What Music Knows About Accessibility

When a child with chronic obstructive pulmonary disease blows into a small wooden flute, a forest on the screen begins to recover. Rain falls and puts out the fires. Wind clears the smoke. Trees come back. The child is doing pursed-lip breathing — a standard rehabilitation exercise in which the patient inhales gently through the nose and then exhales slowly through lightly pursed lips, with the exhalation lasting at least twice as long as the inhalation. The pursed lips create back-pressure in the airways that helps stop them from collapsing too early, which is one of the things that goes wrong in COPD. It is not a complicated exercise, but it is one that has to be practiced regularly at home, and adherence to that practice, when it is prescribed as an exercise, sits somewhere between sixteen and thirty-two percent. Patients say it is monotonous and they cannot tell whether they are doing it right (Valente et al., 2026).

Put the same breathing pattern inside a flute, give the flute two airflow sensors so it physically refuses to be played any other way, and wrap the whole thing in a story about an Australian lyrebird whose habitat burned down. The patient is no longer being asked to comply with a regimen; they are learning a tune. The exercise still happens, but it happens as a side-effect of trying to make music. That is a small but important shift.

This is the kind of thing that comes up repeatedly when you read research on music and digital accessibility. Music is already what software people would call multimodal — that is, it carries information in more than one channel at once. It lives in the ear, but it lives just as comfortably in the body, in the hands, in a pattern of light, in a felt vibration on the skin, in a shape on a page. Working in music tends to force designers to think about access from the inside out, because there is no single channel that is "the real one" and the others mere translations.

I want to walk through what some of that work looks like — across hearing, sight, motor function, cognition, and mental health — and ask, quietly, what the rest of accessibility might pick up from it. I will explain the technical vocabulary as I go, because the field uses several specialist terms that obscure how simple the underlying ideas often are.

The thing about music is that nobody actually hears it the same way

Start with Deaf and hard-of-hearing musicians. The dominant research framing for many years has been that DHH people are music consumers who need access to the audio. The job of accessibility, on that view, was to translate a sound experience into something that could be received some other way — captions for lyrics, perhaps a vibrating cushion, perhaps a coloured visualiser bouncing on a screen. That is not wrong; it is incomplete.

Cavdir and colleagues, working with CymaSpace — a Deaf-owned music institution in Portland, Oregon — spent fifteen months on what is called a group autoethnography, which is a research method where the team itself is the subject and members systematically reflect on their own experiences. Their conclusion is that DHH musicians are not just consumers; they are technology developers (often building the assistive tools themselves), studio hackers (repurposing commercial gear that was never designed for them), and performers who treat visibility, vibration, and visual rhythm as primary musical material rather than substitutes for sound. The authors call this sonic agency: the right to influence sound meaningfully through whatever channel the musician's body actually uses (Cavdir et al., 2025).

Once you take that framing seriously, the design space opens up.

μCap (the name is short for "music captions") tackles a small but telling problem: in conventional television and film captioning, vocal songs get their lyrics displayed, but instrumental music is usually dismissed with a tag like "[instrumental music]." Ahn and colleagues built a system that automatically writes time-aligned captions for instrumental passages using sound-mimetic syllables — strings like "Tta-da-ding" or "dun-dun tung," similar to the scat singing that vocal jazz uses to imitate instrumental phrases. They map loudness to font size and pitch to vertical position on the line, so a quiet low note appears as small text near the bottom and a loud high note as large text near the top. In their evaluation, DHH viewers preferred captions over sign-language interpretation by a wide margin, and the system significantly outperformed the obvious rule-based heuristics for generating such captions (Ahn et al., 2026).

FAME goes considerably further. Yoo and colleagues argue that captions plus an abstract spectrum visualiser still leave most of a song's musical content out of reach. So they built a photorealistic facial avatar — modelled on real vocal production — that visually performs the song: the face's expression carries emotion, the mouth carries lyrics, head movement carries rhythm, and so on. In a comparison study, DHH viewers correctly identified what music was playing 95.8% of the time using FAME versus 66.7% using a more conventional spectrum visualiser, with the largest gains on lyrical songs where abstract visualisations tend to obscure the words (Yoo et al., 2026).

