Sound Data Research Articles

Modeling noise from large-scale data center equipment

Data centers present unique noise modeling challenges due to the scale of equipment needed to meet cooling and backup power demands. Three-dimensional environmental noise modeling software that is based on ISO 9613-2 "Attenuation of sound during propagation outdoors" offers a range of modeling techniques. While more detailed modeling techniques may offer more accuracy in the near field, they also increase model calculation run times. Additionally, manufacturer sound emissions data often more closely align with simplified modeling techniques. This evaluation reviews common data center noise sources, reviews the range of manufacturer noise emissions data typically available for such sources, and compares noise modeling techniques with reference to on-site observations.

Down-the-hole drilling construction acoustic research

nd The Alaska Department of Transportation and Public Faculties (DOT&PF) is conducting research to characterize underwater sounds from down-the-hole (DTH) drilling activities. This method of drilling is commonly used to drill holes through hard or rock substrates to support piles or tensions anchors for piles. A pneumatic DTH hammer is driven by pressured air coming from an air compressor that provides the energy to power the percussion piston that hits the drill bit with a specific impact frequency and energy to break the rock and advanced the hole. This activity produces continuous and impulsive sounds that affect marine mammals. Pile installation in marine environments typically uses vibratory drivers and/or impact pile driving. Underwater sounds from these activities have been well documented in publications, for example, there are compendiums prepared by Caltrans, and the U.S. Navy, along with many project-specific compliance reports submitted to resource agencies. However, there is limited acoustical data available for DTH activities, as the first published measurements were only made recently. The National Marine Fisheries Services (NMFS) recognizes the acoustical issues with DTH noise and is encouraging the collection and dissemination of additional sound data. This paper describes the research conducted and available DTH acoustic data.

Empowering college students with personal privacy protection and anti-fraud on the internet: A comprehensive learning approach

As college students increasingly navigate their academic pursuits and social lives through the digital realm, safeguarding personal privacy on the internet becomes a pressing concern. For example, deep forgery technology has further increased personal privacy concerns. This technique can produce fake videos, pictures, and sounds that are highly realistic, making it a tool for the misuse of personal and sound data. Several artificial intelligences (AI) systems require large amounts of data for training and improvement, which can involve collecting and storing sensitive personal information. If this data is not adequately protected, and there is no clear privacy policy, it may lead to the misuse or disclosure of personal information. In response to the rapid development of artificial intelligence and the impact of deep forgery technology on privacy and security, this research first tested the students' understanding of deep forgery technology, then introduced the dangers of this technology, and finally gave them a reasonable and effective solution. To promote active learning, activities are integrated to simulate real-world scenarios that college students might encounter. These include creating strong passwords, identifying phishing attempts, and understanding the implications of oversharing on social media platforms. Additionally, students are encouraged to examine case studies documenting privacy breaches and their impact on individuals and society. This essay aims to present an effective learning approach for college students to enhance their understanding of personal privacy protection online and provide an effective way of solving personal problems.

Speed of sound measurements and derived third and fourth acoustic virial coefficients of supercritical neon

We report comprehensive and accurate measurements of the speed of sound in neon. These measurements were carried out by a double-path-length pulse-echo technique and cover the temperature range between 200 K and 420 K with pressures up to 100 MPa. The standard uncertainties are 1.9 mK in temperature, 22 parts in 106 in pressure and 35 parts in 106 in speed of sound. The third and fourth acoustic virial coefficients of neon were derived from the speed of sound data in the temperature range of the measurements by fitting a fourth-order acoustic virial expansion in pressure with the second acoustic virial coefficient constrained from first-principles calculations. To support our claimed uncertainty, we determined the ratio M/γ0 between the molar mass M and the ideal-gas heat capacity ratio γ0 of the neon sample with a relative standard uncertainty of 7.7 parts in 106 by additional speed of sound measurements using a spherical resonator at 273.16 K.

DCNN for Pig Vocalization and Non-Vocalization Classification: Evaluate Model Robustness with New Data.

