A teleprompter is a screen that shows your script to you while you look straight down the lens, so you can read without appearing to read. On a full rig the words are reflected off a piece of angled glass mounted in front of the camera. In a browser or on a tablet, the same job is done by the screen itself.
That is the entire idea. Everything else is a detail of how the words get in front of your eyes without the camera seeing them.
How the glass works
The part that confuses people is why the camera cannot see the text. The answer is a beam splitter: a piece of glass coated so that it reflects light from one side while letting light pass through from the other.
The screen lies flat, facing up, underneath the glass. The glass sits at forty-five degrees in front of the lens. Your script bounces off the underside of the glass towards you, and the camera looks straight through the same piece of glass at your face.
Because the reflection reverses the text, the screen has to show it mirrored. Every teleprompter has a mirror setting for exactly this reason, and forgetting to turn it on is the most common first mistake on a rig.
The two kinds you will meet
A beam-splitter rig is the one described above. The camera and the glass are mounted together, usually with a hood to keep stray light off, and it is what you see on a set or in a studio. It puts the words exactly on the lens axis, which is why the eyeline is perfect.
A screen on its own is a tablet, laptop or phone running prompter software, placed as close to the lens as it can get. There is no glass, so you are not reading through anything, and your eyeline is a few degrees off the lens rather than dead on it. For most filming that difference is invisible, and it is how the majority of pieces to camera are now shot.
The second kind is what most people mean when they search for a teleprompter today. It costs nothing, it travels in a bag you already carry, and it is good enough for everything short of a presenter standing at a lectern.
Why the eyeline is the whole point
A viewer can tell when someone is reading. Not because the words sound wrong, but because the eyes move: a small sweep left to right, and a flick back at the end of each line.
A prompter fixes this by making the reading distance short and the text large, so the sweep is small, and by putting the words where the lens already is. The closer the script sits to the lens axis, the more it looks like eye contact.
This is also why prompter text is set in short lines rather than full-width paragraphs. Fewer words per line means less eye travel, which means less of the tell.
A confidence monitor is not the same thing
The two get mixed up. A confidence monitor is a screen placed for a speaker rather than for a camera: on the floor at the front of a stage, or on a wall at the back of a room, so a presenter can see their slides or their notes while addressing an audience.
The difference is who it is for. A teleprompter hides the reading from the camera. A confidence monitor hides nothing, because the audience is in the room and can see the speaker glance across at it.
How much script actually fits
Most people read to camera at somewhere between 130 and 160 words a minute, slower than conversation because delivery to a lens tends to be more deliberate. A two minute piece is therefore roughly 280 to 320 words.
A script that runs long is the most common reason a piece to camera has to be shot again, and it is much easier to cut on the page than to speed up on the day.
Does it make you look wooden?
It can, and it is worth being honest about why. A prompter removes the pauses. When you speak without one you stop, think, and start again, and those small hesitations are what make speech sound like a person rather than a document. A scrolling script never hesitates, so it is easy to end up reading at a flat, even pace that no one uses in real life.
The fix is in the writing rather than the kit. Scripts that work to camera are written the way you talk: short sentences, contractions left in, one idea per line. Read your script aloud before it goes anywhere near a prompter, and change every phrase you would not actually say out loud.
The other half is rehearsal. Read it through two or three times before recording, so you are recognising the next line rather than discovering it. That is the difference between a presenter who is reading and a presenter who already knows what is coming.
What goes wrong on the day
The same handful of things, on almost every shoot.
The mirror is off. On a beam-splitter rig the text has to be reversed, and the moment to find that out is not while the client is watching.
The text is too small. People set the size while standing at the screen, then walk back to their mark and cannot read it. Set the size from where you will actually stand.
The script is too long. Three hundred words is about two minutes. A page and a half of A4 is not a two minute video, and finding that out during the take costs the morning.
The scroll speed is fought rather than followed. With a fixed scroll you end up chasing the words, speeding up and slowing down to stay with them, and it is audible. This is the reason prompters that follow the voice exist: the script waits for the speaker instead of the speaker chasing the script.
Nobody checked the eyeline. If the screen sits well below the lens, you get a presenter who appears to be reading something on the floor. Get the reading line as close to lens height as the setup allows.
Where the word comes from
The first one was built in the early 1950s for American television, using a paper roll in a suitcase turned by a motor, with an operator winding it to match the presenter. The brand name was TelePrompTer, and like Hoover and Biro it became the word for the thing itself.
The mechanism changed completely and the name stayed. What has not changed is the job: put the words where the eyes already are.
Trying one
You do not need a rig to find out whether prompting suits you. Open the free teleprompter on a laptop or tablet, paste a script, and read it to your phone camera. It is free, you can use it without an account, and the whole question of whether you can read without looking like you are reading is answered in about five minutes.
How Miracue works explains what happens to a script once you paste it in.
Written 13 August 2026 by Lumira Studio, a video production company in Hertfordshire, United Kingdom. We built Miracue for our own filming days.