AES Europe 2025

In auditory spatial perception, horizontal sound image localization and a sense of spaciousness are based on level and time differences between the left and right ears as cues, and the degree of correlation between the left and right signals is thought to contribute to the sense of horizontal spaciousness, in particular [Hidaka1995, Zotter2013]. For the vertical image spread (VIS), spectral cues are necessary. The change in VIS due to the degree of correlation between the vertical and horizontal signals depends on the frequency response [Gribben2018]. This paper investigated the influence of different correlation values between the top and middle layers of loudspeaker signals within a 3D audio reproduction system on listening impressions through two experiments. The results of experiments using pink noise with different correlation values for the top and middle layers show that the lower the vertical correlation values are, the wider the listening range is, where the impression does not change from the central listening position. From the results of experiments using impulse responses obtained by setting up microphones in an actual concert hall, a tendency to perceive a sense of spaciousness at the off-center listening position was found when cardioid microphones were used for the top layer that were spaced apart from the middle layer. The polar pattern and height of the microphones may have resulted in lower correlation values in the vertical direction, thus widening the listening range of consistent spatial impression outside of the central listening position (i.e., “sweet spot”.)

This paper proposes the method that plane wave field creation with spherical harmonics for a non-spherical array. In sound field control, there are physics-acoustic models and psycho-acoustic models. Some former are allowed in the location of each loudspeaker, but the sound have the differences between the auditory and the reproduction sound because phantom sources are constructed. The latter developed with wave equation under circle or spherical array conditions which are located strictly, and with high order Ambisonics (HOA) based on spherical harmonics which express only a single point. Therefore, we consider requiring the method which physically creates actual waveforms and provides flexibility in the shape of the loudspeaker array. In this paper, we focus on the Lamé function, changing its order as well as the shape of spatial figures, and propose formulating the distance between the center and each loudspeaker using the function in a polar expression. As the simulation experiment, in the inscribed region, the proposed plane wave can create the same waveform as the spherical one under high order Lamé function which is close to rectangular shape.

This paper presents a recursive solution to the Broadband Acoustic Contrast Control with Pressure Matching (BACC-PM) algorithm, designed to optimize sound zones systems efficiently in the time domain. Traditional frequency-domain algorithms, while computationally less demanding, often result in non-causal filters with increased pre-ringing, making time-domain approaches preferable for certain applications. However, time-domain solutions typically suffer from high computational costs as a result of the inversion of large convolution matrices.
To address these challenges, this study introduces a method based on gradient descent and conjugate gradient descent techniques. By exploiting recursive calculations, the proposed approach significantly reduces computational time compared to direct inversion.
Theoretical foundations, simulation setups, and performance metrics are detailed, showcasing the efficiency of the algorithm in achieving high acoustic contrast and low reproduction errors with reduced computational effort. Simulations in a controlled environment demonstrate the advantages of the method.

Accurate and efficient simulation of room impulse responses is crucial for spatial audio applications. However, existing acoustic ray-tracing tools often operate as black boxes and only output impulse responses (IRs), providing limited access to intermediate data or spatial fidelity. To address those problems, this paper presents GSound-SIR, a novel Python-based toolkit for room acoustics simulation that addresses these limitations. The contribution of this paper includes the follows. First, GSound-SIR provides direct access to up to millions of raw ray data points from simulations, enabling in-depth analysis of sound propagation paths that was not possible with previous solutions. Second, we introduce a tool to convert acoustic rays into high-order Ambisonic impulse response synthesis, capturing spatial audio cues with greater fidelity than standard techniques. Third, to enhance efficiency, the toolkit implements an energy-based filtering algorithm and can export only the top-X or top-X-% rays. Fourth, we propose to store the simulation results into Parquet formats, facilitating fast data I/O and seamless integration with data analysis workflows. Together, these features make GSound-SIR an advanced, efficient, and modern foundation for room acoustics research, providing researchers and developers with a powerful new tool for spatial audio exploration.

This paper proposes a new algorithm for enhancing the spatial resolution of measured first-order Ambisonics room impulse responses (FOA RIRs). It applies a separation of the RIR into a salient stream (direct sound and reflections) and a diffuse stream to treat them differently: The salient stream is enhanced using the Ambisonic Spatial Decomposition Method (ASDM) with a single direction of arrival (DOA) per sample of the RIR, while the diffuse stream is enhanced by 4-directional (4D-)ASDM with 4 DOAs at the same time. Listening experiments comparing the new Salient/Diffuse S/D-ASDM to ASDM, 4D-ASDM, and the original FOA RIR reveal the best results for the new algorithm in both spatial clarity and absence of artifacts, especially for its variant, which keeps the DOA constant within each salient event block.

Head-related transfer functions (HRTFs) are used in auditory applications for spatializing virtual sound sources. Listener-specific HRTFs, which aim at mimicking the filtering of the head, torso and pinnae of a specific listener, improve the perceived quality of virtual sound compared to using non-individualized HRTFs. However, using listener-specific HRTFs may not be accessible for everyone. Here, we propose as an alternative to take advantage of the adaptation abilities of human listeners to a new set of HRTFs. We claim that agreeing upon a single listener-independent set of HRTFs has beneficial effects for long-term adaptation compared to using several, potentially severely different HRTFs. Thus, the Non-individual Ear MOdel (NEMO) initiative is a first step towards a standardized listener-independent set of HRTFs to be used across applications as an alternative to individualization. A prototype, NEMObeta, is presented to explicitly encourage external feedback from the spatial audio community, and to agree on a complete list of requirements for the future HRTF selection.

Multimodal research and applications are becoming more commonplace as Virtual Reality (VR) technology integrates different sensory feedback, enabling the recreation of real spaces in an audio-visual context. Within VR experiences, numerous applications rely on the user’s voice as a key element of interaction, including music performances and public speaking applications. Self-perception of our voice plays a crucial role in vocal production. When singing or speaking, our voice interacts with the acoustic properties of the environment, shaping the adjustment of vocal parameters in response to the perceived characteristics of the space.

This technical report presents a real-time auralization pipeline that leverages three-dimensional Spatial Impulse Responses (SIRs) for multimodal research applications in VR requiring first-person vocal interaction. It describes the impulse response creation and rendering workflow, the audio-visual integration, and addresses latency and computational considerations. The system enables users to explore acoustic spaces from various positions and orientations within a predefined area, supporting three and five Degrees of Freedom (3Dof and 5DoF) in audio-visual multimodal perception for both research and creative applications in VR.

The design of this pipeline arises from the limitations of existing audio tools and spatializers, particularly regarding signal latency, and the lack of SIRs captured from a first-person perspective and in multiple adjacent distributions to enable translational rendering. By addressing these gaps, the system enables real-time auralization of self-generated vocal feedback.

Immersive Audio Media and Formats (IAMF), also known as Eclipsa Audio, is an open-source audio container developed to accommodate multichannel and scene-based audio formats. Headphone-based delivery of IAMF audio requires efficient binaural rendering. This paper introduces the Open Binaural Renderer (OBR), which is designed to render IAMF audio. It discusses the core rendering algorithm, the binaural filter design process as well as real-time implementation of the renderer in a form of an open-source C++ rendering library. Designed for multi-platform compatibility, the renderer incorporates a novel approach to binaural audio processing, leveraging a combination of spherical harmonic (SH) based virtual listening room model and anechoic binaural filters. Through its design, the IAMF binaural renderer provides a robust solution for delivering high-quality immersive audio across diverse platforms and applications.