Choi and colleagues looked at a different kind of access. Music psychotherapy is a clinical practice in which therapists use songwriting, improvisation, and reflective listening to help patients work through emotional rather than auditory issues, but it has historically excluded DHH clients because the therapy models themselves are auditory-centric. Choi's team co-designed a generative-AI music psychotherapy tool with licensed Korean music therapists, using text-based interaction and visual metaphors so DHH users could write songs about their experiences. Twenty-three participants completed full songwriting sessions averaging forty minutes, with high reported self-disclosure and music satisfaction (Choi et al., 2026). "Access to music," it turns out, is not a single thing — it includes access to the uses music gets put to, including therapy.

For cochlear implant (CI) users, the situation is different again. A cochlear implant bypasses the damaged parts of the inner ear and stimulates the auditory nerve directly with electrical pulses, allowing many users to hear and understand speech. Music, however, is more demanding than speech: pitch and timbre carry much of the emotional content, and CI signals reproduce those poorly. Kim and colleagues asked whether CI users might be helped not by replacing the audio with something else (the conventional "sensory substitution" route, where sound becomes vibration or vision) but by giving them controls to retune the music until it carries the feeling they want. Their interface offers emotion-oriented sliders — bright/warm, calm/energetic — rather than the raw equaliser and compressor knobs an audio engineer would use. The acoustic-feature analysis underlying the design showed that, for CI listeners, spectral-centroid variability (a measure of how much the brightness of a sound changes over time) and tempo combined with overall loudness predict emotional response, and these features are not the same as the predictors for normal-hearing listeners. The implication is practical: a CI user cannot simply borrow a sighted producer's EQ; the controls need to map to the features the CI signal actually preserves (Kim et al., 2026).

The hardware story matches the perception story. Petry and colleagues built MuSS-Bits — small wearable units that take an audio signal and convert it directly into vibrations on the skin (an approach called vibrotactile feedback). Critically, MuSS-Bits connect on the fly to any audio source: an instrument, a phone, an environmental sound. A Deaf student can pick up a guitar they have never seen before and immediately feel what they are producing, which is what most music learning actually requires (Petry et al., 2016). Iijima and colleagues ran what is called a gesture elicitation study, in which DHH participants are shown a target action ("play this percussion instrument," "bend a note up") and asked to invent the gesture they would naturally use to produce it. Across eleven participants and ten instruments, they collected 110 gestures and produced a usable taxonomy that does not assume hearing-centric mappings (Iijima et al., 2022).

Twenty years before any of the above, Hiraga and Kato ran a quiet little study comparing how hearing-impaired and hearing participants recognised emotions in drum performances paired with various visual stimuli — drawings, abstract motion, and so on. The difference between the groups was statistically insignificant. Hearing-impaired participants read drawings of musical emotion just as accurately as hearing participants did (Hiraga & Kato, 2006). Music's emotional content is not locked in the ear. It never was.

And then there is sight

Music notation — the dots and lines on a page that tell a musician what to play — is famously visual. It is a graphical, two-dimensional language with detail at one zoom level (which note, which fingering) and structure at another (the shape of the phrase, the architecture of the piece), and reading it well requires moving fluidly between those scales. For most blind musicians, the historical alternatives have been braille music, a tactile notation system in which braille cells encode notes, durations, and articulations, and audio learning, where the musician memorises a piece by ear. Braille music is excellent but cannot easily be exported to sighted collaborators, and the few specialty production tools for it tend to lag years behind mainstream music software.

SoundCells, from a team at NYU and the Filomen M. D'Agostino Greenberg Music School, takes a different angle. It uses ABC notation — a plain-text syntax in which a piece of music is written as ASCII characters, like |: G3 GAB | c3 BAG :| — as input, and simultaneously renders three outputs: audio playback, a visual print score, and braille music. Because the underlying representation is text, the whole interface is naturally compatible with screen readers and braille displays. The two expert blind musicians who co-designed the system were composing original music within minutes of seeing it. In a six-week study with six blind and visually impaired musicians, every participant composed a piece, and the pieces were performed live by a professional musician (Payne et al., 2022a; Payne et al., 2022b). The lesson generalises: in music as in HTML, a text-based representation is more accessible than a graphics-based one, because it is easier for assistive technologies to interrogate.

