly hear other beginners do not automatically receive the intended example. Conversely, adding strong players does not automatically solve the problem. They must actually be audible from the learner's seat, the same or a related part must be recognisable, and the task must permit simultaneous or alternating playing.

9.11 Seating position and conflicting intonation references

Seating is not a logistical detail: it determines the available information. A wind player does not hear their own sound, neighbouring players, section and overall ensemble in the same proportions as the conductor or an audience member. The instruction “listen to one another” therefore assumes skills that must be explicitly learned: recognising relevant parts, hearing differences, choosing a reference and knowing when to adjust or hold steady.

Consider a row of five players. Player 2 hears player 4 producing a different pitch. Which reference applies? Should player 2 adapt, should player 4 adjust, or should both follow a third reference? Merely stating that they must listen to each other does not establish who supplies the reliable reference. Conflicting pitches make both perception and the control decision ambiguous.

9.12 Practical experience and the limits of the usual tuning procedure

The author bases the urgency of this question partly on personal experience: four years of instrumental lessons and fifteen ensemble-years across three ensembles, with recurring section intonation problems and no experienced explicit instruction on reference selection when pitches conflict. This is personal experience, not a representative study.

The tuning procedure described in section 8.5—hearing poor intonation, playing an isolated G4 or C5, reading the tuner, moving the tuning slide in or out, playing again, checking the section and continuing—can correct an immediate problem. If poor intonation subsequently returns, this has not demonstrated that players have learned what to hear within the passage, who supplies the reference, and when to adjust or hold steady.

9.13 What Temple clarifies and does not resolve

Temple (2010) describes relevant conditions for ensemble listening: each player's seat provides a different perspective, musicians must monitor their own tone, section and ensemble, and the conductor must help determine whom and what to listen to. He also discusses how listening relationships connect with layout, musical function and players' strengths or weaknesses.

These suggestions reveal the geographical and relational nature of ensemble listening, but do not solve the concrete problem of reference selection. When players 2 and 4 audibly differ, “listen to the players around you” does not establish who should adjust and who should hold steady. Temple is therefore used as a source for conditions and questions, not as empirical proof that a complete teaching route is available.

9.14 Five variables that must be recorded

When exposure to models becomes a curriculum variable, simply adding up minutes is insufficient. At least five features must be recorded:

• duration: for how many minutes or hours an external model was available;
• specificity: how closely instrument, register, dynamics, style and musical task matched what the learner played;
• perceptibility: how separate, clear and visible the relevant model was;
• connection to action: whether the learner only listened, imitated afterwards, alternated or played simultaneously;
• seating and listening relationship: which player, part or recording was audible from the actual playing position and which information was masked by section, volume or layout.

These records may reveal fundamentally different learning environments for two learners with the same number of playing hours. One hundred hours playing alone, one hundred in a loud ensemble and one hundred playing along with a clear solo model are not equivalent hours. This distinction permits investigation of which combination of environment, duration and task correspondence contributes most to particular aspects of instrumental learning.

10. Two factors working together to the player's advantage

Play-alongs with sound and image bring together two different conditions for development. The first concerns the amount of playing; the second the direction of learning. They must be named separately, because more hours without a clear target model are not the same as more hours with a perceptible reference.

10.1 Play-alongs can encourage more playing hours

With a music video, the player enters a familiar musical world, follows synchronised notation and becomes a participant rather than solely a listener. The attraction of favourite music can encourage voluntary repetition. A play-along thereby increases opportunities to accumulate actual instrument-playing hours.

10.2 An available model gives those hours direction

When a suitable player on the same or a related instrument also remains audible and visible during playing, each attempt gains a concrete reference. The learner can isolate brief applications—a note onset, articulation pattern, dynamic arc or phrase—but can also continue following the model within the complete music. “Listen to a good player” thus becomes a recurring action during playing itself.

The play-along can increase playing hours. The model makes a larger share of those hours goal-directed.

11. The same media at home and in the teaching room

The greatest opportunity arises when home and teaching room use the same audio, video, notation and controls. In the teaching room, the learner learns not only to play a piece, but also to use these media: balancing model and personal volume, knowing when watching is useful and when listening alone is needed, repeating a fragment, and perceiving differences without excessive verbal analysis.

