Particle sonorization based on user-centred and universal design

Particle sonorization based on user-centred and universal design Emmanouil Chaniotakis𝑎,∗for the REINFORCE Collaboration𝑏 𝑎Ellinogermanki Agogi, 15351 Attica, Pallini, Greece 𝑏European Gravitational Observatory, Via E Amaldi, Pisa, Italy Full author list: https://reinforce.sonouno.org.ar/wp-content/uploads/sites/11/2025/ 05/REINFORCE-Collaboration.pdf E-mail: [email protected] Researchers have long highlighted the need for multi-modal approaches in teaching and learning environments using inclusive material, based on user-centred and universal design. In most cases, the sonification mapping of the dataset was defined by its creator and shared as a final product or even with some musical products, not clearly devoted to the study of the data, the identification of features or to research. Taking into account the increasing examples of sonification in astrophysics, the development of a sonification software that translates data in different formats into sound, which allows users to open different datasets, and explore them through visual and auditory display, the last permitting them to adjust visual and sound settings to enhance their perception, was done. It is an open-source application, based on a modular design, which allows users to access different datasets, and explore them through visual and auditory display, the last permitting them to adjust visual and sound settings to enhance their perception The software emerged at the early stages of the REINFORCE project, and, because of this, it was possible to add multiple new functionalities and developments, including the sonification of diverse sets of data from the demonstrator-projects. In this contribution, we present for a first time the particle sonorization, applied to two specific cases: •New Particle Search at CERN, with an innovative approach in order to be able to classify different particles using only sound. •Cosmic Muon Images, the representation with sound was based on the possibility of correlating the deposit of energy through the three layers of the detector Next, and after two training courses, we will present the results of impact and improvement of detection of data features using the multi-modal approach to the study of scientific data. 39th International Cosmic Ray Conference (ICRC2025) 15 – 24 July, 2025 CICG - International Conference Centre - Geneva, Switzerland ∗Speaker ©Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0). https://pos.sissa.it/ Particle sonorization Emmanouil Chaniotakis 1. Introduction Sonification, leveraging our highly developed sense of hearing [1,2], shows promise for data analysis. However, its use for detecting specific features in natural or scientific data is limited. Despite an increase in astrophysics [3,4], sonification isn’t widely adopted in science. Notable sonification applications include a platform for listening to ATLAS experiments in real-time [5], and software such as STRAUSS1, starSound2, Astronify3, and sonoUno4. The last stands out for its user-centered design, catering to users with and without disabilities. sonoUno aims to provide multisensory (visual and auditory) deployment of astronomical data, enabling scientific analysis. The REsearch INfrastructures FOR Citizens in Europe (REINFORCE) project 5[5,6] facilitated the web version of sonoUno and sonifications of Large Research Infrastructures in Physics6 7. This opened avenues for particle sonification, a less formally researched area. sonoUno’s development has consistently prioritized open access, cross-platform code, user-testing, and documentation. To integrate particle physics data, new sonification codes were developed for sonoUno, allowing event classification solely through sonification, moving beyond the initial column-format data. 2. The REINFORCE Initiative The REINFORCE was an original EU proposal aimed to foster citizen science by actively engaging and supporting citizens in scientific research. It was focused on bridging the gap between citizens and researchers by enabling citizens to contribute meaningfully to scientific endeavors. Through innovative approaches, REINFORCE sought to democratize science and increase public understanding and engagement with research. Within the REINFORCE collaboration, research developments and work were divided into different work teams, and the commitment made regarding sonification was to enable the demonstrators developed by the work groups linked to various experiments to be used by a wide audience, including those with visual impairments. Regarding the aforementioned demonstrators, they were based on projects developed on the Zooniverse platform, which allows the inclusion of general information and different activities so that everyone can do science. In this work, we will focus on the challenge that sonification posed in the case of particles, since when signals can be transformed into data such as mathematical functions, stored in two-column tables, and represented with simple graphic resources, the result is simple to display, both visually and in its sound format [3], but when dealing with discrete data, such as particles generated in accelerators or cosmic rays, we face a special data interpretation process and a challenge when sonifying them so that plots and sound are meaningful to the user. Throughout the process, the aim is to reproduce the visually analyzed data clearly and effectively into the sound language, without losing significance. 