For low-vision musicians — a population that, between sighted print readers and braille music readers, may actually be larger than either, but has been studied less — Payne and An interviewed sixteen low-vision musicians and two teachers across four countries. Their finding is that no single notation solution fits everyone; accessibility depends on the interaction between the musician's vision, their instrument, the genre, and the venue. Most participants used mainstream sheet-music apps on tablets (forScore, in particular), with pinch-zoom, contrast adjustment, and colour annotation as their everyday working tools. They distinguish large-print music (the standard score made physically bigger) from modified-stave notation (the score actually redrawn with thicker lines and clearer spacing), and note that real modified-stave notation is extremely laborious to produce (Payne & An, 2025). Anken and colleagues, meanwhile, asked what happens if you put the score in extended reality, the umbrella term covering virtual reality and augmented reality together. Their prototype shows a customisable horizontal band of music floating in front of the player, with adjustable size, position, colour, contrast, and continuous scrolling — so a low-vision musician can read the score and play simultaneously rather than memorising small sections one at a time (Anken et al., 2025).

For blind music producers working in digital audio workstations (DAWs — software like Pro Tools, Logic, Reaper, or Ableton Live, the modern equivalent of a recording studio in software), there is a particular problem. A screen reader reads interface text aloud, in speech, through the same audio channel as the music being mixed. The two collide. Karpodini and colleagues asked whether vibrotactile patterns — short structured vibrations on a wearable device, sometimes called Tactons — could replace the screen reader for conveying equaliser settings (an equaliser, or EQ, is the part of audio software that lets you boost or cut specific frequency ranges to shape the tone of a sound). All three of their conditions — screen reader, screen reader plus haptics, haptics alone — hit over 94% accuracy on identifying EQ settings. But the haptics-only condition reached an "Excellent" usability score of 90.28, and crucially, it did not collide with the music in the audio channel (Karpodini et al., 2025; Karpodini, 2022).

Lu and colleagues spent years interviewing forty BLV (blind and low-vision) musicians and music teachers, and then ran co-design workshops with ten teachers and learners, including four blind, four low-vision, and two deafblind participants. Their findings are practical and humane. Vibrotactile alerts work well for "your turn," "start," "stop," and "you are flat"; intensity and pattern can carry musical concepts like dynamics (loudness changes) and articulation (how notes are connected or separated); and material aesthetics — how the device feels in the hand, whether it is breathable, whether it dampens unwanted sound, whether it bends the way the body expects — matter a great deal, because devices that feel pleasant get worn longer (Lu, 2022; Lu et al., 2023a; Lu et al., 2023b).

The adaptation story matters too. Payne and colleagues interviewed eleven blind composers, producers, and songwriters and found extraordinary ingenuity — bending tools like Finale, Lime, Pro Tools, and Sonar to work for them — combined with three persistent problems. Accessible music software is severely limited (their review identified only three accessible digital musical instruments, out of eighty-three reviewed, that target visual impairment). Sighted assistance is still required at critical stages, particularly when a composer needs to deliver a visually formatted score. And third-party accessibility scripts — small patches written by community members to expose hidden interface elements to screen readers — break almost every time the underlying software gets updated (Payne et al., 2020). Lucas and colleagues interviewed twenty visually impaired sound creatives across ten countries and framed the gaps as educational, informational, and tooling, drawing on Ivan Illich's idea of convivial tools — tools that ordinary people can master and use to express themselves, as opposed to opaque industrial systems that require specialists. Their argument is that accessibility is not just a property of the software itself; it is a property of the whole ecosystem of tutorials, manuals, online communities, and teacher knowledge that surrounds it (Lucas et al., 2025).

The systematic literature review of this space is worth pausing on. Zhang and colleagues mapped fifty-four papers on music technology for blind and low-vision people, following the PRISMA methodology (a structured approach that documents which papers were included or excluded and why). They identified four cross-cutting design insights: spatial awareness can be carried by audio-spatial mapping and haptics; alternative representations (sonification, haptic notation, verbal description) can replace visual notation when designed thoughtfully; embodied physical engagement supports memory and skill retention; and multi-sensory collaboration enables genuine inclusion rather than parallel participation (Zhang et al., 2025).

Notice how those four insights generalise. They are not really about music.

