# Choose voice-follow or fixed-speed scrolling

Understand what voice-follow does, its on-device requirements, and how to recover when speech recognition is unavailable.

Source: https://teleprompterfocus.com/guides/voice-follow-scrolling/

Updated: 2026-11-04

QUICK ANSWER

## Should I use voice-follow or fixed-speed scrolling?

Use fixed-speed for a rehearsed, steady pace and voice-follow when you want the prompt to advance from your spoken script. Voice-follow is paid and needs supported on-device speech recognition, microphone access and speech permission.

## Choose the mode for the take

Fixed-speed scrolling advances at the speed you set. It is included in the free tier and does not depend on recognising your voice. Rehearse once and adjust it until you can finish sentences comfortably.

Voice-follow compares recognised words with your script to advance the reading position. It can suit a delivery with varied pacing and pauses. It follows the written words; it is not a guarantee that the prompter will find your place after a long detour or repeated passage.

## Test voice-follow before a full recording

1.  Choose your script, then enable voice-follow from camera tools. A subscription or eligible trial is required.
2.  Allow Speech Recognition and Microphone access when prompted.
3.  Read a short section aloud and check that the prompt advances at the right place.
4.  Use a quiet enough setting for clear speech and keep the wording close to the script.

Recognition runs on the iPhone. Availability depends on the device, script language and locally installed speech support. Missing language models can prevent voice-follow from starting; test your setup before relying on it offline.

## If the text stops following

Check the app’s error message first. Denied permissions can be changed in iPhone Settings under Privacy & Security. If the language or local model is unavailable, use fixed-speed scrolling rather than depending on recognition for that take.

If speech recognition stops during capture, the app keeps the take and manual prompting controls available. Check the prompt and use pause/resume or fixed-speed control as needed; stop and retry if the delivery has become difficult.

There is no cloud-recognition fallback. Your recording still contains its microphone audio: “on-device” describes where processing happens, not a silent video. Read more about [local storage and sharing](https://teleprompterfocus.com/guides/private-on-device/).