1Website: https://github.com/james-trayford/strauss 2Website: https://www.jeffreyhannam.com/starsound 3Website: https://astronify.readthedocs.io/en/latest/ 4Website: https://reinforce.sonouno.org.ar/ 5Home REINFORCE. (2022, 31 octubre). https://www.reinforceeu.eu/ 6CERN, https://home.cern/ 7Virgo, https://www.virgo-gw.eu/ 2 Particle sonorization Emmanouil Chaniotakis 3. New Particles Search at CERN The Large Hadron Collider (LHC) at CERN is the world’s most powerful particle accelerator. In its two 27km-long circumference and the proton beams, circulating in opposite directions, intersect at four Interaction Points (IPs).To study the products of proton collisions, four large experiments (particle detectors) are installed around the IPs. The largest general-purpose detector is called ATLAS8. The discovery of the Higgs boson in 2012 completed the last elementary constituent missing from our present description of nature and its laws, the Standard Model (SM). Despite the success of the SM certain open questions remain, and there are several theoretical models Beyond the Standard Model (BSM) that try to answer them. Some of these models predict the existence of new, long-lived particles with lifetimes of the order of picoseconds to nanoseconds, which decay to known SM particles several millimeters or centimeters away from the IP at the center of the detector. Due to their significant distance from the IP, such decay points are called Displaced Vertices (DVs) and their detection is an important subject of research for new physics. The citizen scientists were asked to help the ATLAS physicists in the search of such new physics particle. The "New Particle Search at CERN" demonstrator available on Zooniverse9, in addition to the detailed explanation about the project available on the demonstrator’s website, directs users to the HYPATIA program which presents a visual representation of the products of proton-proton collisions as registered by the detector [7]. This program allows the user to identify the particles (electrons, muons, photons, converted photons) produced in different events at the LHC. The identification is visual and based on the different signatures which the above mentioned particles leave in the detector.. It is evident from the nature of these data and their graphical representation that sonorizing them requires a different approach than the one SonUno had developed up to the time we joined this proposal. The data from which the translation into a sonorous representation must be produced cannot be found in a two-column table that can be represented in a 2D graph, and sonorizing the entire image would not provide much information to differentiate between particles. The design and development steps of the code for sonorizing particles are described in section 3.1. 3.1 LHC data Sonification To be able to discriminate between particles, originating from the proton beams that are accelerated to nearly the speed of light before making them collide, the detector features different detector subsystems: the inner detector, the calorimeters, the muon spectrometer, and the magnet systems. The ATLAS detector at the LHC uses several subsystems to identify and measure particles from high-energy proton collisions. The inner detector precisely tracks charged particles, while calorimeters (electromagnetic and hadronic) measure the energy of electrons, photons, and hadrons. The muon spectrometer specifically detects penetrating muons. Several magnet systems bend the charged particle trajectories to determine their momentum. 8https://atlas.cern 9https://www.zooniverse.org/projects/reinforce/new-particle-search-at-cern 3 Particle sonorization Emmanouil Chaniotakis Knowing the structure of the detector and how it interacts with and detects different particles, the characteristics of each will be summarized and how they are characterized visually to then relate them to the sonification process. The particles of interest for sonification in this first approach are: electron, photon, converted photon and muon, in addition to particles that do not fit among the previous ones and that are called unknown. Each of these particles describes a characteristic trajectory (Figure 1): an electron has a trajectory in the inner detector pointing a cluster in the electromagnetic calorimeter; if instead of one trajectory has two very close, it is a converted photon; on the other hand, a cluster in the electromagnetic calorimeter without a trace represent a photon; a long trajectory passing through all layers of the detector is a muon; and any other representation that not fit the above is categorized as unknown. (a) HYPATIA (b) sonoUno Figure 1: Views of one event, representing the trajectories of the particles through the detector. Knowing the structure of the detector and how it interacts with and detects different particles, the characteristics of each particle will be summarized, along with how they are visually characterized, to relate them to the sonification process. To represent this characteristic through sonification: the presence or absence of the trajectory in the inner detector will be represented by one continuous sound or silence; if the trajectory proceeds beyond this detector, the sound continues playing; if there are two close trajectories in the inner detector, two continuous sound of different frequency will be played; then, the presence or absence of the cluster in the electromagnetic calorimeter will be represented by a specific compilation of short sounds with volume related to energy. In addition, to represent the beginning of the playback and the change from one detector to another, a tick mark is sonified. This process translates the information from text files, usually seen as visual data for analysis, into sound patterns. This ensures that data from various events can be consistently and coherently represented through standardized sonification parameters. 