Background
Semi-occluded vocal tract exercises (SOVTE) are widely used in both clinical rehabilitation as well as habilitation of voice for pathological and healthy professional voice users. SOVTEs have been shown particularly useful in the treatment of dysphonia. However, potential contraindications for SOVTEs are sparsely studied, with some clinical results suggesting that fatigue and tiredness may result from SOVTEs in both patients and singers, at least short-term. A recent study demonstrated how SOVTEs increased vocal loading in a healthy male producing loud phonation using various SOVTEs.

Objective:
To study the effects of several SOVTEs in 2 healthy professional singers, one male and one female, producing a variety of voice conditions at various loudness levels and qualities.

Methods:
Double-case study of a male and a female singer, both singers were recorded while performing p+vowel syllables. Airflow, oral pressure (Poral), audio, and EGG signals were recorded. Poral during p-occlusion was used to estimate subglottic pressure (Psub). A total of 8 different singing sounds varying in loudness and qualities were studied for all SOVTE interventions. The severity subscale for the Vocal Tract Discomfort Scale was taken before and after each SOVTE separately. Aerodynamic, EGG, and acoustic parameters at baseline as well as before, during, and after the following SOVTEs were recorded: 3mm straw, 10cmH20 Water Resistance Therapy, Shaker Deluxe at 5cmH20, and a voiced vvvf-fricative. Specifically for the 3mm straw, a continuous sample was recorded with phonation initiated and maintained without the straw, adding the straw, and removing the straw to capture baseline, SOVTE production and post-SOVTE maintenance, specifically. All other SOVTEs were practiced for 2 minutes prior to measurement to familiarise and condition the singers to the specific SOVTE production with 30 second repetition of the SOVTE before each 8 conditions in the protocol. Auditory perceptual assessment for each condition before and after each SOVTE was performed by a trained singing teacher.

Results:
VTDS increased from baseline 0 to 5 with the 3mm straw, while Shaker Deluxe at 5cmH20 increased to 3, the 10cmH20 WRT increased to 2, and the voiced fricative showed the least discomfort at a score of 1. Comparing baseline and SOVTE target aerodynamic targets, certain sounds and qualities aligned better with specific SOVTEs compared to others. Specifically, SOVTEs targeting higher airflow aligned better with softer-sounding target productions where increased ml/sec was observed, compared to louder and metallic-sounding target productions where decreased airflow and increased Psub was needed. Interestingly, for loud phonation where airflow was decreased by WRT and Shaker Deluxe interventions, Psub increased considerably, while VTDS increased. Some SOVTEs impacted the auditory perceptual production more than others.

Conclusions:
More research into the alignment between target singing output in terms of loudness and sound quality and appropriate SOVTE for the specific sound is warranted.