SELVI SUPPORT

Back to your next take.

From your first script to your finished video. Find a quick answer or get in touch with the developer.

Talk to us

Include your device model, iOS and Selvi versions, and a short description of the issue. For voice following, include your script language. You do not need to send private scripts or recordings.

These instructions cover the upcoming Selvi 3.0. The App Store currently offers version 2.2 with a different interface. Using that version? Email us for help.

How do I create or edit a script?

Tap + in the script library, enter a title and your text, then tap Save. Both fields must contain text. To edit a saved script, open its three-dot menu and choose Edit. Tap the script card or Open camera when you are ready to record.

How do I move, resize, or adjust the text window?

Drag the top bar to move the window. Drag its bottom-right corner to change its width and height. Open Text settings to adjust text size, scroll speed in Timed mode, and background opacity. Use Reset window position to restore the default layout. You can also scroll the script with your finger.

What is the difference between Timed and Follow voice?

In Text → Playback, choose Timed for a fixed scroll speed or Follow voice to advance with recognized speech. Rehearse lets you practice before filming. Starting a recording restarts the script from the beginning. Pausing the text, or reaching its end, does not stop the video: use the recording button to finish filming.

Does voice following need an internet connection?

The first setup may need internet to download a speech model. Selvi uses on-device Apple recognition where supported, or a multilingual Whisper model of about 150–220 MB, plus additional preparation files. After the required model is downloaded and prepared, recognition can work offline. Your scripts, timed scrolling, and video recording do not need an internet connection.

Which languages can I use for voice following?

Open Text → Playback → Follow voice → Script language to see the available languages. Automatic detects one main language from your script; choose it manually for short or mixed-language scripts. Whisper includes 99 languages, including English and Ukrainian, but recognition quality varies with language, accent, noise, and device. The script language is independent of the interface language.

What if voice following pauses or falls behind?

Check the script language and wait for model preparation to finish. Speak clearly in a quieter space. Selvi may wait for more words when a match is uncertain. To choose a different position, tap a word or scroll, then press play. You can switch to Timed mode; a recognition error does not stop an ongoing video recording.

Why does Selvi need camera and microphone access?

The camera records your video, and the microphone records its audio and supplies speech for voice following. Both permissions are required for recording. If you previously declined, use the Settings button in Selvi to enable camera and microphone access in iOS Settings. Photos access is requested separately when you save a video there.

Where are recordings saved, and how do I share them?

When you stop recording, Selvi saves the finished video in the app and opens a preview. Choose Save to Photos to add a copy to your photo library, or use the share button to send or save the video file. Earlier recordings are available from My videos in the script library. If saving to Photos fails, the original recording remains in Selvi.

How do I delete a script or video, and can I recover it?

Delete a script from its three-dot menu. In My videos, swipe a recording and choose Delete. These deletions cannot be undone in Selvi. Deleting a video from the app leaves any separately exported copy in Photos. Removing Selvi from your device also removes its local scripts and recordings, so copy important text and export videos first.

Will the script appear in my saved video?

No. The text window and controls are visible while you film, but they are not added to the saved video. Your recording contains the camera image and microphone audio. Selvi does not turn the script into video captions.