Over the recent years, it became more evident in soundscape studies realised in urban parks the importance of natural sound sources as a restorative resource for human health and well-being. This study investigates psychoacoustic and bioacoustics patterns of natural sound sources in urban soundscapes to understand and enhance restorative environments. Four areas with garden and park functionalities in Aachen, Germany, were investigated. In one of the surveyed areas, it is possible to verify seasonality differences. Sound classification is possible with the help of Machine Learning techniques, subdividing the recordings’ sound sources into anthropogenic, geophonic, and biophonic sounds. Subjective responses help with the sound quality information of these sound sources. After the categorisation of sound sources, psychoacoustic and bioacoustics indicators are analysed. These indicators help to understand the sound qualities of natural sound sources in urban areas.