Abstract by W. Zachary Horton
W. Zachary Horton
Relating Perceived Audio Quality to Personal Factors
Modern techniques for recording music almost always involve converting live analog signals into digital files. Original studio files are often quite large and thus, are not suitable for commercial distribution. Before being sold through digital marketplaces or other media, these files are compressed using algorithms that take advantage of limitations in human hearing. Procedures such as reducing sampling rates can shrink file sizes by tens or hundreds of times and are designed to preserve overall sound quality; however, high enough levels of compression will begin to have an audible effect. Listening tests that assess the ability to discern between levels of audio quality across several genres were administered to volunteers. A generalized linear model is employed to relate perceived audio quality with factors such as age, gender, musical preferences, and prior musical training.