Voice therapy in general, and semi-occluded vocal tract (SOVT) therapy in particular, deal with two big challenges that will be addressed in this research proposal. The first big challenge is how to carry-over or transfer newly established vocal behaviors outside the clinic. To data, research mainly focused on the effects determined during or immediately after a SOVT exercise (= performance). It is unknown whether permanent changes in the vocal system occur (= learning). The most concrete way to determine permanent changes in the vocal system is to actually observe the level of the larynx and pharynx. Therefore, the first aim of this research is to investigate both immediate (performance), short-term (performance to learning) and long-term (learning) effects of three SOVT therapy programs (straw phonation in air, straw phonation in water - 2cm, straw phonation in water - 5cm) on vocal behavior of patients with functional dysphonia by direct visualization of the (supra)glottic activity in the larynx and pharynx using flexible laryngovideostroboscopy. The second big challenge is how to select the most appropriate exercise for a specific vocal pathology. In voice therapy, there is a lack of efficacy studies that compare different exercises in groups with different diagnoses. Therefore, the second aim of this research is to find the best matched SOVT exercises for specific vocal pathologies (diagnosis-specific treatment).