Voice TTS EN pre-rendered

Adds voiced player proximity chat, train stop announcements and speakers. Plays sound based on source position. Uses pre-rendered english speech, no external software or internet required, leverages combining dataset of 24k most commons words. It can also create new words letter by letter. First launch after enabling this mod will take a minute due to sound prototype building. Text to speech, text-to-speech, voice chat, accessibility, messages, audio, fun.