Talking to YO ATOM. Push-to-talk is the main way. The talk button works two ways: hold it, talk, and let go, or tap it, talk, and tap again. Either way your question sends itself and ATOM answers out loud, so a loud bus or venue never leaves it waiting for you to go quiet. Pressing the button while ATOM is talking cuts it off; in hands-free, saying "Yo ATOM" while it talks does the same and it listens to you. Anything ATOM offers to change still shows as a card you tap to save; nothing saves by voice. Hands-free is the optional mode: turn it on and say "Yo ATOM" to wake it, best in a quiet room. The reply follows the question: ask out loud and it answers out loud, type and it stays quiet. The speaker button is there if you want every answer read aloud. Money and personal details are not spoken aloud until you confirm; the answer is always on screen either way.
HOW IT LISTENS AND TALKS: signed in and online, what you say is turned into text by ATOM's listening service, the one that knows your tour's words; with no signal or no login, your phone or computer's own dictation is used instead, in your ATOM language (English, Spanish or Korean). ATOM answers in its own voice and starts talking as soon as the first sentence is ready, while the rest is still being written.
IT KNOWS YOUR TOUR'S WORDS: when you talk, ATOM listens for your tour's own venues, cities, the airports for those cities and your crew's names, plus touring words like FOH, buyout, per diem and shore power. It only uses what you can already see on this tour. It never swaps a word on its own: if it heard something one letter off one of your tour's words, it asks first, out loud and on screen: "I heard VNA. Did you mean BNA, Nashville?" Tap Yes, No (keeps what it heard) or Edit (puts your question in the box to fix by hand). When you ask out loud, the answer starts with what ATOM understood ("For BNA, Nashville:") so a mishearing is easy to catch, and stays short; anything longer is on screen.
VOICE CHECK: Settings > YO ATOM > ATOM's Voice > Voice check tests it all on your device in about 15 seconds: ATOM's own voice, your microphone, and what ATOM hears when you say a line. Anything marked Problem, tap Copy results and send it to your TM or ATOM support.
WHAT IS NOT READ ALOUD: ATOM decides by what the information is, not by money-sounding words. Held back unless you allow it: money amounts (any currency, "bucks", cash, buyouts, pay, settlements, in English, Spanish or Korean), door and access codes, passwords and Wi-Fi keys, phone numbers, email addresses, your room number, and ID documents. Spoken normally: showtimes, the schedule, venue and hotel names, and production talk like "run a wire to the balance channel". Type COPY to hear a held answer, or open read-aloud in Settings > YO ATOM > Sensitive read-aloud when you're somewhere private.
VOICE AND YOUR ACCOUNT: voice runs on your own login and your own allowance, so a crew sharing one venue Wi-Fi can't use it up for each other. Turning ATOM Intelligence off also stops voice: nothing you say is sent anywhere. If hands-free can't hear you because the signal dropped or a limit was hit, ATOM tells you out loud and on screen instead of going quiet.
Numbers and pace: ATOM says money and big numbers in full words, exact to the cent ("one million, nine hundred forty-seven thousand, four hundred eighty-seven dollars and twelve cents"); the screen still shows the figures. In conversation mode it takes about one second of quiet as the end of your question (it used to wait almost two), and it pulls your tour details faster, so answers start sooner. In hands-free, ATOM now hears your words while you are still talking, so your question is ready the moment you stop (if the connection can't keep up, it uses the recording, the same as before). Long answers are spoken as one smooth answer: each sentence is voiced knowing what came before it, and anything in brackets is said as a plain pause.