Since pig vocalization is an important indicator of monitoring pig conditions, pig vocalization detection and recognition using deep learning play a crucial role in the management and welfare of modern pig livestock farming. However, collecting pig sound data for deep learning model training takes time and effort. Acknowledging the challenges of collecting pig sound data for model training, this study introduces a deep convolutional neural network (DCNN) architecture for pig vocalization and non-vocalization classification with a real pig farm dataset. Various audio feature extraction methods were evaluated individually to compare the performance differences, including Mel-frequency cepstral coefficients (MFCC), Mel-spectrogram, Chroma, and Tonnetz. This study proposes a novel feature extraction method called Mixed-MMCT to improve the classification accuracy by integrating MFCC, Mel-spectrogram, Chroma, and Tonnetz features. These feature extraction methods were applied to extract relevant features from the pig sound dataset for input into a deep learning network. For the experiment, three datasets were collected from three actual pig farms: Nias, Gimje, and Jeongeup. Each dataset consists of 4000 WAV files (2000 pig vocalization and 2000 pig non-vocalization) with a duration of three seconds. Various audio data augmentation techniques are utilized in the training set to improve the model performance and generalization, including pitch-shifting, time-shifting, time-stretching, and background-noising. In this study, the performance of the predictive deep learning model was assessed using the k-fold cross-validation (k = 5) technique on each dataset. By conducting rigorous experiments, Mixed-MMCT showed superior accuracy on Nias, Gimje, and Jeongeup, with rates of 99.50%, 99.56%, and 99.67%, respectively. Robustness experiments were performed to prove the effectiveness of the model by using two farm datasets as a training set and a farm as a testing set. The average performance of the Mixed-MMCT in terms of accuracy, precision, recall, and F1-score reached rates of 95.67%, 96.25%, 95.68%, and 95.96%, respectively. All results demonstrate that the proposed Mixed-MMCT feature extraction method outperforms other methods regarding pig vocalization and non-vocalization classification in real pig livestock farming.

Can we still ensure no one is left behind by 2030? Demonstrating the potential of the implementation of the WHO Functioning and Disability Disaggregation Tool (FDD11) in existing survey platforms for disaggregating SDG indicators by disability

Purpose: The world is approaching the sustainable development goals deadline, but many countries still do not produce the data required to track their indicators by disability. Integrating valid and relievable disability measurement tools into existing data platforms is key to ensuring that “no one is left behind.” In this paper, we aim to demonstrate that it is possible to gather valid data on disability for disaggregation using the WHO Functioning and Disability Disaggregation Tool.Materials and methods: Using representative data from India, Lao PDR, and Tajikistan collected through the Gallup World Poll, we estimated the likelihood of a positive sustainable development indicator by disability level. Logit regression was used, adjusted for age, sex, household size, number of children, marital status, urban or rural area, and country-fixed effects.Results: Our estimates showed a consistent disability gradient across all countries and indicators: the higher the level of disability, the lower the probability of having a positive outcome in barely any sustainable development goal.Conclusion: Our study demonstrates that it is not too late to generate sound and precise data about inequalities faced by persons with mild, moderate, or severe disability. This data is essential for reducing inequalities through evidence-based policymaking.

Frequency-focused sound data generator for fault diagnosis in industrial robots

Abstract A frequency-focused sound data generator was developed for the in situ fault sound diagnosis of industrial robot reducers. The sound data generator, based on a conditional generative adversarial network, selects a target frequency range without relying on domain knowledge. A sound dataset of normal and faulty harmonic drive rotations of in situ industrial robots was collected using an attachable wireless sound sensor. The generated sound data were evaluated based on the fault diagnosis accuracy of a simple classifier trained using the generated data and tested using real data. The proposed method well-defined the frequency feature clusters and produced high-quality data, exhibiting up to 16.0% higher precision score on normal and 13.0% higher accuracy on weak-fault harmonic drive compared with the conventional methods, achieving fault diagnosis accuracy of 95.6% even in situations of fault data comprising only 5% of the normal data.

P-717 Add-ons availability and quality of online information provided by Brazilian IVF clinics