Music can also do work that other things cannot

A well-known and slightly unsettling clinical observation: people in the late stages of dementia, who may no longer recognise their own family members, will often sing along to songs they learned in their twenties. Music finds something in the brain that other inputs do not. Seymour and colleagues built AMI (Adaptable Music Interface), a tangible, modular music player whose physical controls — knobs, sliders, rocker switches, buttons — can be swapped in and out by a caregiver as the person's abilities change. Across feedback sessions with ten people with dementia, eight unique configurations emerged from eight completed devices. The opposite of a one-size-fits-all interface; an interface that is supposed to keep changing (Seymour et al., 2017).

For autistic children and adults, several research groups have used music as a vehicle for communication that does not depend on speech. McGowan and colleagues built CymaSense, an interactive system based on the visual phenomenon of Cymatics — the patterns that sound vibrations create in physical media like water, sand, or salt on a metal plate. CymaSense generates real-time 3D Cymatic shapes from sound input: amplitude controls shape size, pitch selects from twelve different forms, and timbre affects surface quality. In a twelve-week study with eight autistic adults, both musical and non-musical communicative behaviours rose significantly during the intervention phases (McGowan et al., 2017; McGowan et al., 2021). Ragone built OSMoSIS, which uses a Microsoft Kinect motion sensor to map body positions to instrument sounds — guitar, marimba, voice — without anyone having to touch an instrument. Eleven autistic children showed marked engagement increases when the sonification (the conversion of movement into sound) was switched on (Ragone, 2020). Lobo and colleagues' CHIMELIGHT augments handchimes — small handheld bells used in music therapy — with embedded sensors and visual feedback, so therapy with children with neurodevelopmental disorders gets both objective performance data and immediate visual reward in the same physical object the children already trust (Lobo et al., 2019).

Ostiz-Blanco and colleagues used music in two surprising directions. ACMUS, a multimedia music tool, ran nine half-hour sessions with twelve adults living with severe psychiatric conditions in residential care (mostly schizophrenia or related diagnoses, average duration since diagnosis 33.6 years). Scores on the Comprehensive Occupational Therapy Evaluation, which measures cognitive, social, and manipulative skills, improved significantly between the first and last sessions (Ostiz-Blanco et al., 2018b). The same research group, separately, used music as a screening mechanism. There is established evidence that dyslexia is associated not only with reading difficulty but with auditory and temporal processing differences — children with dyslexia tend to be less accurate when they tap in time to a beat. So a serious game built around rhythm and auditory perception can pick out at-risk children before they reach reading age and start losing confidence in school (Rauschenberger, 2016; Ostiz-Blanco et al., 2018a). Del Sette and Saitis, working with people living with chronic pain conditions like fibromyalgia and non-specific low back pain, found that music's value in self-management came not from "calming genres" or prescribed playlists but from self-selection and change: participants moved through songs and audio media as their mood shifted, and repetition reduced effectiveness. Music's role here was less to soothe than to externalise — to make a private, easily-disbelieved pain experience visible (Del Sette & Saitis, 2026).

The thread running through these is that music functions, in design terms, as contingent feedback at human emotional resolution. That is, an action produces a response immediately, and the response carries enough emotional texture to feel meaningful. That combination is rare in software, and useful in places that are not obviously about music at all.

And music remains, quietly, a thing people want to make

Here is the part that the wider accessibility field still under-appreciates. The Flote, built at Stanford in 2008, is a digital wind instrument played with breath and head tilt. Its hardware is a headset with breath sensors and head-tracking; its software calibrates per user, so that a person with very limited head range still has access to the full musical scale (Aziz et al., 2008). Brulé's open-source head-tracking music software was co-designed with a single man, M., who lives in a French nursing home with severe motor, dexterity, and speech impairments; the system reads an infrared dot stuck to his forehead and gives him both solo and ensemble play (Brulé, 2016). Anderson and Smith's E-Scape, from the Drake Music Project in 1996, supported single-switch composition with auditory scanning — an interaction pattern in which the system reads each option aloud in turn and the user activates a single switch when the desired option is announced. E-Scape predates most modern accessibility frameworks and still articulates principles — multiple input pathways, structured composition activities, constrained option sets — that current tools rediscover (Anderson & Smith, 1996).

