Table 5.
Mean intelligibility, acceptability, and phrasal stress classification accuracy scores (PSCA) for speakers with typical voices (T) and speakers with laryngectomy (L) according to primary voice source (natural, electrolarynx, or electrolaryngeal [EL]) and the synthetic voice matched to the speaker.
| ID | Primary voice source (natural, EL) |
Synthetic voice |
|||||
|---|---|---|---|---|---|---|---|
| Voice source | Intelligibility | Acceptability | PSCA | Intelligibility | Acceptability | PSCA | |
| T1 | Natural | 93.8 | 88.3 | 94.4 | 78.1 | 65.4 | 86.8 |
| T2 | Natural | 92.3 | 88.0 | 86.1 | 73.7 | 60.8 | 86.8 |
| T3 | Natural | 86.9 | 75.1 | 88.2 | 63.9 | 52.9 | 86.1 |
| T4 | Natural | 93.2 | 85.7 | 90.3 | 70.9 | 56.3 | 86.1 |
| L1 | EL | 54.3 | 21.8 | 53.5 | 72.7 | 58.9 | 78.5 |
| L2 | EL | 48.0 | 17.5 | 95.1 | 72.5 | 90.0 | 77.1 |
| L3 | EL | 51.3 | 21.6 | 45.8 | 66.6 | 51.5 | 81.3 |
| L4 | EL | 52.4 | 19.3 | 50.7 | 63.3 | 50.1 | 87.5 |