Abstract Study question How do Brazilian fertility clinic websites advertise four IVF add-ons: assistedhatching (AH), preimplantation genetic testing for aneuploidies (PGT-A), ERA and sperm DNA fragmentation test. Summary answer Available information, about ivf add-ons in Brazil, on researched websites is heterogeneous and is not evidence-based according to the last published recommendations. What is known already Over 1 billion people worldwide use online resources to obtain health-related information. People struggling with infertility are increasingly turning to online resources for infertility-related content and social support. “Add-on” procedures are actively promoted on some fertility clinic websites as means to improve IVF success rates. However, the actual contribution of these interventions to IVF success rates remains debatable. ESHRE has recently issued 42 recommendations were formulated on the use of add-ons in IVF. At present, no study has evaluated the type and quality of the information provided on the Brazilian IVF clinics’ websites regarding the merits of “add-ons. Study design, size, duration Prospective evaluation of the availability and quality of information on four common add-ons provided by IVF clinics in the southeast of Brazil between October-November 2023. Clinics were identified using the national registry (SisEmbrio) as well as through manual online search. As only public online available information was used no ethical approval was necessary. Statistical analyses were restricted to descriptive of the frequency of each add-on and frequency of claims made in relation to their use. Participants/materials, setting, methods One reviewer took screenshots of the included clinic pages and claims and a second reviewer double-checked everything. The folowing information was colletced: availability of of the four add-ons; information on the effectiveness on: implanation rate(IR), pregnancy rate (PR) or live-birth rate (LBR); information on the scientifc uncertainty of the use of add-on and possible negative impact; for PGT-A if the clinic informed that the data obtained should be confirmed by prenatal diagnostic testing. Main results and the role of chance The websites of 122 IVF clinics were included. Nine clinics had no website or were offline. PGT-A was the most frequent available add-on (n = 77; 68%) but only two of them (2,5%) informed on PR. There was no information on IR nor LBR. Thirteen clinics (16,8%) acknowledged it could have adeverse effects but only one mentioned the scientifc uncertainty related to its use. Forty (35%) websites displayed AH with 7 reporting on possible negative effects but none disclosed success rates (PR, IR or LBR). Eight clinics admitted AH had uncertain benefits and seven noted possible negative impact. The sperm DNA fragmentation test was presented in 37 (33%) websites but none reported on potential adverse effects nor success rates in terms of PR, IR nor LBR. No website provided information on the uncertainty of the benefit of the test. The ERA-test was advertised in 28 (25%) websites with no information on its benefits on sucess rates. Three clinics advised it could have a negative impact and six agreed its effectiveness was unclear. The cost was never pesented for any of the add-ons researched in accordance to national regulations. Of note, 36 websites advertised other add-ond such as MSOME, ZYMOT and endometrial scrathcing. Limitations, reasons for caution The study was restricted to 122 clinics websites in the southeast of Brazil. Evalaution was restricted to four commons add-ons and the quality of the information provided could not be assessed as the clinics follow no specific regulation on how to advertise scientifc information to their patiens and general public. Wider implications of the findings The information provided by Brazilian IVF clinics could be misleading. The use of add-ons is surrounded by scientific uncertainly related to their benefits and cost-effectiveness therfeore it is of utmost importance that clinics reviewt heir websites and offer sound scientific data to patients and the general public as well. Trial registration number Not applicable

Contrasting effect of hybridization on genetic differentiation in three rockfish species with similar life history.

Hybridization can provide evolutionary benefits (e.g., population resilience to climate change) through the introduction of adaptive alleles and increase of genetic diversity. Nevertheless, management strategies may be designed based only on the parental species within a hybrid zone, without considering the hybrids. This can lead to ineffective spatial management of species, which can directly harm population diversity and negatively impact food webs. Three species of rockfish (Brown Rockfish (Sebastes caurinus), Copper Rockfish (S. auriculatus), and Quillback Rockfish (S. maliger)) are known to hybridize within Puget Sound, Washington, but genetic data from these species are used to infer population structure in the entire genus, including in species that do not hybridize. The goal of this project was to estimate the hybridization rates within the region and determine the effect of hybridization on geographic patterns of genetic structure. We sequenced 290 Brown, Copper, and Quillback rockfish using restriction-site associated DNA sequencing (RADseq) from four regions within and outside Puget Sound, Washington. We show that (i) hybridization within Puget Sound was asymmetrical, not recent, widespread among individuals, and relatively low level within the genome, (ii) hybridization affected population structure in Copper and Brown rockfish, but not in Quillback Rockfish and (iii) after taking hybridization into account we found limited directional dispersal in Brown and Copper rockfish, and evidence for two isolated populations in Quillback Rockfish. Our results suggest that rockfish population structure is species-specific, dependent on the extent of hybridization, and cannot be inferred from one species to another despite similar life history.