A computational framework is proposed for analyzing the temporal evolution of perceptual attributes of sound stimuli. As a paradigm, the perceptual attribute of envelopment, which is manifested in different audio sound reproduction formats, is employed. For this, listener temporal ratings of the envelopment for mono, stereo, and 5.0-channel surround music samples, serve as the ground truth for establishing a computational model that can accurately trace temporal changes from such recordings. Combining established and heuristic methodologies, different features of the audio signals were extracted at each segment that envelopment ratings were registered, named long-term (LT) features. A memory LT computational stage is proposed to account for the temporal variations of the features through the duration of the signal, based on the exponentially weighted moving average of the respective LT features. These are utilized in a gradient tree boosting, machine learning algorithm, leading to a Dynamic Model that accurately predicts the listener’s temporal envelopment ratings. Without the proposed memory LT feature function, a Static Model is also derived, which is shown to have lower performance for predicting such temporal envelopment variations.

The Honeybee is an insect known to almost all human beings around the world. The sounds produced by bees is a ubiquitous staple of the soundscape of the countryside and forest meadows, bringing an air of natural beauty to the perceived environment. Honeybee-produced sounds are also an important part of apitherapeutic experiences, where the close-quarters exposure to honeybees proves beneficial to the mental and physical well-being of humans. This research investigates the generation of synthetic honeybee buzzing sounds using Conditional Generative Adversarial Networks (cGANs). Trained on a comprehensive dataset of real recordings collected both inside and outside the beehive during a long-term audio monitoring session. The models produce diverse and realistic audio samples. Two architectures were developed: an unconditional GAN for generating long, high-fidelity audio, and a conditional GAN that incorporates time-of-day information to generate shorter samples reflecting diurnal honeybee activity patterns. The generated audio exhibits both spectral and temporal properties similar to real recordings, as confirmed by statistical analysis performed during the experiment. This research has implications for scientific research in honeybee colony health monitoring as well as apitherapy research. and artistic endeavours, for example in sound design and immersive soundscape creation, the trained generator model is publicly available on the project’s website.

Existing methods for moving sound source localization and tracking face significant challenges when dealing with an unknown number of sound sources, which substantially limits their practical applications. This paper proposes a moving sound source tracking method based on source signal envelopes that does not require prior knowledge of the number of sources. First, an encoder-decoder attractor (EDA) method is used to estimate the number of sources and obtain an attractor for each source, based on which the signal envelope of each source is estimated. This signal envelope is then used as a clue for tracking the target source. The proposed method has been validated through simulation experiments. Experimental results demonstrate that the proposed method can accurately estimate the number of sources and precisely track each source.

Traditional methods for inferring room geometry from sound signals are predominantly based on Room Impulse Response (RIR) or prior knowledge of the sound source location. This significantly restricts the applicability of these approaches. This paper presents a method for estimating room geometry based on the localization of direct sound source and its early reflections from First-Order Ambisonics (FOA) signals without the prior knowledge of the environment. First, this method simultaneously estimates the Direction of Arrival (DOA) of the direct source and the detected first-order reflected sources. Then, a Cross-attention-based network for implicitly extracting the features related to Time Difference of Arrival (TDOA) between the direct source source and the first-order reflected sources is proposed to estimate the distances of the direct and the first-order reflected sources. Finally, the room geometry is inferred from the localization results of the direct and the first-order reflected sources. The effectiveness of the proposed method was validated through simulation experiments. The experimental results demonstrate that the method proposed achieves accurate localization results and performs well in inference of room geometry.

In recent years, there has been an increasing interest in binaural technology due to its ability to create immersive spatial audio experiences, particularly in streaming services and virtual reality applications. While audio localization studies typically focus on individual sound sources, ensemble width (EW) is crucial for scene-based analysis, as wider ensembles enhance immersion. We define intended EW as the angular span between the outermost sound sources in an ensemble, controlled during binaural synthesis. This study presents a comparison between human perception of EW and its automatic estimation under simulated anechoic conditions. Fifty-nine participants, including untrained listeners and experts, took part in listening tests, assessing 20 binaural anechoic excerpts synthesized using 2 publicly available music recordings, 2 different HRTFs, and 5 distinct EWs (0° to 90°). The excerpts were played twice in random order via headphones through a web-based survey. Only a subset of ten listeners, of which nine were experts, passed post-screening tests, with a mean absolute error (MAE) of 74.62° (±38.12°), compared to MAE of 5.92° (±0.14°) achieved a by pre-trained machine learning method using auditory modeling and gradient-boosted decision trees. This shows that while intended EW can be algorithmically extracted from synthesized recordings, it significantly differs from human perception. Participants reported insufficient externalization, front-back confusion (suggesting HRTF mismatch). The untrained listeners demonstrated response inconsistencies and a low degree of discriminability, which led to the rejection of most untrained listeners during post-screening. The findings may contribute to the development of perceptually aligned EW estimation models.

This research aims to provide a systematic approach for the analysis of geometrical and material characteristics of traditional frame drums using deep learning. A data-driven approach is used, integrating supervised and unsupervised feature extraction techniques to associate measurable audio features with perceptual attributes. The methodology involves the training of convolutional neural networks on Mel-Scale spectrograms to estimate wood type (classification), diameter (regression), and depth (regression). A multi-labeled dataset containing recorded samples of frame drums of different specifications is used for model training and evaluation. Hierarchical classification is explored, incorporating playing techniques and environmental factors. Handcrafted features enhance interpretability, helping determine the impact of construction attributes on sound perception, ultimately aiding instrument design. Data augmentation techniques, including pitch alterations, additive noise, etc. are introduced to expand the generalization of the approach and dataset expansion.

This paper discusses the process of generating natural language music descriptions, called captioning, using deep learning and large language models. A novel encoder architecture is trained to learn large-scale music representations and generate high-quality embeddings, which a pre-trained decoder then uses to generate captions. The captions used for training are from the state-of-the-art LP-MusicCaps dataset. A qualitative and subjective assessment of the quality of created captions is performed, showing the difference between various decoder models.

Artificial intelligence (AI) tools are transforming the way music is being produced. The rate of development is expeditious, and the associated metamorphosis of audio education is abrupt. Higher-level education is largely built around the objectives of knowledge transmission and skills development, evidenced by the emphasis on learning in the cognitive domain in University programmes. But the cohort of skills that music producers will require in five years’ time is unclear, making skills-based curriculum planning challenging. Audio educators require a systematic approach to integrate AI tools in ways that enhance teaching and learning.

This study uses speculative design as the underpinning research methodology. Speculative design employs design to explore and evaluate possible futures, alternative realities, and sociotechnical trends. In this study, the practical tasks in an existing university module are modified by integrating available GAI tools to replace or augment the task design. This tangible artefact is used to critique prevailing assumptions concerning the use of GAI in music production and audio education. The findings suggest that GAI tools will disrupt the existing audio education paradigm. Employing a process-centred approach to teaching and learning may represent a key progression for educators to help navigate these changes.

