Feed Entities Podcasts Deep Dives Developers About

[KongchangAI]

Real-time AI knowledge base. Every claim is traceable to its source, with data freshness under 24 hours.

Powered by 240+ first-hand AI sources

Knowledge

Feed
Hot Entities
Podcasts

Developers

CLI
MCP Server
REST API
About

© 2026 KongchangAI. The freshest, most accurate AI knowledge

Unverified50% confidenceFactExact time

职业同声传译员的平均延迟（EVS, Ear-Voice Span）约为2-5秒

1

Sources

50%

Confidence

Long-term

Relevance

7/20/2026

First Seen

Sources

OpenAI新语音模型：实时多人多语言同声传译全解析

Related Claims

Unverified优秀的同声传译员会维持约2-3秒的认知缓冲（Ear-Voice Span），在语义完整节点触发输出78% similar Unverified双向语音对讲存在约0.5-1.5秒延迟，半双工机型（对讲时听不到对方）不适合与老人实时沟通，看护老人应选全双工对讲机型75% similar Unverified双向语音对讲存在延迟，主流机型延迟0.5-2秒，回声消除(AEC)功能影响通话体验；看护沟通场景应选带全双工通话（可同时说话）的机型，半双工需轮流讲话不利紧急沟通75% similar Unverified双向语音（对讲）功能实测延迟差异大，选延迟＜0.5秒的型号才能有效隔空安抚哭闹；部分低端款延迟达2-3秒，安抚时机滞后失去意义74% similar Unverified传统语音AI采用串行管线设计，即VAD→ASR→LLM→TTS四个独立模块依次传递数据，每个环节引入约100-300毫秒延迟，四段叠加总延迟往往超过1秒73% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/573858

API

curl https://kongchang.com/api/v1/knowledge/claims/573858

MCP

get_claim(id=573858)