Digital transformation, blockchain, and the music industry: A review from the perspective of performers’ collective management organizations

This study examines the challenges related to the music industry's digital transformation and potential role of blockchain from the perspective of collective management organizations (CMOs). Building on desk research and primary data from semi-structured surveys conducted with C-level executives and managers, this empirical analysis identifies major projects, their state of development, and prospects for digital transformation based on blockchain in the music industry. The findings reveal that there are a limited number of blockchain projects led by and/or with the relevant participation of CMOs. However, most are just research projects, proofs of concept, or pilots, showing that blockchain is currently in an experimental phase of development on the periphery of the music industry's digital transformation. This is not very different from analysts' understanding of the current situation and perspectives on the mass adoption of blockchain in other industries. In summary, blockchain is neither at the core of the music industry's digital transformation nor a priority for CMOs leading this process from the perspective of intellectual property rights management. The limited quality of music metadata, sparsity of sound data sources, and absence of a common data governance framework among different stakeholders within the industry are the main impediments to transforming blockchain into a global solution with transformative potential. Overall, the results provide a snapshot of the current status and possible future trajectories of blockchain adoption as a paradigm for intellectual property rights exchange within the music industry.

Wind direction and speed calculation model with deep learning

Abstract In view of the problems of complex process and weak representation in regular atmospheric wind field detection, this paper adopts deep learning method, uses global high-altitude meteorological detection data, and establishes a deep learning model for calculating wind direction and speed with different altitudes and temperatures by using keras software package. The model is verified by using third-party independent sounding data from a meteorological observatory in Shanghai. The calculation accuracy of the model above 2000 m is 0.9830 and the value of loss function is 0.0482. The accuracy under 2000 m is 0.9164 and the value of loss function is 0.0377. There are significant differences in the performance of the model between under 2000 meters and above 2000 meters due to surface friction. The model shows that wind direction and speed of different height layers can be calculated by using only height and temperature at the same height. This model can also be used to check whether the quality of regular wind detection work is good or not with big old data.

Prediction method of human defecation based on informer audio data augmentation and improved residual network

Defecation care for disabled patients is a major challenge in health management. Traditional post-defecation treatment will bring physical injury and negative emotions to patients, while existing pre-defecation forecasting care methods are physically intrusive. On the basis of exploring the mechanism of defecation intention generation, and based on the characteristic analysis and clinical application of bowel sounds, it is found that the generation of desire to defecate and bowel sounds are correlated to a certain extent. Therefore, a deep learning-based bowel sound recognition method is proposed for human defecation prediction. The wavelet domain based Wiener filter is used to filter the bowel sound data to reduce other noise. Statistical analysis, fast Fourier transform and wavelet packet transform are used to extract the integrated features of bowel sound in time, frequency and time-frequency domain. In particular, an audio signal expansion data algorithm based on the Informer model is proposed to solve the problem of poor generalization of the training model caused by the difficulty of collecting bowel sound in reality. An improved one-dimensional residual network model (1D-IResNet) for defecation classification prediction is designed based on multi-domain features. The experimental results show that the proposed bowel sound augmentation strategy can effectively improve the data sample size and increase the sample diversity. Under the augmented dataset, the training speed of the 1D-IResNet model is accelerated, and the classification accuracy reaches 90.54 %, the F1 score reaches 83.88 %, which achieves a relatively good classification stability while maintaining a high classification index.

The efficiency assessment of using hybrid neural networks for the detection of forged audio data in socially oriented systems

The subject of the research is the problem of detecting falsified data, in particular in audio format, in socially oriented systems. The goal of the work is to develop an effective model based on recurrent and convolutional neural networks for determining the fact of forgery of sound data, using MapReduce technology for parallelization. The article addresses the following tasks: determining the features of audio in socially-oriented systems, conducting an analysis of algorithms for processing audio information both in the form of text and in the form of a signal, forming a list of target architectures of neural networks and revealing the features of their implementation, conducting an experimental test of effectiveness selected approaches. The following methods used are – analytical and inductive method for determining the target set of neural network architectures; expert assessment for the formation of the most influential efficiency factors; experimental, multi-criteria evaluation and statistical methods of data augmentation to determine the most effective model. The following results were obtained: an audio data reprocessing algorithm was developed for the possibility of using recurrent and convolutional networks. Several approaches to data classification using augmentation based on vector autoregression and MapReduce parallelization technology have been implemented. It was determined that the most effective model for the multi-criteria selection problem is a combination of a bidirectional recurrent neural network with support for short- and long-term memory with several convolutional networks. The advantages of using MapReduce technology to optimize training time and data processing are shown, and a set of open questions for further research and applied implementation is defined. Conclusions: he application of an analytical and inductive approach followed by experimental verification made it possible to develop an effective (with an accuracy of more than 96%) a mechanism for detecting fabricated data both in the form of a signal and in text form. The obtained result makes it possible to assert the feasibility of implementing the proposed approach, and, accordingly, makes it possible to reduce the influence of such information in socially oriented systems, especially during crisis events.