In automotive audio playback systems, dynamically increasing driving sounds are typically taken into account by applying a generic, i.e., non-individualized, increase in overall level and low-frequency amplification to compensate increased masking. This study investigated the degree of individuality regarding the preferences of noise-dependent level and equalizer settings. A user study with 18 normal-hearing participants was conducted in which individually preferred level-dependent and frequency-dependent amplification parameters were determined using a music-based procedure in quiet and in nine different driving noise conditions. The comparison of self-adjusted parameters suggested that, on average, participants adjusted higher overall levels and more low-frequency amplification in noise than in quiet. However, preferred self-adjusted levels differedmarkedly between participants for the same listening conditions but were quite similar in a re-test session for each participant, indicating that individual preferences were stable and could be reproducibly measured with the employed personalization scheme. Furthermore, the impact of driving noise on individually preferred settings revealed strong interindividual differences, indicating that listeners can differ widely with respect to their individual optimum of how equalizer and level settings should be dynamically adapted to changes in driving conditions.

Cyclical formal reviews are essential to keep Music and Audio Technology degree programmes current. Whilst clear institutional guidance exists on the requisite documentation to be submitted, there is little guidance concerning the process used to gather the information. To address this issue, a 12 step collaborative and reflective framework was developed to review a degree programme in Music Technology.

This framework employs Walker’s ‘Naturalistic’ process model and design thinking principles to create a dynamic, stakeholder-driven review process. The framework begins with reflective analysis by faculty, helping to define program identity, teaching philosophy, and graduate attributes. Existing curricula are evaluated using Boehm et al.’s (2018) tetrad framework of Music Technology encompassing the sub-disciplines of production, technology, art, and science. Insights from industry professionals, learners, and graduates are gathered through semi-structured interviews, surveys, and focus groups to address skill gaps, learner preferences, and emerging trends. A SWOT analysis further refines the scope and limitations of the redesign process, which culminates in iterative stakeholder consultations to finalise the program’s structure, content, and delivery.

This process-centred approach emphasises adaptability, inclusivity, and relevance, thus ensuring the redesigned program is learner-centred and aligned with future professional and educational demands. By combining reflective practice and collaborative engagement, the framework offers a comprehensive, replicable model for educators redesigning degree programmes in the discipline. This case study contributes to the broader discourse on curriculum design in music and audio degree programmes, demonstrating how interdisciplinary and stakeholder-driven approaches can balance administrative requirements with pedagogical innovation.

In this contribution we present subjective tests of loudspeaker virtualization, a technique enabling the application of specific target behaviors to the physical loudspeaker system. In this work, loudspeaker virtualization is applied to virtualize a closed box car audio subwoofer to replicate the performance of a larger vented enclosure. The tests are designed to determine if any reduction in sound quality is detected by a panel of listeners when a virtualized loudspeaker is used.

Acoustic Sovereignties (2024) is a First Nations, anti-colonial spatial audio exhibition held in Naarm (Melbourne), Australia. Through curatorial and compositional practices, Acoustic Sovereignties confronts traditional soundscape and Western experimental sound disciplines by foregrounding marginalised voices.
As this research will show, the foundations of sound-based practices such as Deep Listening and Soundscape Studies consisted of romanticised notions of Indigenous spirituality, in addition to the intentional disregard for First Nations stewardship and kinship with the land and its acoustic composition. Acoustic Sovereignties aims at reclaiming Indigenous representation throughout sound-based disciplines and arts practices by providing a platform for voices, soundscapes and knowledge to be heard.

Car audio systems aim to provide information, entertainment, and acoustic comfort to drivers and passengers in cars. In addition to basic audio functions for broadcasting, playing chimes, warning sound, and music, there are special audio features such as vehicle noise compensation, spatial sound effects, individual sound zone, and active noise control. In this paper, commonly used objective measurement methods for basic sound quality and special features in cars are reviewed and discussed. All objective measurements are proposed to use the 6-unit microphone array specified in the White Paper for In-car Acoustic Measurements released by AES Technical Committee on Automotive Audio in 2023, and the main parameters to be measured are frequency responses and sound pressure levels in the car when the specially designed test signals are played back. General measurement frameworks and procedures for basic sound quality and each feature are presented. The advantages and weakness of using these parameters to characterize the basic sound quality and special features of a car audio system are discussed, and the challenges and future directions are explored.

Experiments testing sound for augmented reality can involve real and virtual sound sources. Paradigms are either based on rating various acoustic attributes or testing whether a virtual sound source is believed to be real (i.e., evokes an auditory illusion). This study compares four experimental designs indicating such illusions. The first is an ABX task suitable for evaluation under the authenticity paradigm. The second is a Yes/No task, as proposed to evaluate plausibility. The third is a three-alternative-forced-choice (3AFC) task using different source signals for real and virtual, proposed to evaluate transfer-plausibility. Finally, a 2AFC task was tested. The renderings compared in the tests encompassed mismatches between real and virtual room acoustics. Results confirm that authenticity is hard to achieve under nonideal conditions, and ceiling effects occur because differences are always detected. Thus, the other paradigms are better suited for evaluating practical augmented reality audio systems. Detection analysis further shows that the 3AFC transfer-plausibility test is more sensitive than the 2AFC task. Moreover, participants are more sensitive to differences between real and virtual sources in the Yes/No task than theory predicts. This contribution aims to aid in selecting experimental paradigms in future experiments regarding perceptual and technical requirements for sound in augmented reality.

In recent years, professional and consumer audio and music technology has advanced in several areas, including sensory immersion, electronic transmission, content formats, and creation tools. The production and consumption of immersive media experiences increasingly rely on a global network of interconnected frameworks. These experiences, once confined to separate content markets like music, movies, video games, and virtual reality, are now becoming interoperable, ubiquitous, and adaptable to individual preferences, conditions, and languages. This article explores this evolution, focusing on flexible immersive audio creation and reproduction. We examine the development of object-based immersive audio technology and its role in unifying broadcast content with embodied experiences. We introduce the concept of Acoustic Objects, proposing a universal spatial audio scene representation model for creating and distributing versatile, navigable sound in music, multimedia, and virtual or extended reality applications.

Immersive audio mix presentations involve transmitting and rendering several audio elements simultaneously. This enables next-generation applications, such as personalized playback. Using immersive loudspeaker and headphone MUSHRA tests, we investigate rate vs. quality for a typical mix presentation use case of a foreground stereo element, plus a background Ambisonics scene. For coding, we use Immersive Audio Model and Formats, a recently proposed system for Next-Generation Audio. Excellent quality is achieved at 384 kbit/s, even with reasonable amount of personalization. We also propose a framework for content-aware analysis that can significantly reduce the bitrate even when using underlying legacy audio coding instances.