4. Muon Tomography The application of muon tomography in geosciences and archaeology exemplifies particle physics’ interdisciplinary impact [8]. This technique utilizes atmospheric muons, which originate from extended atmospheric showers created when cosmic ray particles interact with the atmosphere. By leveraging atmospheric muons’ long penetration depths, and good knowledge of angular and energy distributions, muon tomography enables the study of geological, archaeological or other structures. By placing a detector after muons have traversed an object, and then reconstructing their 4 Particle sonorization Emmanouil Chaniotakis trajectories, it’s possible to estimate the average density of the traversed matter based on how the muons were absorbed or scattered within the object. Signal detection in muon tomography can be complicated by various background sources. To improve background understanding and enhance muon reconstruction efficiency, the Diaphane muon detector project10 has developed the "Cosmic Muon Images" REINFORCE demonstrator. This initiative records background events and engages citizens to assist in their classification, aiming for more effective background rejection. What they seek is to be able to classify between different data files, identifying which ones represent the existence of a muon and which ones do not, by determining its trajectory using three scintillators located one above the other: muons will pass through the three detector layers without deviation, allowing the muon’s trajectory to be "drawn". Although muons are particles, the way of displaying and analyzing the data is completely different from that described for the LHC, so a new approach was required for their sonification. (a) 1D view zooniverse. (b) sonoUno display. Figure 2: Views of the event 5565547_19009797 representing the trajectory of a muon. 4.1 Muography sonification The 1D representation of muography (Figure 2a) shows the ’x’ and ’y’ views of the 3D phenomenon in each column of graphs, and the rows represent the three layers of the detector (a detail to keep in mind is that the middle detector has 16 channels instead of 32 like the upper and lower ones). It was decided to sonify the 1D representation because sonification of a moving object in a 3D space does not guarantee effectiveness in this case yet, and presents several open questions. Evaluating the amount of data per detector, it was evident that there was a maximum of 32 channels. Therefore, piano notes will be used to represent each pair of channels; to match the top and bottom layers with 32 points, with the center layer with 16 points (that is, two channels per piano key in the upper and lower layers, and one channel per key in the middle layer), the sonification use a 10https://www.diaphane-muons.com/#HOME 5 Particle sonorization Emmanouil Chaniotakis piano keyboard with 16 keys sonifying the piano notes of the channels with energy deposits. The playback consist on one tickmark indicating the beginning, one sound of piano keys per detector (up, center, bottom) on the left, a tickmark indicating the pass from left column of detectors to the right; and one sound of piano keys per detector (up, center, bottom) on the right. With these premises, the sonification approach was carried out, which yielded good results. By sonifying one layer after another and mapping the channels as if they were a piano keyboard, a correlation between the different detector layers was established. In this way, the approach proved to be promising. 5. Results 5.1 LHC Sonification Following the described method, the developed script generates the sonification from a columnformatted text file provided by the LHC. The script can display the sonification with its visual complement on the screen11. Additionally, Figure 1shows the same plot for an event represented in HYPATIA and with the sonoUno script, allowing for a visual comparison between both platforms. The algorithm could recognize if the file provide one or more events. In the case of multiple events, they are separated and labeled by its name, so the sound file and plot image (optional) is generated by event and storage in the indicated folder. Therefore, it could even be used to produce the sonification (the graphic is optional) of the large quantities of events that exist in the LHC without modifications or extra work by a command-line use, which would work autonomously, generating the sounds of each event and saving them in a folder with the same name as the file that originated it and the label of each event. 5.2 Muography Sonorization The events in this case were sonified by layer from top to bottom, since this corresponds to the way the muon is detected. Figure 2b shows a case of muon existence where a line can be drawn between the energy deposits in the detector12. In addition to the sonification, in order to investigate new data access modalities, focusing on a multimodal display of data, changes in the graphic display synchronized with the auditory display were included. Thus, each point being sonified changes color, adopting a shade from red to violet corresponding to the sound frequency; that is, red for low frequency and violet for high frequency. Once the point has been sonificated, it is left black13. While the script represents a promising step forward, there’s room for improvement in how it translates the nuances of the data into sound. Future development could focus on enhancing the sound representation of energy deposits, improving the handling of complex events, and implementing controls that allow users to customize the sonification process. In this way, the script can become a more robust and versatile tool for analyzing muography data, potentially aiding in scientific discovery and data exploration. 11Video: https://youtu.be/XMaYIJkJIHg. Event of Figure 1at minute 0:29. 