Sporka and colleagues' interviews with three motor-disabled musicians — including Peter, a composer with ALS who described his assistive-technology needs as "a moving target" because his condition continuously changes — show how professional creative software treats motor accessibility as an afterthought, even though adaptive consumer instruments like Skoog (a soft, squeezable cube that produces musical notes when pressed) exist (Sporka et al., 2013). Omori and Yairi built a tabletop tangible composition interface inspired by the Reactable — a well-known electronic music instrument where physical pucks placed on a glass surface generate and modulate sound. In their adaptation, all the tangible objects are connected by physical chains a blind user can trace with their fingers, encoding timing and pitch by position. In a six-pair study, blind participants led collaborations rather than following sighted partners, which is unusual enough in mixed-ability work to be worth noticing (Omori & Yairi, 2013).

Then there is the work on Accessible Digital Musical Instruments (ADMIs). A digital musical instrument, in this field, is any electronic instrument where the input device (a controller, a sensor, a movement) is decoupled from the sound generator — the physical action does not produce sound directly, but rather sends signals that software turns into notes. That decoupling is what makes accessibility possible, because the mapping between movement and sound can be re-tuned to whatever the player's body can do.

Ilsar and Kenning take an existing professional instrument, the gestural AirSticks (controllers held in the hands and moved through the air to produce sound), and adapt it for inclusive use mostly through mapping changes rather than rebuilding the hardware. Their case studies include a teenager with leukodystrophy (a progressive neurological condition) who performed publicly at a music festival after his AirStick mappings were tuned to his movement capabilities (Ilsar & Kenning, 2020; Ilsar et al., 2022). Trolland and colleagues went further. They worked with the artist Mel — a painter — to build a personalised instrument that uses a paintbrush as the gesture sensor (because she has more reliable control gripping a brush than wearing a wrist sensor) and her own digitised paintings as the visual material the audience sees. They argue, drawing on Amartya Sen's Capability Approach — a framework from welfare economics that defines wellbeing in terms of the opportunities a person actually has reason to value — that designing deeply for one person produces transferable knowledge that designing shallowly for everyone does not (Trolland et al., 2026).

Förster and Lepa surveyed 745 special-needs schools across Germany about digital musical instrument use. DMIs of any kind appear in only 42% of those schools, and purpose-built ADMIs in barely any of them. The reason is not teacher resistance: 84% of teachers reported they had simply never been trained on the instruments during their music education. The bottleneck, in other words, is not technology but teacher preparation (Förster & Lepa, 2023).

Game accessibility belongs in this story too, because rhythm-action games are essentially musical interfaces. Yuan and Folmer's Blind Hero (2008) replaced Guitar Hero's scrolling visual cues with vibration motors on the fingertips of the left hand, one motor per coloured button. Visually impaired players actually outperformed sighted blindfolded controls, possibly because they already practise "feel-then-act" patterns in everyday life (reading braille, navigating by touch) while sighted players are trained from childhood to "visualise-then-act" (Yuan & Folmer, 2008). Allman and colleagues' Rock Vibe extended the same idea to the drumming part of Rock Band, with motors on the upper and lower arms (mapped to four drum colours) and one on the ankle for the kick pedal (Allman et al., 2009). Miller, Parecki, and Douglas's Finger Dance went the opposite direction. Rather than retrofitting an inherently visual game with sound, they designed an audio rhythm-action game from the audio out, since the underlying gameplay — matching key presses to musical beats — is fundamentally auditory anyway (Miller et al., 2007). And Payne and colleagues' Non-Visual Beats — the accessible redesign of NYU's Groove Pizza, a popular educational drum step-sequencer — uses the structural fact that even-numbered beats sit "off the pulse" in Western music to keep keyboard mappings musically coherent for screen-reader users, mapping odd beats to one row of keys and even beats to another (Payne et al., 2019).

What Sara Bly figured out in 1982

In a small CHI paper from Lawrence Livermore National Laboratory in 1982, Sara Bly asked whether a computer could present multivariate data — data with many simultaneous dimensions — in sound well enough for an analyst to use. Her experiment took six-dimensional data sets and rendered them as tones, mapping each dimension of the data to a different audio parameter (pitch, loudness, duration, timbre, and so on). With six-dimensional samples, seven listeners hit 92% correct on a strong translation difference, 70% on a moderate one, and 60% on a correlation difference. In a follow-up with seventy-five subjects split into sound-only, graphics-only, and combined-presentation conditions, sound-only was competitive with graphics-only, and combined sound-plus-graphics did best of all (Bly, 1982).