This study presents an ethnographic analysis of current immersive music production workflows, examining industry trends, tools, and methodologies. Through interviews and participant observations with professionals across various sectors, the research identifies common patterns, effective strategies, and persistent obstacles in immersive audio production. Key findings highlight the ongoing struggle for standardized workflows, the financial and technological barriers faced by independent artists, and the critical role of collaboration between engineers and creatives. Despite the growing adoption of immersive formats, workflows still follow stereo conventions, treating spatialization as an afterthought and complicating the translation of mixes across playback systems. Additionally, the study explores the evolving influence of object-based and bed-based mixing techniques, monitoring inconsistencies across playback systems, and the need for improved accessibility to immersive production education. By synthesizing qualitative insights, this paper contributes to the broader discourse on immersive music production, offering recommendations for future research and industry-wide best practices to ensure the sustainable integration of spatial audio technologies.

A psychoacoustic experiment is conducted to evaluate and compared the auditory distance perception in reflected sound field by using static and dynamic VAD. The binaural signals creased by a point source at different distances in a rectangular room are simulated. The contribution of direct sound to binaural signals is simulated by near-field head-related transfer function filters and a gain factor to account for the propagation attenuation of spherical surface wave. The contribution of early reflections up to the second order and later reverberation are respectively simulated by the image source method and Schroeder reverberation algorithm. The results of psychoacoustic experiment indicates that there are still significant differences between the perceived distances created by static VAD and these created by dynamic VAD in the simulated reflected condition, although the differences are not so large as those in the simulated free-field case. The results of dynamic VAD are more consistent with these of real sound source. Therefore, simulating reflections reduces the in-head-localization and thus improves the control of perceived distance in headphone presentation, but static VAD is still less effective in creating different distance perception. Dynamic VAD is still needed in the distance perception experiment for hearing researches even if simulated reflections are included. In practical applications, dynamic VAD is advocated for recreating virtual source at different distance.

With the rapid development of virtual reality (VR) and augmented reality (AR), spatial audio recording and reproduction have gained increasing research interest. Higher Order Ambisonics (HOA) stands out for its adaptability to various playback devices and its ability to integrate head orientation. However, current HOA recordings often rely on bulky spherical microphone arrays (SMA), and portable devices like smartphones are limited by array configuration and number of microphones. We propose a method for HOA encoding using a smartphone microphone array (SPMA). By designing beamformers for each order of spherical harmonic functions based on the array manifold, the method enables HOA encoding and up-scaling. Validation on a real SPMA and its simulated free-field counterpart in noisy and reverberant conditions showed that the method successfully encodes and up-scales HOA up to the fourth order with just four irregularly arranged microphones.

Through a practice-oriented study, various coincident, near-coincident, and non-coincident immersive microphone arrays were compared during drum recordings for different contemporary popular music genres. In a preliminary study, the OCT-3D, PCMA-3D, 2L-Cube, Hamasaki Square, IRT Cross, Ambisonics A-Format, and native B-Format were informally compared, revealing that the differences between non-coincident systems were much smaller than the differences between coincident and non-coincident systems. This led to a reduction in microphone systems for the final drum recordings. Four microphone techniques were selected: OCT-3D, native B-Format, Ambisonics A-Format, and IRT Cross. These were compared within the context of two different songs – a calm pop track and an energetic rock song – where the drums were respectively recorded in a dry drum booth and a large studio hall. Through a listening test with a small sample group, it was determined which microphone technique was best suited for each song. Participants were also asked to identify the general favorite, without musical context, as well as how the spatiality, timbre, and height were perceived. It was concluded that the choice of immersive microphone technique depends on the musical context. Conclusions from more objective studies focus primarily on accurate localization, with non-coincident systems consistently performing the best. However, these studies do not take into account the musical context, where accurate localization does not always take precedence. Furthermore, it was noted that height perception in music is not solely created by speakers in the height range. The comparative drum recordings are published through https://www.immersive.pxl.be/immersive- microphone-techniques-for-drums/.

This study investigated whether categorical perception—a phenomenon observed in speech perception—extends to the discrimination of vowel-like timbre qualities. Categorical perception occurs when continuous acoustic variations are perceived as distinct categories, leading to better discrimination near category boundaries than within a category. To test this, discrimination thresholds for the center frequency of a one-third-octave band formant introduced into the spectrum of a pink noise burst were measured in five subjects using an adaptive psychophysical procedure. Thresholds were assessed at distinctive formant frequencies of selected Polish vowels and at boundaries between adjacent vowel categories along the formant-frequency continuum. Results showed no reduction in discrimination thresholds at category boundaries, suggesting an absence of categorical perception for vowel-like timbre. One possible explanation for this finding lies in the listening mode—a concept from ecological auditory research—describing cognitive strategies in auditory tasks. The design of both the stimuli and the experimental procedure likely encouraged an acousmatic listening mode, which focuses solely on the sensory characteristics of sound, without reference to its source or meaning. This may have suppressed cues typically used in the categorical perception of speech sounds, which are associated with the communication listening mode. These findings highlight the importance of considering listening mode in future research on categorical perception of timbre and suggest that vowel-like timbre discrimination may involve perceptual mechanisms distinct from those used in speech sound discrimination.

Individual head-related transfer functions (HRTFs) are instrumental in rendering plausible spatial audio playback over headphones as well as in understanding auditory perception. Nowadays, the numerical calculation of individual HRTFs is achievable even without high-performance computers. However, the main obstacle is the acquisition of a mesh of the pinnae with a submillimeter accuracy. One approach to this problem is the photogrammetric reconstruction (PR), which estimates a 3D shape from 2D input, e.g., photos. Albeit easy to use, this approach comes with a trade-off in the resulting mesh quality, which subsequently has a substantial impact on the HRTF's quality. In this study, we investigated the effect of PR on HRTF quality as compared to HRTFs calculated from a reference mesh acquired with a high-quality structured-light scanner. Additionally, we applied two pinna deformation methods, which registered a non-individual high-quality pinna to the individual low-quality PR pinna by means of geometric distances. We investigated the potential of these methods to improve the quality of the PR-based pinna meshes. Our evaluation involved the geometrical, acoustical, and psychoacoustical domains including a sound-localization experiment with 9 participants. Our results show that neither PR nor PR-improvement methods were able to provide individual HRTFs of sufficient quality, indicating that without extensive pre- or post-processing, PR provides too little individual detail in the HRTF-relevant pinna regions.

Published studies indicate that musicians outperform non-musicians in a variety of non-musical auditory tasks, a phenomenon known as the “musicians’ hearing advantage effect.” One widely reported benefit is enhanced speech-in-noise (SIN) recognition. It was observed that musicians’ speech-in-noise (SIN) recognition thresholds (SRTs) are lower than those of non-musicians, though findings—mainly from English-language studies—are mixed; some confirm these advantage, while others do not. This study extends SRT measurements to Polish, a language with distinct phonetic characteristics. Participants completed a Polish speech intelligibility test, reconstructing sentences masked by multitalker babble noise by selecting words from a list displayed on a computer screen. Speech levels remained constant while masking noise was adjusted adaptively: increasing after each correct response and decreasing after each error. Three groups were tested: musicians, musically trained audio engineers, and non-musicians. Results showed that musicians and audio engineers had SRTs 2 and 2.7 dB lower than non-musicians, respectively. Although audio engineers exhibited slightly lower SRTs than musicians, the difference was minimal, with statistical significance just above the conventional 5% threshold. Thus, under these conditions, no clear advantage of audio engineers over musicians in SIN performance was observed.

