Frequently asked questions

Voice check-in stops after each question and waits for your answer. Once your speech is captured, 1.5 seconds of detected quiet triggers a reply. Speaking again restarts the pause. Silence before any answer cannot advance the conversation. There is no send button; use Mute if you need a longer thinking break. Headphones help prevent the speaker audio from returning through your microphone.

Use Sign in in the header with the ChatGPT account that owns this demo. Admin controls are restricted to that account. In Frequency, choose your starting phase, feeling ratings, breathing rhythm, and spoken cues, then select Save my demo setup. Reset demo stops the current run, clears its transcript, voice results, exercise state, and unsaved drafts, restores your setup, and begins a fresh session. Saved songs, plans, and feeling snapshots are kept. Other users are unaffected. For Voice check-in, end the conversation and use Clear this check-in.

Voice check-in is a short conversation with a waveform, transcript, expression timeline, and an end-of-session summary. Frequency combines conversation, meaningful songs, optional breathing and reset exercises, and a small next-step plan. The overall session summary is currently available in Voice check-in.

Anyone with the link can open the demo. Saving Frequency feelings, a soundtrack, playlist link, or plan requires signing in with ChatGPT. Spotify listening may have its own account requirements; signing into this site does not connect Spotify.

No. In Frequency you can use the on-screen controls and choose channels in any order. Voice check-in needs microphone input to create voice-expression results. You can mute or end a conversation at any time.

Those channels use a separate, calm guide voice for conversation and soothing exercise cues. You can turn spoken exercise cues off.

Allow microphone access in your browser’s site settings and check the microphone selected by your device. Close other apps using it, then start again. If you see Enable agent audio, select it. Headphones can reduce echo. Refreshing will clear the current check-in results.

A check-in stops after five minutes. Network interruptions or the demo’s voice-service credits, usage limits, or simultaneous-session limit can also stop it. Follow the displayed message. If it refers to credits or a monthly limit, the demo owner needs to resolve it; changing your microphone will not fix that.

In The Intro or The Remix, name the song and artist, for example, ‘Here Comes the Sun by The Beatles.’ Check the song card, then say ‘save song’ or select Save song. You can also type or correct the fields. Wait for a successful saved confirmation. ‘Yes’ alone does not save a song.

Check that the card has both a title and artist and that you are signed in. Review any save error. If your profile did not load, use Retry loading. Your soundtrack holds up to five songs; remove one to make room. Saving the same title and artist again updates its link rather than adding another copy.

Not in this demo. Add a full Spotify track link, select Open player here, and press play in the embedded Spotify player. Find on Spotify opens a separate search so you can choose the right recording. Voice conversation stops while you listen.

Spotify determines what its player makes available for the recording, account, browser, and location. A preview or full song is not guaranteed for every link. If inline playback is unavailable, open the recording in Spotify.

The demo supports Spotify song and playlist links and embedded players. It does not currently connect your Spotify account, search its catalog for you, import playlist tracks, or control Spotify playback by voice. Save playlist link keeps the link only. The site does not ask for your Spotify password.

The microphone conversation is stopped so you can listen or follow the exercise. Use Close player or the exercise’s stop control, then start talking again when you are ready. Music is not analyzed as your speech.

Yes. Under Where are you today?, select all that fit and adjust each slider from 1 (slight) to 5 (very strong). A newly selected feeling starts at 3 until you adjust it. Select it again to remove it. Sign in and select Save feelings to record a snapshot for your account and current Frequency session. Save again to record a change. Your saved feelings shows dated snapshots without judging them. Voice commands add a named feeling without clearing your other selections; use the sliders to choose its strength.

Prosody is the pattern of pitch, rhythm, emphasis, and tone in speech. The demo displays estimated expression signals from your voice. These estimates cannot establish your actual feelings or provide a diagnosis.

No. Scores run from 0 to 1 and represent the model’s expression estimates. A score of 0.70 does not mean you are 70% likely to feel that emotion. In Frequency, the feelings and 1–5 strengths you choose yourself are kept separate from these estimates.

After a Voice check-in ends, the summary ranks expression signals by their average valid score across captured speech segments. Each segment has equal weight, regardless of duration. Missing or invalid scores are excluded. The summary also shows selected signals’ lowest and highest observed scores.

The chart compares neighboring segments using the average absolute change in scores shared by both segments. It highlights up to three of the largest nonzero changes in that session. These are descriptive comparisons, not judgments, statistical findings, or explanations. View the linked segments to read the words alongside their scores.

The check-in must end with at least one captured speech segment for the summary to appear. Comparisons need at least two segments with valid shared signals. With too little data, no shared scores, or no observed change, there may be nothing to highlight. If a connection stopped early, the summary covers only what was captured.

Voice check-in transcripts, expression charts, and summaries stay in the current tab. They are lost when you refresh, clear the check-in, or start a new one; there is no report download or session-history feature. When signed in, Frequency saves confirmed song metadata and links, your playlist link, your plan, and feeling snapshots you save to your account. Feeling snapshots include your selected feelings, 1–5 strengths, time, channel, and session. Each visit or admin reset starts a new session. Your saved feelings shows the latest 50 snapshots; unsaved choices are lost on leaving or refresh.

External voice services process microphone audio to support conversation and expression analysis. The app’s saved Frequency profile does not include audio, transcripts, or expression scores. Service providers have their own data handling; this demo does not promise immediate deletion or zero retention. Use mute or end the voice session to stop sharing microphone input.

This is an adult wellness-reflection demo. Its exercises and voice estimates are not a clinical assessment or treatment. The Support option offers ways to reach a person; the demo does not contact anyone, monitor your safety, or dispatch help.