This study is designed to refine vocal analysis techniques by integrating distinct vocal characteristics, known as vocal biomarkers (VB), which are indicative of a variety of health states. Conventionally, VB development has utilized "cued" voice samples—these are recordings captured when participants are prompted to produce sounds or speech based on specific guidelines. Established VB models have shown success in discerning between certain respiratory conditions (such as asthma, chronic obstructive pulmonary disease (COPD), and COVID-19) and healthy states, through the elicitation of sustained vowel sounds, particularly the "ahh" sound. Moreover, analyses of vocal features from 30-second segments of natural speech have been effective in differentiating levels of mental health symptomatology. This research intends to advance the application of VB methods to "passive" voice recordings. These recordings are captured via devices that continuously process audio data and identify individual users without active user engagement. By utilizing these passively obtained signals for VB analysis, the study aims to remove the requirement for overt participant interaction, thereby allowing the technology to function unobtrusively, much like current fitness tracking devices. To achieve this transition, it is necessary to adapt the established cued VB techniques to a passive collection environment. This adaptation will require the compilation of a new passive VB dataset, which will then be compared and validated against cued recordings from the same participants. The successful completion of this process is crucial to enhance the practicality and efficacy of VB technology for health state assessment.

 Â·       Respiratory Patients: 400 adults with Asthma (200) or COPD (200).

·       Healthy Volunteers: 100 adults without any chronic or acute health, especially respiratory conditions.

Criteria:

·       Age: 18+