Effect of Additive (1-Butyl-1-methyl Pyrrolidinium Tetrafluoroborate) on Volumetric, Acoustic and Conductance Properties of Binary Aqueous Solutions of Benzylamine at T = (288.15, 298.15, 308.15, 318.15) K

Effect of additive, 1-butyl-1-methyl pyrrolidinium tetrafluoroborate [BmPyrr+][BF4–] on the volumetric, acoustic and conductance properties of binary aqueous solution of benzylamine has been analyzed at equidistant temperature range using different techniques. Using density and speed of sound data, volumetric and acoustic parameters like apparent and partial molar volumes (Vφ and Vφ0), apparent and partial molar isentropic compressions (Kφ,S and K0φ,S), partial molar volume of transfer and partial molar isentropic compression of transfer (ΔVφ0 and ΔK0φ,S) have been calculated. Conductance parameters like molar conductance (Λm), limiting molar conductance (Λ°m) and activation energy are also calculated using conductance data. All these parameters were used to analyze various types of interactions present in the system. Furthermore, density functional theory (DFT) calculations were conducted for comprehensive understanding of structural variations of ion pairs affecting their physical properties. The hydrogen bond formation in the mixture components was examined by using IR spectrum. Both DFT and Hartee-Fock (HF) methods were used to analyze the results. In addition, the DFT/B3LYP-D3 calculations were also carried out to gather data on the molecular geometry, interactions and other molecular characteristics of the ionic liquid under study.

The thermodynamic behavior of binary systems isoamyl acetate and alcohol through experimental measurements and modeling

The thermodynamic and transport properties are experimentally determined for three binary liquid mixtures consisting of isoamyl acetate and different alcohols at ambient pressure and in temperature range. From measured density, viscosity and refractive index are calculated excess molar volume, viscosity and refractive index deviations. Isentropic compressibility and excess molar Gibbs free energies of viscous flow are determined combining density with speed of sound and viscosity data. All mentioned derived properties are used to analyze the changes taking place in investigated binary systems causing deviations from the ideal mixture. Additionally, partial molar volumes and excess partial molar volumes, as well as thermal expansion coefficient, were calculated. The results are discussed in terms of chemical structure of used compounds and the influence of the alcohol present in the mixture on the mixture behavior. The study showed that higher influence has branching of the alcohol molecule and position of the –OH group rather than the length of the primary alcohol chain. Beside the experimental part, the modeling of the excess molar volume and viscosity was also performed, and the percentage deviations are used as the indicator of model performance in terms of model approach, predictive or correlative, and number of model parameters, one to three parameters optimized.

Моделирование КВ-радиотрасс на основе волноводного подхода

We present a scheme for modeling HF radio signal characteristics along paths of different lengths, which is based on the waveguide approach — the normal mode method. We use a representation of the recorded signal field in the form of Green function products of the angular operator, excitation coefficients, and reception coefficients of individual normal modes. Algorithms have been developed for calculating distance-frequency, frequency-angular, and amplitude characteristics of signals in large spatial regions through analysis and numerical summation of normal mode series. We have implemented a complex algorithm for simulating propagation conditions of HF radio signals, which includes a medium model, algorithms for calculating signal characteristics, and operational diagnostics of radio channel. We have compared the results of the HF signal propagation characteristic modeling and the experimental oblique sounding data obtained along paths of different lengths and orientation. To analyze experimental ionograms, determine the maximum usable frequencies for propagation modes along radio paths, we employ the method of automatic processing and interpretation of oblique sounding ionograms.

HF radio path modeling by waveguide approach

We present a scheme for modeling HF radio signal characteristics along paths of different lengths, which is based on the waveguide approach — the normal mode method. We use a representation of the recorded signal field in the form of Green function products of the angular operator, excitation coefficients, and reception coefficients of individual normal modes. Algorithms have been developed for calculating distance-frequency, frequency-angular, and amplitude characteristics of signals in large spatial regions through analysis and numerical summation of normal mode series. We have implemented a complex algorithm for simulating propagation conditions of HF radio signals, which includes a medium model, algorithms for calculating signal characteristics, and operational diagnostics of radio channel. We have compared the results of the HF signal propagation characteristic modeling and the experimental oblique sounding data obtained along paths of different lengths and orientation. To analyze experimental ionograms, determine the maximum usable frequencies for propagation modes along radio paths, we employ the method of automatic processing and interpretation of oblique sounding ionograms.