An individual human pinna geometry can be used to achieve plausible personalized audio reproduction. However, an accurate acquisition of the pinna geometry typically requires the use of specialized equipment and often involves time-consuming post-processing to remove potential artifacts. To obtain an artifact-free but individualized mesh, a parametric pinna model based on cubic Bézier curves (BezierPPM) can be used to represent an individual pinna. However, the parameters need to be manually tuned to the acquired listener’s geometry. For increased scalability, we propose Mesh2PPM, a framework for an automatic estimation of BezierPPM parameters from an individual pinna. Mesh2PPM relies on a deep neural network (DNN) that was trained on a dataset of synthetic multi-view images rendered from BezierPPM instances. For the evaluation, unseen BezierPPM instances were presented to Mesh2PPM which inferred the BezierPPM parameters. We subsequently assessed the geometric errors between the meshes obtained from the BezierPPM parametrized with the inferred parameters and the actual pinna meshes. We investigated the effects of the camera-grid type, jittered camera positions, and additional depth information in images on the estimation quality. While depth information had no effect, the camera-grid type and the jittered camera positions both had effects. A camera grid of 3×3 provided the best estimation quality, yielding Pompeiu-Hausdorff distances of 2.05 ± 0.4 mm and 1.4 ± 0.3 mm with and without jittered camera
positions, respectively, and root-mean-square (RMS) distances of 0.92 ± 0.12 mm and 0.52 ± 0.07 mm. These results motivate further improvements of the proposed framework to be ultimately applicable for an automatic estimation of pinna geometries obtained from actual listeners.

The multi-stimulus test with hidden reference and anchor (MUSHRA) is a prevalent method for the subjective audio quality evaluation. Despite its popularity, the technique is not immune to biases. Empirical evidence indicates that the presence of labels (quality descriptors) equidistantly distributed along the rating scale may be the cause of its non-linear warping; however, other factors could evoke even stronger non-linear effects. This study aims to investigate the hypothesis that stimulus spacing bias may induce a greater magnitude of non-linear warping of the quality scale compared to that caused by the presence of labels. To this end, a group of more than 120 naïve listeners participated in MUSHRA-compliant listening tests using labeled and unlabeled graphic scales. The audio excerpts, representing two highly skewed distributions of quality levels, were reproduced over headphones in an acoustically treated room. The findings of this study verify the postulated hypothesis and shed new light on the mechanisms biasing results of the MUSHRA-conformant listening tests.

In order to reproduce audio over headphones as in-
tended, it is essential to have well-defined and con-
sistent references of how headphones should sound.
With the aim of stereo reproduction in mind, the field
has established a de-facto reference target curve called
the Harman Target Curve to which headphone transfer
functions are commonly compared. This contribution
questions if the same target curve is suitable when used
for the reproduction of spatial audio. First, the ori-
gins the Harman Curve are revisited; it is motivated by
the frequency response of loudspeaker playback in a
specific listening room. The necessary measurement
procedures are described in detail. Then, the paper
discusses the applicability of existing targets to spa-
tial audio. Therein, it is possible to embed convincing
spatial room information directly into the production,
thereby calling into question the motivation for incor-
porating a listening room in the headphone target. The
paper concludes with a listening experiment that com-
pares the preference of different target curves for both
spatial audio and stereo

This research aims to explore the impact of variations in apparent sound source width and position on emotional and physiological responses among listeners, with a particular focus on the domain of virtual reality applications. While sound is recognized as a potent elicitor of strong emotions, the specific role of spatial characteristics, such as apparent sound source width, has not been systematically analyzed. The authors’ previous study has indicated that the spatial distribution of sound can alter perceptions of scariness. In contrast, the current study explores whether adjustments in apparent sound source width can significantly affect emotional valence and arousal, as well as human physiological metrics. The objective of this study was to investigate the impact of a single sound source width and its horizontal position on emotional engagement, thereby providing valuable insights for advancements in immersive audio experiences. Our experiments involved conducting listening tests in a spatial sound laboratory, utilizing a circular setup of sixteen loudspeakers to present a range of audio stimuli drawn from five selected recordings. The stimuli were manipulated based on two key parameters: the apparent sound source width and the spatial positioning of the sound source (front, back, left, or right). Participants assessed their emotional reactions using the Self-Assessment Manikin (SAM) pictogram method. Physiological data, including electroencephalogram, blood volume pressure, and electrodermal activity was collected in real-time via wearable sensors consisting of an EEG headset and a finger-attached device.

In recent years, applications such as virtual reality (VR) systems and room acoustics simulations have brought the modeling of sound source directivity into focus. An accurate simulation of directional responses of sound sources is essential in immersive audio applications.

Real sound sources have directional properties that are different from simple sources such as monopoles, which are sources frequently used for modeling more complex acoustic fields. For instance, the sound level of human speech as a sound source varies considerably depending on where the sound is recorded with respect to the talker’s head. The same is true for loudspeakers, which are considered linear and time-independent sources. When the sound is recorded behind the speaker, it is normal to observe differences of up to 20 dB SPL at some frequencies. The directional characteristics of sound sources become particularly pronounced at high frequencies. The propagation of real sound sources, such as human voices or musical instruments, differs from simple source models like monopoles, dipoles, and quadrupoles due to their physical structures.

The common approach to measuring directivity patterns of sound sources involves surrounding a sound source in an anechoic chamber with a high number of pressure microphones on a spherical grid and registering the sound power at these positions. Apart from the prohibitive hardware requirements, such measurement setups are mostly impractical and costly. Audio system manufacturers have developed various methods for measuring sound source directionality over the years. These methods are generally of high technical complexity.

This article proposes a new, reduced-complexity directivity measurement approach based on the spherical harmonic decomposition of the sound field. The method estimates the directional characteristics of sound sources using fewer measurement points with spherical microphone arrays. The spherical harmonic transform allows for the calculation of directivity using data collected from spherical microphone arrays instead of pressure sensors. The proposed method uses both the pressure component and spatial derivatives of the sound field and successfully determines directivity with sparse measurements.

An estimation model based on the spherical Fourier transform was developed, measurements were carried out to test this model, and preliminary results obtained from the estimation model are presented. Experiments conducted at the METU Spatial Audio Research Laboratory demonstrated the effectiveness of the proposed method. The directivity characteristics of Genelec 6010A loudspeaker are measured using eight 3rd-order spherical microphone arrays. The directivity functions obtained were highly consistent with the data provided by the loudspeaker manufacturer. The results, especially in low and mid-frequency bands, indicate the utility of the proposed method.

This paper describes ongoing research on integrating algorithmic reverberation tools designed for audio post-production into virtual acoustics, focusing on using Impulse Responses (IRs) captured from the legendary Lexicon 960L hardware reverberation unit. While previous research from the McGill University Virtual Acoustics Technology (VAT) Lab has utilized room impulse responses (RIRs) captured from various performance halls to create active acoustic environments in the recording studio, this study analyzes the perceived differences between the two listening environments and the effect of the VATLab speakers and effect of room acoustics on IRs captured from 5.0 multichannel reverb presets. Three of these multichannel IRs have been chosen to simulate a Lexicon 960L “environment” in a physical space.