A lesson can begin with the learner's current favourite music video. Playing together breaks the ice and immediately gives the teacher information about timing, sound, range, reading behaviour and musical engagement. The teacher can then choose a smaller fragment, isolate the model and direct attention to the musical result. At the end, the application returns to the complete clip.

This creates a continuous learning loop:

MODEL → PLAY → LISTEN → COMPARE → ADJUST

The teacher remains necessary for task selection, diagnosis, quality monitoring and personal feedback. Recordings do not replace the teacher; they extend the model's availability into independent playing hours.

12. Common hardware configurations

Playing with an audible or visible model requires three sources of information: the notes the player performs, the sound of the musical model and, optionally, the image of the performing player or original music video. These sources can be distributed across paper and one or two devices. In practice, three main configurations arise.

Table 8. Three main configurations for distributing notation, sound and image.

12.1 One screen or two screens

In the first configuration, the player reads notes from paper. A phone, tablet, laptop or television supplies sound and, when video is available, the image of the model player or original music video. In the second, notes appear on a screen and sound and any image come from the same device. This is the simplest fully digital setup.

The third configuration distributes the sources across two screens. Screen 1 shows the notes. Screen 2 shows the model player or original music video and supplies the accompanying sound. Notes on screen 1 should run approximately in synchrony with music on screen 2. Perfect synchronisation is unnecessary as long as several bars remain visible. A small difference between the screens does not then immediately disrupt playing: the player can look ahead in the visible notation and continue following the music. Both sources should nevertheless begin at approximately the same musical point.

In all three configurations, sound can come from the device's built-in speakers, open-ear earphones, in-ear earphones, headphones, external speakers, a soundbar or a stereo system. Notes must be clearly readable, the model must sound sufficiently clear, and the player must continue hearing their own instrument while playing along.

12.2 Sound and audibility of the player's own performance

While playing along, the player must hear both the model and their own instrument. This is necessary to compare the performances. If the recorded model overwhelms the player's instrument, precisely the information needed to perceive differences in timing, note onset, articulation, intonation and timbre disappears.

Open-ear earphones, such as JBL Sense, keep the player's own instrument more audible than sealed in-ear earphones. The ear canal remains open, allowing the instrument to be heard directly from the room while the recorded model sounds close to the ears. In-ear earphones seal the canal. The player's instrument is consequently perceived differently, and the playback model can readily dominate.

Volume must be carefully adjusted even with open-ear earphones. The model need not rise above the instrument. It should be clear enough to follow timing, articulation, melodic line and sound direction, while the player's own tone remains separately recognisable. Built-in speakers provide the simplest solution. External stereo speakers or a good television sound system can provide a fuller, more spatial model. Headphones or in-ear earphones remain usable when speaker playback is impossible, but require extra attention to audibility of the player's instrument.

The best sound reproduction is therefore not automatically the most powerful or most acoustically closed. A suitable setup keeps model and personal instrument simultaneously and clearly audible as two comparable information sources.

12.3 Stereo as a secondary quality factor

The learning principle does not depend on stereo. The essential point is that a relevant model is available frequently enough, recognisable and clearly audible. Stereo can nevertheless refine reproduction. Brass instruments do not radiate every frequency equally strongly in every direction; the room and reflections also affect what listeners receive. Research on instrument directivity confirms that differences in spatial representation can be perceptible (Corcuera et al., 2023). A good player demonstrating live in the same room therefore remains the richest acoustic and visual reference.

When a live model is unavailable, a high-quality stereo recording through earphones, headphones or two speakers may preserve more spatial detail than one small speaker. This sharpens the auditory reference without changing its pedagogical function. Recording and reproduction should above all be clear and avoid overwhelming the learner's own instrument sound.

Model availability comes first; stereo can subsequently improve reproduction quality.

12.4 Controls at the playing position

Controls should preferably remain within reach. A phone can function as remote control and media source while image and sound run through the large screen. A laptop provides more space for selecting parts, setting starting points and combining windows. In the teaching room, the teacher can demonstrate the same controls the learner uses at home. The technical transition from lesson to independent playing then remains small.