A multimodal multitask deep learning framework for vibrotactile feedback and sound rendering

Data-driven approaches are often utilized to model and generate vibrotactile feedback and sounds for rigid stylus-based interaction. Nevertheless, in prior research, these two modalities were typically addressed separately due to challenges related to synchronization and design complexity. To this end, we introduce a novel multimodal multitask deep learning framework. In this paper, we developed a comprehensive end-to-end data-driven system that encompasses the capture of contact acceleration signals and sound data from various texture surfaces. This framework introduces novel encoder-decoder networks for modeling and rendering vibrotactile feedback through an actuator while routing sound to headphones. The proposed encoder-decoder networks incorporate stacked transformers with convolutional layers to capture both local variability and overall trends within the data. To the best of our knowledge, this is the first attempt to apply transformer-based data-driven approach for modeling and rendering of vibrotactile signals as well as sounds during tool-surface interactions. In numerical evaluations, the proposed framework demonstrates a lower RMS error compared to state-of-the-art models for both vibrotactile signals and sound data. Additionally, subjective similarity evaluation also confirm the superiority of proposed method over state-of-the-art.

Machine learning-based weld porosity detection using frequency analysis of arc sound in the pulsed gas tungsten arc welding process

Automatic welding equipment has replaced human welders in the nuclear industry for safety issues and uniform and high welding quality. However, automatic welding equipment cannot predict porosity defects. So, the weldment must be inspected by non-destructive testing. This inspection was a costly and time-consuming process, and it applies to each weldment even if it welded same material. To improve the welding efficiency, a weld porosity detection system of the same weld material with different material thicknesses was needed. This paper proposed a machine-learned porosity detection system for 3.0 mm plates with welding arc sound data from the pulsed gas tungsten arc welding (P-GTAW) process of 1.6 mm plates. Ensemble-Empirical Mode Decomposition (EEMD) was used to divide the arc sound signal according to the pulse period of P-GTAW. Fast Fourier transform (FFT) was used to convert the arc sound into frequencies for features extraction according to porosity. The validity of these weld frequency features was confirmed through k-fold cross-validation across various machine learning techniques, with evaluation of F-1 scores against experimental weld sounds.

Revised clusters of annotated unknown sounds in the Belgian part of the North sea

Acoustic signals, especially those of biological source, remain unexplored in the Belgian part of the North Sea (BPNS). The BPNS, although dominated by anthrophony (sounds from human activities), is expected to be acoustically diverse given the presence of biodiverse sandbanks, gravel beds and artificial hard structures. Under the framework of the LifeWatch Broadband Acoustic Network, sound data have been collected since the spring of 2020. These recordings, encompassing both biophony, geophony and anthrophony, have been listened to and annotated for unknown, acoustically salient sounds. To obtain the acoustic features of these annotations, we used two existing automatic feature extractions: the Animal Vocalization Encoder based on Self-Supervision (AVES) and a convolutional autoencoder network (CAE) retrained on the data from this study. An unsupervised density-based clustering algorithm (HDBSCAN) was applied to predict clusters. We coded a grid search function to reduce the dimensionality of the feature sets and to adjust the hyperparameters of HDBSCAN. We searched the hyperparameter space for the most optimized combination of parameter values based on two selected clustering evaluation measures: the homogeneity and the density-based clustering validation (DBCV) scores. Although both feature sets produced meaningful clusters, AVES feature sets resulted in more solid, homogeneous clusters with relatively lower intra-cluster distances, appearing to be more advantageous for the purpose and dataset of this study. The 26 final clusters we obtained were revised by a bioacoustics expert. We were able to name and describe 10 unique sounds, but only clusters named as ‘Jackhammer’ and ‘Tick’ can be interpreted as biological with certainty. Although unsupervised clustering is conventional in ecological research, we highlight its practical use in revising clusters of annotated unknown sounds. The revised clusters we detailed in this study already define a few groups of distinct and recurring sounds that could serve as a preliminary component of a valid annotated training dataset potentially feeding supervised machine learning and classifier models.