Objective measurements in McGill University’s Immersive Media Laboratory (IMLAB) Control Room and in VATLab following the ISO 3382 standard measure the effect of the physical room and the omnidirectional dodecahedral speakers used for auralization. Through a subjective pilot study, subjective analysis investigates the perceived differences between the Lexicon IRs in VATLab and a control condition, the IMLAB control room. The results of an attribute rating test on perceived immersion, soundfield continuity, tone color, and overall listening experience between the two spaces helps us better understand how reverberation algorithms designed for multichannel mixing/post-production translate to a virtual acoustics system.
In conclusion, we discuss the perceptual differences between the IMLAB Control Room and VATLab and results of objective measurements.

Immersive sound reinforcement aims to create a balanced perception of sounds arriving from different directions, establishing an impression of envelopment over the audience area. Current perceptual research shows that coverage designs featuring nearly constant decay (0dB per distance doubling) preserve the level balance among audio objects in the mix. In contrast, a -3dB decay supports a more uniform sensation of envelopment, especially for off-center listening positions. For practical reasons, point-source loudspeakers remain widely used for immersive audio playback in mid-sized venues. However, point-source loudspeakers inherently decay by -6dB per distance doubling, and using them can conflict with the design goals outlined above. In this paper, we investigate the perceived differences between point-source and line-source setups using eight surrounding loudspeakers side-by-side covering a 10m x 7m audience area. The perceptual qualities of object level balance, spatial definition, and envelopment were compared in a MUSHRA listening experiment, and acoustic measurements were carried out to capture room impulse responses and binaural room impulse responses (BRIRs) of the experimental setup. The BRIRs were used to check whether the results of the listening experiment were reproducible on headphones. Both the loudspeaker and headphone-based experiments delivered highly correlated results. Also, regression models devised based on the acoustic measurements are highly correlated to the perceptual results. The results confirm that elevated line sources, exhibiting a practically realizable decay of -2dB per distance doubling, help preserve object-level balance, increase spatial definition, and provide a uniform envelopment experience throughout the audience area compared to point-source loudspeakers.

Numerous studies highlight the role of transient behavior in musical sounds and its impact on sound identification. This study compares these findings with established psychoacoustic measurements of detection thresholds for asynchrony in onset and offset transients, obtained using synthesized stimuli that allowed precise control of stimulus parameters. Results indicated that onset asynchrony can be detected at thresholds as low as 1 ms—even half a cycle of the component frequency. In contrast, offset asynchrony detection was found to be less precise, with thresholds ranging from 5 to 10 ms. Sensitivity improves when multiple harmonics are asynchronous. Additionally, component phase significantly influences onset asynchrony detection: at 1000 Hz and above, phase shifts raise thresholds from below 1 ms to around 50 ms, while having little effect on offset detection. Although these findings were based on controlled artificial stimuli, they can provide valuable insight into asynchrony in natural musical sounds. In many cases, detection thresholds are well below the variations observed in music, yet under certain conditions and frequencies, some temporal variations may become not perceptible.

Wide Dynamic Range Compression in hearing aids is becoming increasingly more complex as the number of channels and adjustable parameters grow. At the same time, there is growing demand for customization and user self-adjustment of hearing aids, necessitating a balance between complexity and user accessibility. Compression in hearing aids is governed by the input-output transfer function, which relates input magnitude to output magnitude, and is typically defined as a combination of linear piecewise segments resembling logarithmic behavior. This work presents an alternative to the conventional compression transfer function that consolidates multiple compression parameters and revisits expansion in hearing aids. The
curvilinear transfer function is a continuous curve with logarithm-like behavior, governed by two parameters—gain and compression ratio. Experimental results show that curvilinear compression reduces the amplification of low-level noise, improves signal-to-noise ratio by up to 1.0 dB, improves sound quality as measured by the Hearing Aids Speech Quality Index by up to 6.7%, and provides comparable intelligibility as measured by the Hearing Aids Speech Perception Index, with simplified parameterization compared to conventional compression.
The consolidated curvilinear transfer function is highly applicable to over-the-counter hearing aids and offers more capabilities for customization than current prominent over-the-counter and self-adjusted hearing aids.

Hearing loss is a global public health issue due to its high prevalence and negative impact on various aspects of one’s life, including well being and cognition. Despite their crucial role in auditory rehabilitation, hearing aids remain inaccessible to many due to their high costs, particularly in low- and middle-income countries. Existing open-source solutions often rely on high-power, bulky platforms rather than compact, low-power wearables suited for real-world applications. This work introduces Tiresias, an open-source hearing aid development board designed for real-time audio processing using low-cost electronics. Integrating key hearing aid functionalities into a compact six-layer printed circuit board (PCB), Tiresias features multichannel compression, digital filtering, beamforming, Bluetooth connectivity, and physiological data monitoring, fostering modularity and accessibility through publicly available hardware and firmware resources based on the Nordic nRF Connect and Zephyr real-time operating system (RTOS). By addressing technological and accessibility challenges, this work advances open-source hearing aid development, enabling research in hearing technologies, while also supporting future refinements and real-world validation.

Simple truncation of the reflections in the impulse response of loudspeakers measured in normal rooms will increasingly falsify the response below about 500 Hz for typical situations. Well-known experience and guidance from loudspeaker models allow the determination of the lowest frequency for which truncation suffices. This paper proposes two additional strategies for achieving much improved low-frequency responses that are complementary to the easily-obtained high-frequency response: (a) a previously published nearfield measurement which can be diffractively transformed to a farfield response with appropriate calculations, here presented with greatly simplified computations, and (b) a measurement setup that admits only a single floor reflection which can be iteratively corrected at low frequencies. Theory and examples of each method are presented.

A new one-third-octave-band noise criteria (NC) rating method is presented. one-third-octave-band NC curves from NC 70 to NC 0 are derived from the existing octave-band curves, adjusted for bandwidth, fit to continuous functions, and redistributed progressively over this space. This synthesis is described in detail. The diffuse field hearing threshold at low frequencies is also derived. Several NC curves at high frequencies are shown to be below threshold (inaudible). NC ratings are calculated using both the new one-third-octave-band and the legacy octave-band methods for a number of different room noise spectra. The resulting values were found to be similar for both methods. NC ratings using the new method are particularly applicable to very low noise level critical listening environments such as recording studios, scoring stages, and cinema screening rooms, but are shown to also be applicable to higher noise level environments. The proposed method better tracks the audibility of noise at low levels as well as the audibility of tonal noise components, while the legacy method as originally conceived generally emphasizes speech interference.

We propose a practical method for the measurement of the pressure sensitivity frequency response of a microphone that has been integrated into product mechanics. The method uses a probe microphone to do determine the sound pressure entering the inlet of the integrated microphone. We show that the measurements can be performed in a normal office environment as well as in anechoic conditions. The method is validated with measurement of a rigid spherical microphone prototype having analytically defined scattering characteristics. Our results indicate that the proposed method, called IMPro, can effectively measure the pressure sensitivity frequency response of microphones in commercial products, quite independent of the measurement environment.

