Studies of timbre are usually conducted in a "vacuum" of perfect silence. However, in the real-world, sounds are mostly heard in the presence of competing background noise. A series of pairwise dissimilarity listening tests on musically trained participants demonstrated how different levels of background noise can cause rearrangement of timbre spaces. Furthermore, it was shown that while spectral acoustic descriptors (e.g. spectral centroid or tristimulus values) seem robust under the presence of background noise, descriptors representing deviations from purely harmonic characteristics (e.g. inharmonicity) lose their salience for the higher noise level. Such results suggest that studies of timbre may need to take background noise into account in order to enhance their validity for real world applications.

Mixing is a quintessential optimization problem. Given control of several component tracks, a balance must be struck that reflects a trade-off between engineering methods, artistic objectives, and auditory perceptual constraints. Formally, this balance can be thought of as the optimal solution to a system of mathematical equations that describe the relationships between the component tracks within a mix. Hence, the nature of these equations defines the process by which solutions may be arrived at. As perception is strongly nonlinear, an analytical solution to this set of equations is not possible and so a search must be conducted. Here, taking loudness as an example, we develop an optimization theory treatment of the problem of mixing, complete with case studies to illustrate how auditory perception can complicate the process, not least due to masking-related interactions.

In this tutorial article we provide a condensed, practical step-by-step guide to the excitation pattern loudness model of Moore, Glasberg and Baer [J. Audio Eng. Soc. 45, 224–240, 1997; J. Audio Eng. Soc. 50, 331–342, 2002]. The various components of this model have been separately described in the well known publications of Patterson et al. [J. Acoust. Soc. Am. 72, 1788–1803, 1982], Moore [Hearing (Academic Press), 161-205, 1995], Moore et al. (1997) and Glasberg and Moore (2002). This paper provides a consolidated and concise introduction to the complete model for those who find the disparate and complex references intimidating and who wish to understand the function of each of the component parts. Furthermore, we provide a consolidated notation and integral forms. This introduction may be useful to the loudness theory beginner and to those who wish to adapt and apply the model for novel, practical purposes.

In this paper, we describe the procedure to set up a cyber infrastructure to provide a cloud service to various academic institutions using the Xen Hypervisor according to the Platform as a Service (PaaS) model. The users of this service should be able to access the servers located in the Computer Network Systems and Security (CNSS) Laboratory at Jackson State University remotely. They will use this access in the cloud to run virtual machines and utilize the powerful server hardware. These virtual machines can either be created from user provided images or operating system images that have been previously stored on the server. In order to allow users to use these virtual machines some type of virtual machine manager is needed. For this project, the Xen Hypervisor was chosen to fulfill this need. The hypervisor was installed on two servers located in the laboratory and allows users to remotely log in and create their desired machines.

We examine the effect of listening level, i.e. the absolute sound pressure level at which sounds are reproduced, on music similarity, and in particular, on playlist generation. Current methods commonly use similarity metrics based on Mel-frequency cepstral coefficients (MFCCs), which are derived from the objective frequency spectrum of a sound. We follow this approach, but use the level-dependent auditory spectrum, evaluated using the loudness models of Glasberg and Moore, at three listening levels, to produce auditory spectrum cepstral coefficients (ASCCs). The ASCCs are used to generate sets of playlists at each listening level, using a typical method, and these playlists were found to differ greatly. From this we conclude that music recommendation systems could be made more perceptually relevant if listening level information were included. We discuss the findings in relation to other fields within MIR where inclusion of listening level might also be of benefit.

A model of live performance is presented that includes simplified acoustic-environmental system models and enables the coupling behavior between multiple sources and receivers to be predicted. The model allows the mix at each location within the performance space to be evaluated as a function of the acoustic signals generated by the instruments and of the control parameters on the mixing desk, through which the instrument signals are sent before being reinforced using loudspeakers. For a set of listener locations, which includes both performers and members of the audience, ideal mixes are defined, and an optimization algorithm is developed that sets the control parameters on the mixing desk automatically, to deliver approximations of these ideal mixes to all listener locations simultaneously. The control parameters are constrained during the optimization to prevent the onset of acoustic feedback. The algorithm is examined, and we show that a targeted approach, which first sets the control parameters relating the vocal level, gives a better solution in a shorter time when compared to a brute-force approach.

An algorithm is presented which automatically sets the attack, release, threshold, and hold parameters of a noise gate applied to drum recordings which contain bleed from secondary sources. The gain parameter which controls the amount of attenuation applied when the gate is closed is retained, to allow the user to control the strength of the gate. The gate settings are found by minimising the artifacts introduced to the desirable component of the signal, whilst ensuring that the level of bleed is reduced by a certain amount. The algorithm is tested on kick drum recordings which contain bleed from hi-hats, snare drum, cymbals, and tom toms.

Dynamic spectrum access is a technique to utilize the spectrum resources efficiently in a cognitive radio environment. To use spectrum efficiently various models were designed by researchers. Game theoretical models are few of the efficient techniques recently introduced in wireless communications. Game models are very useful for opportunistic selection of spectrum and help to utilize available spectrum efficiently. We introduced the spectrum underlay model and discussed the current state of allocation algorithms. We then introduced the congestion game model for opportunistic spectrum access with minimum interference. The theoretical results conclude that the congestion game model helps to use the underlay spectrum efficiently with minimum interference.

The image source method is used to identify early reflections which have a re-enforcement effect on the sound traveling within an enclosure. The distribution of absorptive material within the enclosure is optimised to produce the desired re-enforcement effect. This is applied to a monitor mix and a feedback prevention case study. In the former it is shown that the acoustic path gain of the vocals can be increased relative to the acoustic path gain of the other instruments. In the latter it is shown that the acoustic path from loudspeaker to microphone can be manipulated to increase the perceived signal level before the onset of acoustic feedback.

A model has been developed which describes the monitor mix experienced by each participant in a musical performance. Using this model, and the mix requirements of each performer, an objective function was defined. This function was then minimized to find the monitor mix settings that best replicate the target monitor mix of each performer, subject to the side constraints of maximum and minimum allowable sound pressure level and the prevention of acoustic feedback.

A method has been developed for automating the set- tings of a noise gate. The method has been applied to a kick drum track containing bleed from secondary drum sources and white noise. The optimal settings are found by maximising the signal to distortion ratio (SDR). The SDR has contributions from the distortion caused to the kick drum signal, and the residual bleed and noise. These two components are weighted, en- abling the gate to be controlled by a single parame- ter. It is shown that the improvement in the SDR can be obtained when the two components of the SDR are approximated, enabling the optimal settings to be cal- culated from the noisy signal and a single kick drum hit. It is found that the optimal threshold is slightly above the peak level of the noise component of the sig- nal.

Model updating is a powerful technique to improve finite element models of structures using measured data. One of the key requirements of updating is a set of candidate parameters that is able to correct the underlying error in the model. Often regions such as joints are very difficult to parameterise satisfactorily using physical design variables such as stiffnesses or dimensions. Parameters arising from generic element and substructure transformations are able to increase the range of candidate parameters, and furthermore are able to correct structural errors. However, unconstrained generic substructure transformations change the connectivity of the model matrices. In many instances retaining the connectivity is desirable and this paper derives constraint equations to do so. The method assumes that substructure eigenvalues are the parameters used in the global updating procedure and that the substructure eigenvector matrix is optimised to enforce the connectivity constraints. The method is demonstrated on a simple L shape test structure, where the substructure is the corner.