Forty-four years later, every screen-reader audio chart, every sonified line graph in iOS VoiceOver and the Highcharts library, every accessible data dashboard rests on the proposition Bly empirically tested. What she understood, and what every paper in this article reaffirms in its own way, is that sound carries information humans can use. We just got into the habit of putting almost all of it on a screen.

The quiet lesson

If you read enough of this work in one sitting, a few patterns settle out, and they are not really about music.

The first is that the modalities mainstream computing treats as primary — sight for screens, hearing for speech, touch for input — are one arrangement of channels, not the only one. Vibrotactile patterns can carry an equaliser setting (Karpodini et al., 2025). A facial avatar can carry a melody (Yoo et al., 2026). A drawing can carry an emotion as accurately to a hearing-impaired viewer as the original sound carries it to a hearing one (Hiraga & Kato, 2006). A breath can carry a song, and a song can carry the slow recovery of a forest (Valente et al., 2026). When we design as though our default arrangement is the only viable one, we end up doing translation work where we could have done original work.

The second is that accessibility is rarely just a property of the artifact. It tends to be a property of the surrounding ecosystem — the tutorials, the teacher training, the third-party scripts, the documentation, the institutional culture. That is what Lucas and colleagues mean by convivial tools (2025), and what Förster and Lepa quietly demonstrate when they show that the bottleneck for accessible music education in Germany is not equipment but teacher preparation (2023). Building an accessible product is necessary but not sufficient.

The third is something the music research is unusually clear about, possibly because no-one picks up an instrument unless they want to. People do not engage with accessibility tools because the tools are accessible. They engage because the tools let them do something they wanted to do anyway — compose a piece, play a game with a friend, breathe well enough to climb the stairs, sing a half-remembered song from when they were nineteen. Accessibility is not a substitute for desire; it is a way of removing what was in its way.

We did not need music to teach us this. But it has been teaching us anyway, since at least 1982.

---

References

Ahn, S., Baek, I., Kim, K., Truong, K. N., & Hong, J. (2026). μCap: Instrumental Music Captions for Deaf and Hard-of-Hearing Individuals. Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems. https://doi.org/10.1145/3772318.3790729

Allman, T., Dhillon, R. K., Landau, M. A. E., & Kurniawan, S. H. (2009). Rock Vibe: Rock Band® Computer Games for People with No or Limited Vision. ASSETS '09. https://doi.org/10.1145/1639642.1639653

Anderson, T., & Smith, C. (1996). "Composability": Widening Participation in Music Making for People with Disabilities via Music Software and Controller Solutions. Assets '96. https://doi.org/10.1145/228347.228365

Anken, J., Blaess, D., & Müller, K. (2025). XRMusic4VIP: Enabling Simultaneous Sheet Music Reading and Playing for Visually Impaired Musicians through Extended Reality. ASSETS 2025. https://doi.org/10.1145/3663547.3746339

Aziz, A. D., Warren, C., Bursk, H., & Follmer, S. (2008). The Flote: An Instrument for People with Limited Mobility. Assets '08. https://doi.org/10.1145/1414471.1414545

Bly, S. (1982). Presenting Information in Sound. CHI '82. https://doi.org/10.1145/800049.801814

Brulé, E. (2016). Playing Music with the Head. ASSETS '16. https://doi.org/10.1145/2982142.2982146

Cavdir, D., Simone, D., de Bastion, M., Trail, S., & Hergert, N. (2025). Sonic Agency: A Group Autoethnography of Technology-mediated Performance Practice by Deaf and Hard of Hearing Musicians. ASSETS 2025. https://doi.org/10.1145/3663547.3746396

Choi, Y., Moon, J., Yoo, J., Kim, J. G., & Hong, J. (2026). Designing a Generative AI-Assisted Music Psychotherapy Tool for Deaf and Hard-of-Hearing Individuals. CHI '26. https://doi.org/10.1145/3772318.3791385