The pedagogical core is the same in every configuration: the model must be easy to start and repeat, clearly visible and audible, and leave room to perceive one's own performance.

13. WindPlayer as a concrete implementation

WindPlayer organises exposure to models through sound, image, synchronised notation and playing along in an existing collection of play-alongs and instrument-specific models. It is therefore not a future technical concept, but an operating playing practice now being named, organised and further developed pedagogically.

13.1 Two existing forms

The playlist Rode Boekje - Cornet - Speel mee contains ten chorales with René Fennema as an instrument-specific model. The player hears first one cornet part and then the other, and can play along with both. The central action thus already exists: a good player on the same instrument remains available as a musical model outside lesson hours.

The existing YouTube channel also contains WindPlayer playlists for B♭, E♭ and C instruments. These connect parts and synchronised notation with the original music video. The second form therefore also exists: playing along with attractive existing music while image, sound and notation are simultaneously available. One application primarily addresses instrument-specific model formation; the other mainly increases the attractiveness and amount of playing. Together they form WindPlayer's practical basis.

13.2 From existing practice to a pedagogical system

The innovation therefore does not lie in claiming that this is technically possible. It lies in recognising, explaining and systematically organising an existing playing practice. The next development step is to name exposure to models as a curriculum variable, record model hours, apply high recording quality and stereo as quality criteria, connect the approach between lessons and home, and investigate learning outcomes.

For this purpose, WindPlayer connects original music videos, expert instrumental models, synchronised notation and different parts for B♭, E♭ and C instruments. Applications range from isolated model recordings to complete play-alongs with a pop video.

A pedagogically usable implementation needs at least the following design criteria:

• Model quality: performance must be musically convincing and technically reliable;
• Task correspondence: what is visibly and audibly modelled must correspond to what the learner attempts;
• Audible personal production: playback volume must not acoustically conceal the learner;
• Repeatability: fragments must be quickly restarted or slowed without losing musical coherence;
• Level differentiation: lead, supporting and backing parts must allow participation without the task constantly exceeding the player's level;
• Lesson–home continuity: the same content and controls must be introduced in lessons and immediately usable at home;
• Repertoire value: music must be attractive enough to encourage voluntary repetition.

14. Limits of evidence and risks

The argument for organised exposure to models through audio, video and play-alongs consists of several layers that must not be confused.

Table 9. The argument includes supported general learning principles, reasoned pedagogical inferences, teaching positions, practical experience and effects whose magnitude remains to be measured separately.

Important risks include an overly loud model, passive following without independent auditory attention, visual overload, imitation of irrelevant surface movements, dependence on one sound ideal, technically excessive repertoire difficulty and copyright restrictions. Simultaneous playing can also mask differences. Practice should therefore alternate between model alone, model plus learner and learner alone.

15. Research agenda

The central research question is not whether learners need a model, but which arrangement of exposure to models produces the greatest lasting learning gain for which task. An initial research programme can separate the following variables:

• Exposure duration: compare equal total playing time with, for example, 5, 20 and 40 minutes of model availability per week.
• Model timing: compare listening before playing, alternating model and player, and simultaneous playing.
• Modality: compare audio, video, video showing the player and video with synchronised notation.
• Model specificity: compare the same instrument, a related brass instrument, an ensemble model and accompaniment alone.
• Model level: compare an expert model, a coping model that corrects mistakes, and combinations of both.
• Learning outcome: measure timbre, note onset, articulation, timing, intonation, phrasing, retention and transfer separately.
• Motivation and behaviour: measure voluntary playing time, session frequency, dropout and the choice to continue playing without an assignment.
• Independence: investigate whether learners subsequently show a more accurate internal model and better self-correction without a screen.

A practical first experiment could randomly assign beginning brass players to two conditions with the same music, teacher time and total playing duration. The control group uses notation and accompaniment; the intervention group additionally uses a visible and audible model on the same instrument. Independent assessors judge blinded recordings before, immediately after and several weeks after the intervention. Digital logs record actual exposure and playing time. This makes it possible for the first time to ask not only what the learner practises, but for how many minutes the white line is actually present.

