Visual-auditory visualisation of dynamic multi-scale heterogeneous objects.

Abstract

The multi-scale phenomena analysis is an area of active research that is connecting simulations with experiments to get a correct insight into the compound dynamic structure. Visualisation is a challenging task due to a large amount of data and a wide range of complex data representations. The analysis of dynamic multi-scale phenomena requires a combination of geometric modelling and rendering techniques for the analysis of the changes in the internal structure in the case of data coming from different sources of various nature. Moreover, the area often addresses the limitations of solely visual data representation and considers the introduction of other sensory stimuli as a well-known tool to enhance visual analysis. However, there is a lack of software tools allowing perform an advanced real-time analysis of heterogeneous phenomena properties. The hardware-accelerated volume rendering allows getting insight into the internal structure of complex multi-scale phenomena. The technique is convenient for detailed visual analysis and highlights the features of interest in complex structures and is an area of active research. However, the conventional volume visualisation is limited to the use of transfer functions that operate on homogeneous material and, as a result, does not provide flexibility in geometry and material distribution modelling that is crucial for the analysis of heterogeneous objects. Moreover, the extension to visual-auditory analysis emphasises the necessity to review the entire conventional volume visualisation pipeline. The multi-sensory feedback highly depends on the use of modern hardware and software advances for real-time modelling and evaluation. In this work, we explore the aspects of the design of visual-auditory pipelines for the analysis of dynamic multi-scale properties of heterogeneous objects that can allow overcoming well-known problems of complex representations solely visual analysis. We consider the similarities between light and sound propagation as a solution to the problem. The approach benefits from a combination of GPU accelerated ray-casting, geometry, optical and auditory properties modelling. We discuss how the modern GPU techniques application in those areas allows introducing a unified approach to the visual-auditory analysis of dynamic multi-scale heterogeneous objects. Similarly to the conventional volume rendering technique based on light propagation, we model auditory feedback as a result of initial impulse propagation through 3D space and its digital representation as a sampled sound wave obtained with the ray-casting procedure. The auditory stimuli can complement visual ones in the analysis of the dynamic multi-scale heterogeneous object. We propose a framework that facilitates the design of dynamic multi-scale heterogeneous objects visual-auditory pipeline and discuss the framework application for two case studies. The first is a molecular phenomena study that is a result of molecular dynamics simulation and quantum simulation. The second explores microstructures in digital fabrication with an arbitrary irregular lattice structure. For considered case studies, the visual-auditory techniques facilitate the interactive analysis of both spatial structure and internal multi-scale properties of volume nature in complex heterogeneous objects. A GPU-accelerated framework for visual-auditory analysis of heterogeneous objects can be applied and extend beyond this research. Thus, to specify the main direction of such extension from the point of view of the potential users, strengthen the value of this research as well as to evaluate the vision of the application of the techniques described above, we carry out a preliminary evaluation. The user study aims to compare our expectations on the visual-auditory approach with the views of the potential users of this system if it is implemented as a software product. A preliminary evaluation study was carried out with limitations imposed by 2020/2021 restrictions. However, it confirms that the main direction for the visual-auditory analysis of heterogeneous objects has been identified correctly and visual and auditory stimuli can complement each other in the analysis of both volume and spatial distribution properties of heterogeneous phenomena. The user reviews also highlight the necessary enhancements that should be introduced to the approach in terms of the design of more complex user interfaces and consideration of additional application cases. To provide a more detailed picture on evaluation results and recommendations introduced, we also identify the key factors that define the user vision of the approach further enhancement and its possible application areas, such as users experience in the area of complex physical phenomena analysis or multi-sensory area. The discussed in this work aspects of heterogeneous objects analysis task, theoretical and practical solutions allow considering the application, further development and enhancement of the results in multidisciplinary areas of GPU accelerated High-performance visualisation pipelines design and multi-sensory analysis.