This paper introduces a new algorithm for multiposition mixed-phase equalization of slot-loaded loudspeaker responses obtained in the horizontal and vertical plane, using finite impulse response (FIR) filters. The algorithm selects a {\em prototype response} that yields a filter that best optimizes a time-domain-based objective metric for equalization for a given direction. The objective metric includes a weighted linear combination of pre-ring energy, early and late reflection energy, and decay rate (characterizing impulse response shortening) during filter synthesis. The results show that the presented mixed-phase multiposition filtering algorithm performs a good equalization along all horizontal directions and for most positions in the vertical direction. Beyond the multiposition filtering capabilities, the algorithm and the metric are suitable for designing mixed-phase filters with low delays, an essential constraint for real-time processing.

It is challenging to characterize sound across space, especially in small enclosed volumes, using conventional microphone arrays.
This study explores acousto-optic sensing methods to record the sound field throughout an enclosure, including regions close to a source and boundaries.
The method uses a laser vibrometer to sense modulations of the refractive index in air, caused by the propagating sound pressure waves.
Compared to microphone arrays, the sound field can be measured non-invasively and at high resolution which is particularly attractive at high frequencies, in enclosures of limited size or unfavorable mounting conditions for fixtures.
We compensate for vibrations that contaminate and conceal the acousto-optic measurements and employ an image source model to also reconstruct early parts of the impulse response.
The results demonstrate that acousto-optic measurements can enable the analysis of sound field in enclosed spaces non-invasively and with high resolution.

This paper describes the design of an Analog Stomp Box capable of reproducing the effect observed when a loudspeaker is rotated during operation, the so-called Leslie effect. When the loudspeaker is rotating two physical effects can be observed: The first is a variation of the amplitude because sometimes the speaker is aimed at the observer and then, after 180 degrees of rotation, the loudspeaker is aimed opposing to the observer. To recreate this variation in amplitude, a circuit called Tremolo was designed to achieve this effect. The second is the Doppler effect, which was obtained with a circuit designed to vary the phase of the signal (Vibrato). The phase variation simulates a frequency variation for the ears. Assembling these two circuits in cascade, it is obtained the Pseudo Leslie Effect. These Vibrato and Tremolo circuits receive the control signal from a Low Frequency Oscillator (LFO) which controls the effect frequency. To get a high degree of repeatability, which is not simple in analog circuits employing photocouplers, those photocoupler devices were replaced with VCAs. The photocouplers have a great variation of your optical characteristics, so it is hard to obtain the same result in a large-scale production. However, using VCAs it turns to be easily achievable. The THAT2180 IC is a VCCS, Voltage-Controlled Current Source with an exponential gain control and low signal distortion. The term Pseudo was used because, in the Leslie Effect, the rotation of the loudspeaker gives a lag of 90o between the frequency and amplitude variations. This lag has not been implemented, but the sonic result left nothing to be desired.

Recording engineers and producers choose different microphones for different sound sources. It is intriguing that, in the 1950s and 1960s, the variety of available microphones was relatively limited compared to what we have available today. Yet, recordings from that era remain exemplary even now. The microphones used at the time were primarily vacuum tube models.
Through discussions at AES Conventions on improving phantom power supplies and my own experimentation with tube microphones myself, I began to realize that defining attribute of their sound might not stem solely from the tubes themselves. Instead, the type of power supply appeared to play a crucial role in shaping the final sound quality.
This hypothesis was confirmed with the introduction of high-voltage DPA 4003 and 4004 microphones, compared to their phantom-powered counterparts, the 4006 and 4007. In direct comparisons, the microphones with external, more current-efficient power supplies consistently delivered superior sound.
Having worked extensively with numerous AKG C12 and C24 microphones I identified two pairs, one of C12s and one of C24s with identical frequency characteristics. For one C12, we designed an entirely new, pure Class A transistor-based circuit with an external power supply.
Reflecting on my 50-plus years as a sound engineer and producer, I sought to determine which microphones were not only the best, but also the most versatile. My analysis led to four key solutions extending beyond the microphones themselves. Since I had already developed an ideal Class A equalizer, I applied the same technology to create four analog equalizers designed to fine-tune the prototype microphone’s frequency characteristics at the power supply level.

In today's music industry and among musicians, instead of using analog hardware effects to alter sound, digital counterparts are increasingly being used, often in the form of software plugins. The circuits of musical devices often contain nonlinear components (diodes, vacuum tubes, etc.), which complicates their digital modeling. One of the approaches to address this is the use of state-space methods, such as the Euler or Runge-Kutta methods. To guarantee stability, implicit state-space methods should be used; however, they require the numerical solution of an implicit equation, leading to large computational complexity. Alternatively, the K-method can be used that avoids the need of numerical methods if the system meets certain conditions, thus significantly decreasing the computational complexity. Although the K-method has been invented almost three decades ago, the authors are not aware of a thorough computational complexity analysis of the method in comparison to the more common implicit state-space approaches, such as the backward Euler method. This paper introduces these two methods, explores their advantages, and compares their computational load as a function of model size by using a scalable circuit example.

Virtual acoustic systems can artificially alter a recording studio's reverberation in real time using spatial room impulse responses captured in different spaces. By recreating another space's acoustic perception, these systems influence various aspects of a musician's performance. Traditional methods involve recording a dry performance and adding reverb in post-production, which may not align with the musician's artistic intent. In contrast, virtual acoustic systems allow simultaneous recording of both artificial reverb and the musician's interaction using standard recording techniques—just as it would occur in the actual space. This study analyzes immersive recordings of nearly identical musical performances captured in both real concert hall and McGill University's Immersive Media Lab (Imlab), which features a new dedicated virtual acoustics software, and highlights the similarities and differences between the performances recorded in the real space and its virtual counterpart.

Modeling or compensating a given transfer function is a common task in the field of audio. To comply with the characteristics of hearing, logarithmic frequency resolution filters have been developed, including the Kautz filter, which has orthogonal tap outputs. When the system to be modeled is time-varying, the modeling filter should be tuned to follow the changes in the transfer function. The Least Mean Squares (LMS) and Recursive Least Squares (RLS) algorithms are well-known methods for adaptive filtering, where the latter has faster convergence rate with lower remaining error, at the expense of high computational demand. In this paper we propose a simplification to the RLS algorithm, which builds on the orthogonality of the tap outputs of Kautz filters, resulting in a significant reduction in computational complexity.

The acoustic behaviour of an auditorium is analysed after measurements performed according to the ISO 3382:1 standard. The all-pole analysis of the measured impulse responses confirms the hypothesis that all responses have a common component that can be attributed to room characteristis. Results from a subsequent non-parametric analysis allows conjecturing that the overall reponse of the acoustic channel between two points may de decomposed in three components: one related to source position, another related to the room, and the last one depending on the position of the receiver.