A separate recovery study compares instrument performance before and after free buzzing, mouthpiece buzzing and a control period of instrument playing. The relevant outcome is not only sound immediately after the intervention, but attempts and elapsed time until performance returns within the individual's baseline range. Measurements include onset reliability, pitch stability, spectral features of timbre, dynamic control, air use and perceived playing ease. Recovery work thereby becomes visible as a separate cost rather than disappearing within the remainder of the practice session.

15.1 Top-level ensembles as a research question

Top-level ensembles such as Soli Brass and De Wâldsang show that a high-quality shared result is attainable. That result does not yet demonstrate which approach is transferable to a learning ensemble. Research must establish which quality is brought in through selection and personnel, and which listening and adjustment skills develop through teaching and practice. No policy of reseating or replacing players is attributed here to either ensemble without investigation.

A practice-based study can document rehearsals, sectional work, seating, practice materials and the actual audibility of references. Interviews with conductors and players must be connected to observable approaches and concrete practice examples; general statements about “listening well” are insufficient.

Table 10. Factors to investigate in Soli Brass, De Wâldsang and other top-level ensembles before drawing transferable conclusions for learning ensembles.

15.2 Practice materials for conflicting references

Model recordings are available for one-to-one playing along. An in-tune section recording can demonstrate the intended shared result, but does not automatically teach what player 2 should do when player 4 deviates during joint playing. Targeted materials for such conflicting intonation references have not yet been identified in this treatment. This does not mean none exist; a focused inventory should form part of research into top-level ensembles and training practices.

Required materials must make timbre and intonation separately comparable; present the player's part alone, with a model and within the whole; teach recognition of differences; make reference choice explicit; practise adjusting and holding steady as different decisions; check with and without a model; and connect recordings from the player's seat with section or audience perspectives. Only concrete use in rehearsal and independent practice establishes whether such materials actually constitute a learning route.

16. Conclusion

Brass playing requires joint development and maintenance of sound concept, hearing differences, control of performance and the learned connections between them. A good external performance makes comparison possible, but does not control the player. Repeated listening, playing, comparison and renewed attempts establish relationships through which the intended result and the organisation available to achieve it become accessible at the moment of onset.

The fundamental blind spot is that brass pedagogy names model formation but does not organise it. It does not specify how many hours a learner hears a good player, how many occur during the learner's own playing, or how the model moves from lesson to home. A learner may consequently practise hundreds of hours with mainly their own beginner's production as auditory reference.

At the same time, scarce minutes are spent on activities that do not practise the complete learning task. A healthy learner already has sufficient breathing capacity; the learner does not need to learn to breathe, but to play. Buzzing deviates further: it changes the acoustic system and thereby organises different coordination. Its accounting consists of buzzing time, lost instrument-specific practice time and recovery work to rediscover pressure, aperture, resonance, response, note onset and sound on the instrument afterwards.

Ordinary media make a different educational priority practicable. Spotify and other audio sources keep high-quality sound examples available. YouTube and instrument-specific video add the visible performer. Play-alongs let the learner listen and act within the same time course. Two factors can consequently improve together: the learner plays more, and a greater proportion of those hours occurs with a relevant model within hearing and sight.

The model environment must be linked to the learning target. A good player on the same instrument, live and visible in the same room, supplies the richest reference for timbre, onset, articulation and phrasing, but is available for few hours. A high-quality audiovisual recording is less complete but can recur daily. A strong section teaches shared adjustment; a good ensemble supplies a model for pulse, balance, harmonic function and musical coherence. These environments are complementary and must not be counted as equivalent model hours.

WindPlayer is a concrete expression of this shift: from preparatory component activities to the complete musical task, from occasional demonstration to daily exposure to models, and from uncounted recommendations to measurable model hours. Stereo sharpens auditory representation, but is secondary. The essential point is that the white line remains present while the kilometres accumulate.

For ensemble learning, this means that a learning ensemble with strong, actually audible players is preferable, from the perspective of model exposure, to one in which beginners mainly hear other beginners. Seating, reference selection and conflicting pitches nevertheless remain separate teaching problems. The approaches of top-level ensembles and available practice materials must therefore be investigated before their results are presented as a transferable method.