12Video: https://youtu.be/88nKeGJ9two 13These features could be seen in YouTube videos: https://youtu.be/MwZ0_EEuD_g 6 Particle sonorization Emmanouil Chaniotakis 6. Pilot Testing with Citizens A pilot test on muon and particle collision sonification was conducted with teachers and citizens at the REINFORCE summer school 202214. The summer school had clear objectives across four demonstrators (LHC new particles, cosmic muons, gravitational wave "noise," and neutrino detectors). These included collaborating with researchers to optimize physics infrastructures, understanding how citizen science connects research and society, and receiving training in various physics fields via citizen science. Participants also engaged in virtual visits to research infrastructures, learned to use real-world data, and helped shape modern physics curricula. Specifically for sonoUno, a cross-cutting activity for all demonstrators, participants learned to use the software, applied its functionalities to REINFORCE data, received training in feature detection in sound, and evaluated multimodal data analysis resources. The sonification workshop, structured in two sessions, introduced the sonoUno working group’s efforts in inclusion, equity, and multimodal data analysis. Activities included using the web interface and sonification scripts for REINFORCE demonstrators, highlighting the need for further research into perception and training improvement [9]. Training, based on prior experiences [6], involved detecting LHC and muography events, with participants showing improved performance in multisensory signal detection. All workshop participants found multisensory data deployment novel and expressed interest in using it in their research. The challenge of integrating sonification into existing research environments, where it’s not widely used, was acknowledged. This issue is linked to a lack of awareness about sonification techniques and a need for more perception studies to establish their credibility. The multimodal approach was highly valued for its ability to enhance understanding, with participants finding visual, auditory, and tactile data representations very useful. 6.1 Survey Results A satisfaction survey of 13 participants showed that about half were from Greece and the rest from across Europe, with varied professional backgrounds like teaching, engineering, and science. Most had not used mathematical tools for data transformations, though some used Python or MATLAB. Notably, 70% learned about multisensory data display during the workshop, and none had prior experience with sonification tools. All used Windows, and 77% could interpret sonified data; only 7% found muon sonification difficult. Participants described the program as “easy” and “effective” after some initial adjustment. Suggestions included re-listening options, soundonly sessions, and voice feedback. About 77% felt sonification could enhance their work, citing improved efficiency, accessibility, and engagement. However, half found data-to-sound conversion challenging, suggesting current tools may suit data professionals more than general users. Particle sonification was deemed more difficult than muography in terms of script execution and pattern detection. Participants preferred musical or instrumental sounds, suggesting a need to study if this preference changes after specialized training in data sonification. 14https://reinforce.ea.gr/about-summer-school/ 7 Particle sonorization Emmanouil Chaniotakis 7. Conclusions While most participants needed assistance with the tools, 69% did not use the user manual, though those who did found it positive. All participants felt the tool met expectations, calling it interesting, novel, fun, and extensible. Most found the interface intuitive (85%), underscoring the importance of user-centered development. All would recommend sonoUno to colleagues, and most would use it daily in their research, with uncertainty stemming from its novelty in their work environments. References [1] Foran, G., Cooke, J., and Hannam, J. (2022). The power of listening to your data: opening doors and enhancing discovery using sonification. Revista Mexicana de Astronomıa y Astrofısica Serie de Conferencias (RMxAC), 54, 1-8. DOI: 10.22201/ia.14052059p.2022.54.01 [2] Morishima K. et al. (2015) Discovery of a big void in Khufu’s Pyramid by observation of cosmic-ray muons." In: Nature 24647. [3] Casado, J. (2023). Investigación sobre acceso, uso y exploración efectiva de datos observacionales y bibliográficos astronómicos a partir de la sonorización. Tesis doctoral, Universidad de Mendoza, Argentina. DOI: 10.48550/arXiv.2305.05635 [4] Casado, J., García, B. (2024) A multimodal approach to data analysis in astronomy: sonouno applications in photometry and spectroscopy. RAS Techniques and Instruments; Volume 3(1): 625–635. https://doi.org/10.1093/rasti/rzae042 [5] Chaniotakis, S., Sotirious, S., Hemming, G. and Spagnuolo, F. (2023) Bridging the Gap between Large Research Infrastructures in Physics and Society through Citizen Science. Europhysics News, Vol 54, Num 2. pp 16-19. https://doi.org/10.1051/epn/2023203 [6] Angelidakis, S., Avgitas, T., Chaniotakis, E. et al. (2024) Citizen science, accessibility, art & science, critical thinking, policy and engagement: thoughts and lessons learned from the REINFORCE experience. Eur. Phys. J. Plus 139, 868). https://doi.org/10.1140/epjp/s13360024-05313-w [7] Kourkoumelis, C. and Vourakis, S. (2014) HYPATIA-an online tool for ATLAS event visualization, Phys. Educ. 49 (2014) 21 doi:10.1088/0031-9120/49/1/21 http://iopscience.iop.org/0031-9120/49/1/21/ [8] Marteau, J. de Bremond d’Ars, J. Carlus,B., Chevalier, A., Cohu,A. et al. (2022) Development of Scintillator-Based Muon Detectors for Muography. L.Oláh; H.K.M. Tanaka; D.Varga. Muography: Exploring Earth’s Subsurface with Elementary Particles, 270, American Geophysical Union - Wiley, Chapter 17 237-252. Geophysical Monograph Series, 9781119723028; 9781119722748. ff10.1002/9781119722748.ch17ff. ffinsu-03575649f [9] Oxenham, A.J. (2018). How We Hear: The Perception and Neural Coding of Sound. Annu Rev Psychol, 69,27-50. DOI: 10.1146/annurev-psych-122216-011635. 8