Del Sette, B. M., & Saitis, C. (2026). Co-Ideating a Digital Pain Companion: Exploring How Music Can Support Daily Chronic Pain Self-Management. ACM Transactions on Computing for Healthcare. https://doi.org/10.1145/3807781

Förster, A., & Lepa, S. (2023). Digital Musical Instruments in Special Educational Needs Schools. ACM Transactions on Accessible Computing. https://doi.org/10.1145/3616015

Hiraga, R., & Kato, N. (2006). Understanding Emotion through Multimedia: Comparison between Hearing-Impaired People and People with Hearing Abilities. Assets '06. https://doi.org/10.1145/1168987.1169012

Iijima, R., Shitara, A., & Ochiai, Y. (2022). Designing Gestures for Digital Musical Instruments: Gesture Elicitation Study with Deaf and Hard of Hearing People. ASSETS '22. https://doi.org/10.1145/3517428.3544828

Ilsar, A., & Kenning, G. (2020). Inclusive improvisation through sound and movement mapping: from DMI to ADMI. ASSETS 2020. https://doi.org/10.1145/3373625.3416988

Ilsar, A., Kenning, G., Trolland, S., & Frame, C. (2022). Inclusive Improvisation: Exploring the Line between Listening and Playing Music. ACM Transactions on Accessible Computing. https://doi.org/10.1145/3506856

Karpodini, C. (2022). Evaluating Haptic Technology in Accessibility of Digital Audio Workstations for Visual Impaired Creatives. ASSETS '22. https://doi.org/10.1145/3517428.3550414

Karpodini, C., Michailidis, T., Creed, C., Stockman, T., & Williams, I. (2025). Vibration Or Voice? Enhancing EQ Information Accessibility for Blind Music Producers with Haptics. ASSETS 2025. https://doi.org/10.1145/3663547.3746374

Kim, H., Song, T., & Lee, K. M. (2026). Turning the Knobs of Musical Emotion: Designing Emotion-Oriented Audio Control Interface for Cochlear Implant Users. CHI EA '26. https://doi.org/10.1145/3772363.3798345

Lobo, J., Matsuda, S., Futamata, I., Sakuta, R., & Suzuki, K. (2019). CHIMELIGHT: Augmenting Instruments in Interactive Music Therapy for Children with Neurodevelopmental Disorders. ASSETS 2019. https://doi.org/10.1145/3308561.3353784

Lu, L. (2022). Learning Music Blind: Understanding the Application of Technology to Support BLV Music Learning. ASSETS '22. https://doi.org/10.1145/3517428.3550413

Lu, L., Cochrane, K. A., Kang, J., & Girouard, A. (2023a). "Why are there so many steps?": Improving Access to Blind and Low Vision Music Learning. ACM Transactions on Accessible Computing. https://doi.org/10.1145/3615663

Lu, L., Kang, J., Crispin, C., & Girouard, A. (2023b). Playing with Feeling: Exploring Vibrotactile Feedback and Aesthetic Experiences for Developing Haptic Wearables for Blind and Low Vision Music Learning. ASSETS 2023. https://doi.org/10.1145/3597638.3608397

Lucas, A., Cunningham, J., Harrison, J., Schroeder, F., & McPherson, A. (2025). The Qualities of Convivial Tools and Their Relevance to Music Software Accessibility. ACM Transactions on Accessible Computing. https://doi.org/10.1145/3733606

McGowan, J., Leplâtre, G., & McGregor, I. (2017). CymaSense: A Novel Audio-Visual Therapeutic Tool for People on the Autism Spectrum. ASSETS '17. https://doi.org/10.1145/3132525.3132539

McGowan, J. J., McGregor, I., & Leplâtre, G. (2021). Evaluation of the Use of Real-time 3D Graphics to Augment Therapeutic Music Sessions for Young People on the Autism Spectrum. ACM Transactions on Accessible Computing. https://doi.org/10.1145/3445032

Miller, D., Parecki, A., & Douglas, S. A. (2007). Finger Dance: A Sound Game for Blind People. Assets '07. https://doi.org/10.1145/1296843.1296898

Omori, S., & Yairi, I. E. (2013). Collaborative Music Application for Visually Impaired People with Tangible Objects on Table. ASSETS 2013. https://doi.org/10.1145/2513383.2513403

