Creaky voice or “creak” is a type of voice quality in speech. Due to its acoustic properties, creak is often a major source for errors in fundamental frequency (F0) detection algorithms. For prosodic analysis, however, it is essential to avoid F0 detection errors that would lead to a distorted representation of one of the most important acoustic features in speech signal processing. A central part of this project will be the comparison of different F0 detection tools with special focus on their performance and validity in creaky patterns. The goal of this project is the implementation of a tool for automatic creak-detection and -labeling. Key questions will be whether it is possible to detect creak automatically, and to further categorize creak in terms of (a) different acoustic patterns and (b) their co-occurrence with communicative functions in conversations. This implies that there is not one single category of creak. An exploratory section of the project will be to investigate if different acoustic measures can be found which characterize those different kinds of creak.