Modeling perceptual dimensions of synthetic speech in High-SNR public environments: An SD-based evaluation of clarity, comfort, and preference | Synapse