Ostiz-Blanco, M., Lallier, M., Grau, S., Rello, L., Bigham, J. P., & Carreiras, M. (2018a). Jellys: Towards a Videogame that Trains Rhythm and Visual Attention for Dyslexia. ASSETS 2018. https://doi.org/10.1145/3234695.3241028

Ostiz-Blanco, M., Pina, A., Lizaso, M., Astraín, J. J., & Arrondo, G. (2018b). Using the Musical Multimedia Tool ACMUS with People with Severe Mental Disorders: A Pilot Study. ASSETS 2018. https://doi.org/10.1145/3234695.3241016

Payne, W., Ahmed, F., Gardell, M., DuBois, R. L., & Hurst, A. (2022a). SoundCells: Designing a Browser-Based Music Technology for Braille and Print Notation. W4A '22. https://doi.org/10.1145/3493612.3520462

Payne, W. C., Ahmed, F., Zachor, M., Gardell, M., Huey, I., Hurst, A., & DuBois, R. L. (2022b). Empowering Blind Musicians to Compose and Notate Music with SoundCells. ASSETS '22. https://doi.org/10.1145/3517428.3544825

Payne, W. C., & An, Y. L. (2025). Access Beyond the Score: Understanding Notation Needs and Workflows of Low Vision Musicians. ASSETS 2025. https://doi.org/10.1145/3663547.3746322

Payne, W., Xu, A., Hurst, A., & Ruthmann, S. A. (2019). Non-Visual Beats: Redesigning the Groove Pizza. ASSETS 2019. https://doi.org/10.1145/3308561.3354590

Payne, W. C., Xu, A. Y., Ahmed, F., Ye, L., & Hurst, A. (2020). How Blind and Visually Impaired Composers, Producers, and Songwriters Leverage and Adapt Music Technology. ASSETS '20. https://doi.org/10.1145/3373625.3417002

Petry, B., Illandara, T., Forero, J. P., & Nanayakkara, S. (2016). Ad-Hoc Access to Musical Sound for Deaf Individuals. ASSETS '16. https://doi.org/10.1145/2982142.2982213

Ragone, G. (2020). Designing Embodied Musical Interaction for Children with Autism. ASSETS 2020. https://doi.org/10.1145/3373625.3417077

Rauschenberger, M. (2016). DysMusic: detecting dyslexia by web-based games with music elements. W4A '16. https://doi.org/10.1145/2899475.2899503

Seymour, P. F., Matejka, J., Foulds, G., Petelycky, I., & Anderson, F. (2017). AMI: An Adaptable Music Interface to Support the Varying Needs of People with Dementia. ASSETS '17. https://doi.org/10.1145/3132525.3132557

Sporka, A. J., Carson, B. L., Nauert, P., & Kurniawan, S. H. (2013). Toward Accessible Technology for Music Composers and Producers with Motor Disabilities. ASSETS 2013. https://doi.org/10.1145/2513383.2513412

Trolland, S., Smith, M., Ilsar, A., & McCormack, J. (2026). Movement as Medium: Personalisation of Instruments for Inclusive Creative Expression in Disability-Led Performance. CHI '26. https://doi.org/10.1145/3772318.3791955

Valente, A., Cotto, C., Billinghurst, M., & Gupta, K. (2026). LyreFlute: Sensor-Based Musical Experience for COPD Rehabilitation. CHI EA '26. https://doi.org/10.1145/3772363.3799123

Yoo, S., Pan, Y., Thomas, A. A., Singh, K., & Truong, K. N. (2026). FAME: Exploring Expressive Facial Avatars for Lyrical and Non-Lyrical Music Visualization for d/Deaf Individuals. CHI '26. https://doi.org/10.1145/3772318.3790402

Yuan, B., & Folmer, E. (2008). Blind Hero: Enabling Guitar Hero for the Visually Impaired. Assets '08. https://doi.org/10.1145/1414471.1414503

Zhang, S., Masu, R., Bettega, M., & Fan, M. (2025). Accessibility and Social Inclusivity: A Literature Review of Music Technology for Blind and Low Vision People. ASSETS 2025. https://doi.org/10.1145/3663547.3746466