Privacy Policy.

What this policy covers

Noctra is a media player developed by ZIJIU522. This policy covers the Noctra app, its feedback service, and the official website. “We” means the Noctra developer.

Noctra has no separate registration account and does not include a streaming catalog. You add the media sources the app accesses. A media server, Trakt, or another service may require its own account; its privacy policy also applies to the data it receives.

Playback creates local records and sends requests to the media sources you connect. Feedback, synchronization, and remote AI have their own data flows, described below.

Connecting media sources

Server addresses, accounts, sign-in tokens, proxy details, and connection settings are used to sign in, read your media library, and play files. Requests carry the authentication details, media identifiers, or file paths required by the service. They go to the server you select and the network services you configure.

Saved sign-in credentials are kept in the system Keychain. Separate sync settings control whether credentials are also synchronized through iCloud, as explained below.

HTTP sources transmit data without encryption. Use HTTPS where your server supports it, and check its address and certificate. HTTPS for feedback and custom AI does not encrypt an HTTP media-source connection.

What is saved on your device

To resume playback and retain your library settings, Noctra saves playback history, progress, favorites, subscriptions, watch-later records, local indexes, downloaded media, generated subtitles, and preferences as you use those features. Artwork, network responses, and prefetched playback data may also be cached locally.

You grant access to local folders through the system file picker. The app reads media within that permission. If files move or permission changes, you may need to select the folder again. Removing a source disconnects it from Noctra; it does not delete remote files or close the server account.

If server playback reporting is enabled, media identifiers, progress, and playback state are sent to that server. Server records, local records, and offline copies are stored separately. Clearing one does not clear the others.

iCloud and Trakt sync

The main iCloud sync switch is off by default. When you turn it on, selected server and route records, favorites and subscriptions, appearance, and playback preferences sync through your Apple account. You can choose categories before enabling it.

The credentials category and credential-sync setting for new servers are selected by default. Review both before turning on the main switch. Credentials sync through iCloud Keychain only when the main switch, credentials category, and the individual server setting are all enabled. Sign-in credentials are not written into the ordinary sync document.

Trakt requires you to connect an account. Enabling its sync or reporting sends the relevant media identifiers, lists, watched status, progress, and playback events to Trakt. Media-server playback reporting is controlled separately.

Turning off sync or disconnecting a service stops the corresponding future operations. Manage existing records through Apple, Trakt, or your media server. Custom-AI sharing consent does not transfer to other devices through Noctra’s iCloud settings sync.

Title information, subtitles, and danmaku

Browsing, searching, or filling in title information sends media identifiers, search terms, and necessary request details to the services used by that feature and configuration. These may include TMDB information services and your selected gateway, TheTVDB, Fanart.tv, rating, subtitle, or danmaku services. Connecting a third-party account or posting a comment also sends the authentication details or content needed for that action.

The service receiving a network request usually receives your IP address, request time, and request headers too. It may process data outside your region. Check the privacy policy of the service you use.

When you search NOC Studio in the app, or choose the TMDB gateway provided by Noctra, requests pass through our services. Access logs retain the requested path and its search or filter parameters, source IP address, request time, User-Agent (client type), and any referrer information carried by the request. These records remain after the request completes and are used to investigate access failures and abnormal requests, not advertising tracking. With a direct TMDB connection, the corresponding metadata requests do not pass through this gateway.

Noctra does not use these requests for advertising tracking. Disabling an optional service stops future requests for that service. Use its data-management channels for information it has already stored.

Subtitle processing and remote AI

On-device processing does not send dialogue to a custom AI endpoint. Custom AI is off by default. Before sending content, the app requires separate permission for the full endpoint address, model, and data type. Reading or accepting this policy, the Terms of Use, or the User Agreement does not grant that permission. Neither does saving a key or testing configuration. A different endpoint or model requires new permission.

Remote translation sends subtitle text and limited preceding dialogue. Remote speech recognition sends short audio segments from the video being played, not microphone recordings. Requests also include language and model settings and the API key you supply. The receiving service sees your IP address.

Custom endpoints must use HTTPS, and requests do not follow redirects. API keys are saved in the device Keychain. The provider sets its own data-retention, model-training, and billing rules; read them before granting permission. Noctra does not promise on the provider’s behalf that content will not be stored or used for training.

You can revoke permission in Custom Translation API settings to stop future requests. This does not delete subtitles or audio the provider has already received. Contact that service to address those records.

System permissions

System permissions are used for the features that need them. You can change permissions in iOS or iPadOS Settings. Turning one off affects the features that depend on it.

• Local Network: access to media servers and shared sources on your local network.
• Photos: access to permitted screenshots or recordings while preparing feedback. You can also choose attachments through the system picker. A screenshot or capture notification does not contain the media file.
• Speech Recognition: the Apple on-device recognition path processes video audio after authorization. Real-time subtitles do not record the microphone.
• Face ID or device authentication: the system unlocks Private Space. Noctra does not receive facial templates or device passcodes.
• The current version does not use the camera to take photos or record video.