Without relying on audio data as a reference, artificial reverberation models often struggle to accurately simulate
the acoustics of real rooms. To address this, we propose a hybrid reverberator derived from a room’s physical
properties. Room geometry is extracted via Light Detection and Ranging mapping, enabling the calculation of
acoustic reflection paths via the Image Source Method. Frequency-dependent absorption is found by classifying
room surface materials with a multi-modal Large Language Model and referencing a database of absorption
coefficients. The extracted information is used to parametrise a hybrid reverberator, divided into two components:
early reflections, using a tapped delay line, and late reverberation, using a Scattering Feedback Delay Network.
Our listening test results show that participants often rate the proposed system as the most natural simulation of a
small hallway room. Additionally, we compare the reverberation metrics of the hybrid reverberator and similar
state-of-the-art models to those of the small hallway.

Sound fields in enclosures comprise a combination of directional and diffuse components. The directional components include the direct path from the source and the early specular reflections. The diffuse part starts with the first early reflection and builds up gradually over time. An ideal diffuse field is achieved when incoherent reflections begin to arrive randomly from all directions. More specifically, a diffuse field is characterized by having uniform energy density (i.e., independence from measurement position) and an isotropic distribution (i.e. random directions of incidence), which results in zero net energy flow (i.e. the net time-averaged intensity is zero). Despite this broad definition, real diffuse sound fields typically exhibit directional characteristics owing to the geometry and the non-uniform absorptive properties of rooms.

Several models and data-driven metrics based on the definition of a diffuse field have been proposed to assess diffuseness. A widely used metric is the _mixing time_, which indicates the transition of the sound field from directional to diffuse and is known to depend, among other factors, on the room geometry.

The concept of mixing time is closely linked to normalized echo density (NEDP), a measure first used to estimate the mixing time in actual rooms (Abel and Huang, 2006), and later to assess the quality of artificial reverberators in terms of their capacity to produce a dense reverberant tail (De Sena et al., 2015). NEDP is calculated over room impulse responses measured with a pressure probe, evaluating how much the RIR deviates from a normal distribution. Another similar temporal/statistical measure, kurtosis, has been used to similar effect (Jeong, 2016). However, neither NEDP nor kurtosis provides insights into the directional attributes of diffuse fields. While both approaches rely on statistical reasoning rather than identifying individual reflections, another temporal approach uses matching pursuit to identify individual reflections (Defrance et al., 2009).

Another set of approaches focuses on the net energy flow aspect of the diffuse field, providing an energetic analysis framework either in the time domain (Del Galdo et al., 2012) or in the time-frequency domain (Ahonen and Pulkki, 2009). These approaches rely on calculating the time-averaged active intensity, either using intensity probes or first- and higher-order Ambisonics microphones, where a pseudo-intensity-based diffuseness is computed (Götz et al., 2015). The coherence of spherical harmonic decompositions of the sound field has also been used to estimate diffuseness (Epain and Jin, 2016). Beamforming methods have likewise been applied to assess the directional properties of sound fields and to illustrate how real diffuse fields deviate from the ideal (Gover et al., 2004).

We propose a spatio-spectro-temporal (SST) sound field analysis approach based on a sparse plane-wave decomposition of sound fields captured using a higher-order Ambisonics microphone. The proposed approach has the advantage of analyzing the progression of the sound field’s diffuseness in both temporal and spatial dimensions. Several derivative metrics are introduced to assess temporal, spectro-temporal, and spatio-temporal characteristics of the diffuse field, including sparsity, diversity, and isotropy. We define the room sparsity profile (RSP), room sparsity relief (RSR), and room sparsity profile diversity (RSPD) as temporal, spectro-temporal, and spatio-temporal measures of diffuse fields, respectively. The relationship of this new approach to existing diffuseness measures is discussed and supported by experimental comparisons using 4th- and 6th-order acoustic impulse responses, demonstrating the dependence of the new derivative measures on measurement position. We conclude by considering the limitations and applicability of the proposed approach.

Various AD conversion methods exist, and high-speed 1 bit method have been proposed with using a high sampling frequency and 1 bit quantization. The ΔΣ modulation is mainly used, and due to its characteristic, these signals are able to accurately preserve the spectrum of the analog signal and move quantization noise into higher frequency bands, which allows for a high signal-to-noise ratio in the audible range. However, When performing signal processing tasks such as addition and multiplication on high-speed 1 bit signals, it is generally necessary to convert them into multi-bit signals for arithmetic operations. In this paper, we propose a direct processing method for high-speed 1 bit signal without converting them into multi-bit signal and the convolution is realized. In this method, 1 bit data are reordered to achieve operations without arithmetic one. The proposed method was verified through the simulations with using low-pass FIR filters. Frequency-domain analysis showed that the proposed method achieved equivalent performance to conventional multi-bit convolutions with successfully performing the desired filtering. In this paper, we present a novel approach to directly processing high-speed 1 bit signals and suggest potential applications in audio and signal processing fields.

Ambisonic technology has recently gained popularity in room acoustic measurements due to its ability to capture both general and spatial characteristics of a sound field using a single microphone. On the other hand, conventional measurement techniques conducted in accordance with the ISO 3382-1 standard require multiple transducers, which results in more time-consuming procedure. This study presents a case study on the use of ambisonic technology to evaluate the room acoustic parameters of a medium-sized recording studio.
Two ambisonic microphones, a first-order Sennheiser Ambeo and a third-order Zylia ZM1-3E, were used to record spatial impulse responses in 30 combinations of sound source and receiver positions in the recording studio. Key acoustic parameters, including Reverberation Time (T30), Early Decay Time (EDT) and Clarity (C80), were calculated using spatial decomposition methods. The Interaural Cross-Correlation Coefficient (IACC) was derived from binaural impulse responses obtained using the MagLS binauralization method. The results were compared with conventional omnidirectional and binaural microphone measurements to assess the accuracy and advantages of ambisonic technology. The findings show that T30, EDT, C50 and IACC values measured with the use of ambisonic microphones are consistent with those obtained from conventional measurements.
This study demonstrates the effectiveness of ambisonic technology in room acoustic measurements by capturing a comprehensive set of parameters with a single microphone. Additionally, it enables the estimation of reflection vectors, offering further insights into spatial acoustics.

Speech denoising is a prominent and widely utilized task, appearing in many common use-cases. Although there are very powerful published machine learning methods, most of those are too complex for deployment in everyday and/or low resources computational environments, like hand-held devices, smart glasses, hearing aids, automotive platforms, etc. Knowledge distillation (KD) is a prominent way for alleviating this complexity mismatch, by transferring the learned knowledge from a pre-trained complex model, the teacher, to another less complex one, the student. KD is implemented by using minimization criteria (e.g. loss functions) between learned information of the teacher and the corresponding one from the student. Existing KD methods for speech denoising hamper the KD by bounding the learning of the student to the distribution learned by the teacher. Our work focuses on a method that tries to alleviate this issue, by exploiting properties of the cosine similarity used as the KD loss function. We use a publicly available dataset, a typical architecture for speech denoising (e.g. UNet) that is tuned for low resources environments and conduct repeated experiments with different architectural variations between the teacher and the student, reporting mean and standard deviation of metrics of our method and another, state-of-the-art method that is used as a baseline. Our results show that with our method we can make smaller speech denoising models, capable to be deployed into small devices/embedded systems, to perform better compared to when typically trained and when using other KD methods.