Abstract
Three methods of measuring coffee roast degree were compared using titratable acidity as an indicator of roast-dependent flavor change. The first roast degree method was based on prediction of the cracks with online near infrared spectroscopy and partial least squares regression, the second was based on changes in online near infrared absorbance, and the third was the common L* value from the CIELAB color space in the visible spectrum. Roasting trials utilized arabica coffee from eight origins in an air roaster, and results demonstrated the superiority of an online near infrared sensor for real-time roast degree measurement. A second dataset with constant temperature roasts showed how acidity can be controlled by changing both the roasting temperature and roast degree, finding the linear effects of roast time and roast degree on acidity.
Get full access to this article
View all access options for this article.