Not five minutes of demonstration in the weekly lesson, but the white line every day and during playing itself.

References

Baker, D., & Green, L. (2013). Ear playing and aural development in the instrumental lesson: Results from a case-control experiment. Research Studies in Music Education, 35(2), 141-159. https://doi.org/10.1177/1321103X13508254

Berjamin, H., Lombard, B., Vergez, C., & Cottanceau, E. (2015). Time-domain numerical modeling of brass instruments including nonlinear wave propagation, viscothermal losses, and lips vibration. https://arxiv.org/abs/1511.04247

Cambridge English. (n.d.). Guided learning hours. Cambridge University Press & Assessment. https://support.cambridgeenglish.org/hc/en-gb/articles/202838506-Guided-learning-hours

LKCA. (2014/2016). Landelijk raamleerplan blaasinstrumenten (2e druk). https://www.knmo.nl/wp-content/uploads/2020/02/LKCA_RAAMLEERPLAN_BLAASINSTRUMENTEN.pdf

Corcuera, A., Chatziioannou, V., & Ahrens, J. (2023). Perceptual significance of tone-dependent directivity patterns of musical instruments. Journal of the Audio Engineering Society, 71(5), 293-302. https://doi.org/10.17743/jaes.2022.0076

Ferjan Ramírez, N., & Hippe, D. S. (2024). Estimating infants' language exposure: A comparison of random and volume sampling from daylong recordings collected in a bilingual community. Infant Behavior and Development, 75, 101943. https://doi.org/10.1016/j.infbeh.2024.101943

Green, L. (2012). Informal learning and aural learning in the instrumental music lesson. In G. E. McPherson & G. F. Welch (Eds.), The Oxford Handbook of Music Education (Vol. 2, pp. 161-196). Oxford University Press.

Han, Y., Syed Ali, S. K. B., & Ji, L. (2022). Use of observational learning to promote motor skill learning in physical education: A systematic review. International Journal of Environmental Research and Public Health, 19(16), 10109. https://doi.org/10.3390/ijerph191610109

Johnson, K. (n.d.). Listening: The primary musical skill. International Trumpet Guild Youth Masterclass. https://www.trumpetguild.org/images/ITGYouth/Masterclass/keithjohnson.pdf

Krigolson, O. E., Ferguson, T. D., Colino, F. L., & Binsted, G. (2021). Distribution of practice combined with observational learning has time dependent effects on motor skill acquisition. Perceptual and Motor Skills, 128(2), 885-899. https://doi.org/10.1177/0031512520981242

Mattéoli, R., Gilbert, J., Vergez, C., Dalmont, J.-P., Maugeais, S., Terrien, S., & Ablitzer, F. (2021). Minimal blowing pressure allowing periodic oscillations in a model of bass brass instruments. Acta Acustica, 5, 57. https://doi.org/10.1051/aacus/2021049

Meissner, H. (2017). Instrumental teachers’ instructional strategies for facilitating children’s learning of expressive music performance: An exploratory study. International Journal of Music Education, 35(1), 118-135. https://doi.org/10.1177/0255761416643850

Meissner, H. (2021). Theoretical framework for facilitating young musicians’ learning of expressive performance. Frontiers in Psychology, 11, 584171. https://doi.org/10.3389/fpsyg.2020.584171

Ste-Marie, D. M., Lelievre, N., & St. Germain, L. (2020). Revisiting the applied model for the use of observation: A review of articles spanning 2011-2018. Research Quarterly for Exercise and Sport, 91(4), 594-617. https://doi.org/10.1080/02701367.2019.1693489

Temple, M. (2010, February). Teaching students to listen. The Instrumentalist. https://theinstrumentalist.com/february-2010/teaching-students-to-listen/

Woody, R. H. (2003). Explaining expressive performance: Component cognitive skills in an aural modeling task. Journal of Research in Music Education, 51(1), 51-63. https://doi.org/10.2307/3345648

Wulf, G., Shea, C. H., & Lewthwaite, R. (2010). Motor skill learning and performance: A review of influential factors. Medical Education, 44(1), 75-84. https://doi.org/10.1111/j.1365-2923.2009.03421.x