Claims

XTTS and Orpheus achieve the highest human-rated MOS for Singaporean-accented ATC speech synthesis (3.71 and 3.70), with female voices preferred across models

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2029: tool and model accuracy changes quickly

Subclaims

Reading the evidence codes

q is the design tier, i the impact, n the sample, and the kind with its r the rigour judged within that kind: what each code means.

This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.

q2 i? In human MOS ratings, XTTS and Orpheus achieve the highest overall scores (3.71 and 3.70) while CSM trails across all dimensions due to truncations and incomplete utterance endings. → Ethan Chew 2026 q2 i? LLM-based scores follow the same ranking but diverge from human judgment, with XTTS showing the largest human-LLM gap (3.71 vs 3.37), and gender accuracy exceeds 96% for all models. → Ethan Chew 2026

Evidence

Ethan Chew 2026

Ethan Chew, Enjia Wu, Iruss Eng, Ian Lim, Ranen Sim, Brandon Koh, Kaleb Nim, Caden Toh, Wei Dong Soin, Darius Koh, Galen Tay, Prannaya Gupta, Jonathan Koong, Yong Zhi Lim. (2026). ASTRA: A Scalable Next-Generation ATCO Training Simulator with Autonomous Simpilots. https://arxiv.org/abs/2606.18319 doi:10.48550/arxiv.2606.18319

q2 · i? · design · r2

Perceptual MOS evaluation of three fine-tuned TTS models on five dimensions, rated by 21 aviation personnel familiar with ICAO radiotelephony plus an automated LLM evaluator (Gemini 2.5 Flash); each human rater evaluated 9 clips. The paper notes "CSMtrails both mod- els across all dimensions".

"As shown in Table 3,XTTSandOrpheusachieve the highest overall human MOS scores (3.71 and 3.70 respectively)."

Discussion

Page metadata
Field Value
type claim
id astra-tts-mos-xtts-orpheus-best
title XTTS and Orpheus achieve the highest human-rated MOS for Singaporean-accented ATC speech synthesis (3.71 and 3.70), with female voices preferred across models
description XTTS and Orpheus achieve the highest human-rated MOS for Singaporean-accented ATC speech synthesis (3.71 and 3.70), with female voices preferred across models
status review
evidence_strength weak
newest_evidence 2026
recheck_by 2029
generated by process:wiki-ingest on 2026-10-10
trust unverified — no one has confirmed this page's content
sources 1 — ethan-chew-2026

Frontmatter is the page's machine-readable record — see the Schema & Guide.