How to read from a teleprompter without looking like you are reading
The short answer
Three things give a script away: eyes that travel, a flat reading voice, and a pace set by the scroll instead of by you. Put the text directly under the lens and sit at arm's length or further so your eye movement shrinks. Write short lines in spoken language, with a break wherever you would breathe. Set the scroll to your own speed, usually 130 to 160 words a minute, and rehearse once before you record.
Viewers forgive a lot in a video: soft focus, a plain wall, a cough. They do not forgive the feeling of being read to. The good news is that the feeling has three specific causes, and each has a fix that takes minutes.
1. Your eyes are travelling
When the text is below or beside the lens, you look at the text, and the viewer sees someone looking past them. When the lines are long, your eyes sweep left to right, and that scanning is what people notice.
- Put the text right under the lens. On a laptop the camera is at the top of the lid, so the text belongs at the top of the screen. Shrink the browser window and push it up under the webcam.
- Move back. Eye movement is an angle. At arm’s length or further, the same line needs far less movement than it does up close.
- Narrow the column. Short lines mean the eyes barely move sideways.
- Use a bigger font than you think. If you are squinting, you are leaning in, which undoes the distance.
2. It is written for the page
Written sentences are long and nested. Spoken ones are short. Before you record, read the script aloud once and change every place you stumble.
| Written | Spoken |
|---|---|
| “In order to obtain the best results, it is recommended that users calibrate first.” | “For the best results, calibrate first.” |
| “There are three factors which contribute to this.” | “Three things cause this.” |
| “We will not be able to” | “We can’t” |
Then break the lines where you breathe. In the Spolvero prompter every line break you type is kept as a break on screen, so you can lay the script out in phrases. A line you can take in at a glance is one you can say while looking at the lens.
Mark the words you want to stress in capitals, sparingly. A flat voice comes from reading word by word; a stressed word every line or two forces your voice to move.
3. The scroll is setting your pace
If the text moves faster than you speak, you chase it and sound rushed. If it is slower, you pause mid-sentence waiting for the next word. Both look scripted.
Most people present at 130 to 160 words a minute. Rehearse once with the text running and nudge the pace until you forget about it.
| Script length | At 130 wpm | At 150 wpm |
|---|---|---|
| 150 words | 1 min 9 s | 1 min |
| 300 words | 2 min 18 s | 2 min |
| 450 words | 3 min 28 s | 3 min |
| 750 words | 5 min 46 s | 5 min |
If you tend to stop and think, hold something up, or restart a line, a steady scroll will leave you behind. That is what voice follow is for: the text moves only to words it has heard, and waits when you pause.
Recording a take
The online teleprompter does the reading and the recording in one tab:
- Paste your script and press Show on the prompter. The word count under it shows how long the take will run at your pace. The script is saved on your device, so it is still there tomorrow.
- Choose how the text moves: steady speed in words per minute, or follow my voice. If voice follow loses its place, tap the line you are on and it picks up from there.
- Rehearse with S. It runs the text without recording.
- Turn on the camera and press Record (or Space). The text sits over the top of the preview, next to the lens, and starts from the top with the take. P pauses.
- Stop, watch it back, download.
The text is drawn over the preview and is not part of the recording. The preview is mirrored, because that is how people expect to see themselves, and the file is the camera’s real view, so writing on a shirt reads the right way round. Chrome and Safari give an MP4; Firefox gives a WebM.
The script and the video stay on your device: the take is recorded in the browser tab and saved to your downloads, with no upload. Voice follow is the one part to read about first. It uses your browser’s own speech recognition. Recent versions of Chrome can run that on the device. Where the browser cannot, it streams your speech to its maker’s service, and the page asks before it starts. Steady speed never listens at all.
If you only need audio, the same prompter panel is inside the voice recorder: press Read from a script above the timer.
Before the take that counts
- Look through the text at the lens, as if the words were printed on the glass.
- Blink and move. Frozen shoulders read as reading.
- Smile on the first line; it sets your voice for the rest.
- Do two takes. The second is nearly always the one.
For the picture itself, see how to light yourself for a video call, and for the sound, recording a voiceover on a laptop mic.
- Video
- Presenting
