For YouTube
A teleprompter for YouTube Shorts, and for the forty-minute take.
The same app handles a script written to fit sixty seconds and a podcast that runs all afternoon — voice following when you want to sound like you are talking, a fixed pace when the take has to land on a length.
Two kinds of YouTube script, two different modes
A Short is written to a length. Constant mode runs at a fixed speed set in words per minute — the one setting that tells you how long a take will run before you shoot it. Divide the word count by the pace, trim until that number is the runtime you want, and the last word lands on cue.
Anything longer is the opposite problem: over twenty minutes a fixed pace is what makes people sound like they are reading. Voice mode recognises the words you actually said and scrolls to them, so you can stop mid-sentence, take a second run at a name, or slow right down over a figure. Sound mode advances on the presence of speech rather than the words, for conditions where recognition is unreliable. Recognition runs on your iPhone where your language supports it; where it does not, iOS sends the audio to Apple to convert it to text, and the app tells you before and during the take. The voice teleprompter page takes all three apart.
The start delay — what people mean by a countdown
There is no 3-2-1 countdown in Voice or Sound mode, and nothing is missing: both wait for you to speak, so a countdown would be counting down to nothing.
Constant is the mode with a clock in it, because it starts moving whether or not you do. Before the take you pick a start delay — off, or 5, 10 or 20 seconds, labelled “time to get into position”: the gap where you tap record, walk to your mark and get your hands down.
It is drawn inside the floating window as well as full screen — so with the phone on a tripod and the camera app in front, you can still watch the seconds run out.
Mirror and rotation, for a beam-splitter rig
On a beam-splitter — the glass hood that reflects the script into the lens’s line of sight — the text has to be flipped horizontally. Mirror does that.
Rotation is the other half of a rig. The floating window turns −90°, standard, or 90°, and “Rotate 90°” moves the top edge to the right, so the script reads the right way up whichever way round the phone is clamped in the hood. The window is Wide (3:2) or Tall (2:3), with text scale and side margins set separately from the full-screen prompter — glass is not the reading distance your hand is.
Plan one thing around it: the floating window does not accept touch. That is an iOS constraint, and it is why the spoken commands exist — pause, continue and jump between sentences by saying so, in ten languages, with editable phrases. The volume buttons are marked Experimental and do not work inside camera apps, which take them for the shutter. So on a rig it is voice: tracking keeps following while you record in Camera, and if the camera app does take the microphone, Wispim switches to Constant, tells you, and switches back when the microphone returns. The floating window page has both halves of that.
Typography for a script that runs twenty minutes
Text size, side margins, line spacing, font weight and the position of the reading line all adjust. On a Short the defaults are fine — a minute is over before the type can tire you. The forty-minute take is the one that pays for setting them, and the page for teachers works through what each one buys over a long read. The part that matters on a shoot: pinch to change text size mid-take, so a phone clamped further from your eyes than you planned costs a gesture, not a re-shoot.
Video podcasts, where the script is a structure
Two people and a long take rarely follow a script word for word. What you bring is a running order: the questions, and the figure you must not misquote. A fixed pace is the wrong instrument for that, because the order is somewhere to come back to rather than something to keep up with. Voice mode lets the script sit still through the tangent your guest goes on, and when the conversation lands somewhere the order did not, you browse ahead to the question you now want and start reading there — it finds you wherever you pick the script up. The recovery controls behind that, browse mode and tapping a word among them, are set out on the voice teleprompter page.
Getting a long script onto the phone
Long scripts are not written on a phone. Import brings in PDF, Word and RTF, part of Pro; the free tier is 6 prompter takes with every feature and no limit on scripts. Pro removes the take limit and adds the import, as a monthly or yearly subscription or a one-off purchase.
Questions
Is there a countdown before the script starts?
Constant mode has a start delay — off, 5, 10 or 20 seconds — chosen before the take and shown inside the floating window as well as full screen, so you can see it from a tripod. Voice and Sound have no countdown: they wait for you to speak, so there is nothing to count down to.
Does it work with a beam-splitter teleprompter rig?
Yes. Mirror flips the script horizontally for the glass, and the floating window rotates −90° or 90° so it reads the right way up whichever way the phone is clamped under the lens. Rotate 90° moves the top edge to the right.
Which mode suits a long video podcast?
Constant if you are reading word for word at a planned pace, Voice if the script is a running order you talk around. In Voice mode you can browse ahead to the next question while the conversation goes elsewhere, then start reading anywhere and the prompter finds you.
Can I control the prompter while recording in the Camera app?
By voice. Say pause, continue, or move between sentences — the spoken commands work in ten languages and the phrases are editable. The volume buttons are marked Experimental in the app and they do not work inside camera apps, which take the buttons for the shutter.