Diagnostic logs and the feedback credential

The app records diagnostic information locally to investigate faults and filters known credential fields. You choose whether to export logs or send them with feedback. Automatic filtering may miss private content, so review log text, paths, and any images or files you add before sharing.

The feedback service uses a randomly generated installation credential stored in the device Keychain to link your reports and retrieve history and replies. It does not contain a hardware serial number and is not an advertising ID. Reports using the same credential can be linked, so feedback is not completely anonymous.

This credential identifies your installation to the feedback service and is not synced through iCloud. The service receives the request’s IP address and derives identifiers separately from the installation credential and IP address to link reports and limit abusive requests, not for advertising tracking. Keychain entries may survive reinstallation. Do not share the credential publicly.

What sending feedback uploads

After a screenshot or screen recording ends, Noctra may show a feedback hint. You can disable it on the feedback screen. The system notification tells the app about capture state only. Accessing the Photos file still requires permission or your manual selection; the notification does not trigger an upload.

When you tap Send, your description, optional contact details, device information shown on the feedback screen, selected media attachments, and selected redacted logs are sent over HTTPS to Noctra’s feedback service. We use them to investigate, handle, and reply to the report. Before sending, you can remove attachments or deselect logs. Contact details can be left blank.

Viewing feedback history or replies uses the installation credential to retrieve your records and read status. It does not upload local attachments you have not submitted. Feedback information is not used for advertising or tracking.

Attachments may contain playback images, subtitle text, or someone else’s information. Send only what is needed to investigate the issue. Keep the feedback reference number for follow-up or a deletion request.

Storage and deletion

Online feedback reports and attachments are cleared under a 30-day retention policy, counted from when the report was created. Uploads that were started but not submitted become eligible for cleanup after 24 hours. Cleanup runs on a schedule, so reaching a deadline does not mean immediate deletion at that moment. You can request early deletion with your reference number. Report cleanup includes associated attachments, replies, and association records in the online database; it does not remove the installation credential from your device Keychain.

Database backups kept during maintenance may still contain past feedback. Clearing online records does not delete backup copies at the same time. Include your feedback reference number in a deletion request so the related online records and backup contents can be checked.

Local records remain until you remove them or their storage rules clear them. Artwork, network caches, precached video, offline models, and saved subtitles are managed separately. Clear All Caches only clears the caches managed on that screen, not all user data. Turning off subtitle-result saving does not erase previously saved subtitles.

Deleting the app removes local app data, but Keychain entries, system backups, and cloud copies may remain. Removing a server connection does not close its remote account. Turning off iCloud, Trakt, or AI permission does not delete records already held by those services. Use the controls of the service holding the data, or contact that service.

Access logs for the website, NOC Studio API, and the TMDB gateway provided by Noctra are stored separately from feedback reports. The current configuration checks rotation daily and retains the current log plus up to 10 rotated files. Empty logs are not rotated, so this is not a promise to delete every entry exactly 10 days after it was created. Deleting a feedback report does not also clear these access logs. Third-party records are managed by the corresponding service.

The website and external links

The website provides product information, screenshots, help, and policies. It currently has no advertising, visitor-analytics scripts, or contact-information forms. Chinese and English use different page addresses; cookies and local storage are not used to remember a language choice. A light or dark appearance choice is stored only in this browser’s localStorage; selecting “System” removes that record. The preference is not sent to our server or used for tracking.

The web server receives the IP address, browser request headers, requested path, and time needed to serve pages. It keeps access and error logs to investigate access failures and abnormal requests. Logs rotate and are cleared under the hosting configuration. Contact us through the support link below with questions about these records.

Discover preview uses our API to read the selected list’s data source. Successful results are cached on the server for up to five minutes. Search and filter values appear in the page URL and may be recorded with request paths in server logs. Douban posters are requested through our image service and cached for up to one hour; the image provider receives our server’s IP address and its required request headers. Your browser loads other posters and backdrops directly from their image provider, which receives your IP address and request headers; browser image requests omit the website Referer. Module source, configuration, content keys and provider credentials remain on the server. Public downloads contain encrypted packages only.

When you open an external link such as GitHub, that service handles visits and submissions under its own privacy policy. GitHub Issues are public. Do not post credentials, private attachments, or playback links containing authentication information there.

Privacy requests and policy updates

To ask about, correct, or request deletion of feedback information, or to ask about this policy, open Noctra → Settings → Feedback. Describe the data and the action you are requesting. Include the feedback reference number for a report, not server passwords, API keys, or the installation credential.

Use the same route if a report includes someone else’s private information or media sent by mistake. We will use the details you provide to locate the report. Local, cloud, and third-party data need to be handled separately. If the app will not open, the support page lists a public contact route; leave out private information.

We update this page and its date when features or data handling change. New sharing still follows the relevant permission steps in the app. Publishing a new policy does not turn on sync, send feedback, or authorize remote AI for you. The Chinese and English versions describe the same practices.