Chapter 21 · Formats
Two-Lane Video Scripts
A video script has to say two things at the same moment: what the viewer sees, and what they hear while they are seeing it. Storskarvia writes that as two lanes, either as a running list of voiceover and visual blocks or as the broadcast split page with VIDEO down the left and AUDIO down the right in locked rows. Either way, a clock runs against the words.
Starting a video script
In the Studio, choose Video script in the Type list of the New project form. Chapter 2 covers getting a document open.
The editor opens with a Timing tab beside Home, Review and View, holding the reading rate, the row duration control, the teleprompter and the total running time. This format's two view switches sit on View, covered in Chapter 12.
The three blocks
Before the grid is on, a video script is a stack of blocks of three kinds, chosen from the dropdown at the left of Home.
| Block | Key | How it sets |
|---|---|---|
| Section | ⌘⌥1 | Monospaced capitals with a rule under them, and what the outline navigator lists. |
| Voiceover | ⌘⌥2 | The spoken line, on white with a dark rule down its left edge. |
| Visual / B-roll | ⌘⌥3 | A blue tinted band labelled VISUAL. You do not type that label. |
Tab toggles between voiceover and visual, as the hint under the dropdown says.
Turning on the two-column script
Open the View tab and tick Two-column script. Storskarvia converts what is already there:
- A visual block followed by a voiceover block becomes one row, picture left, words right.
- A voiceover with no visual before it becomes a row with an empty VIDEO cell.
- An empty voiceover block is dropped, along with the blank heading a new document opens on. A blank line reads as nothing in a list and as a tall empty row in a grid.
- With nothing to convert, one empty row is laid down to type into.
The setting is stored with the document rather than your browser, so the script opens in the same shape tomorrow. On a shared document the toggle is disabled and reads "The document owner sets this."
Turning it off takes the rows apart, and each row becomes a Visual block and a Voiceover block again. Empty cells are dropped. A duration you set by hand is kept, parked on whichever half has the words in it, and it comes back on the row when you switch the grid on again. Nothing in the single-column view shows it, so a script can look untimed while the grid is off and still be carrying every number you typed.
Working in the grid
A row is exactly one VIDEO cell and one AUDIO cell, and that pairing is the point of the format. Rows are locked, so every row operation is its own command. The buttons appear on Home as soon as two-column mode is on.
| To do this | Use | If it refuses |
|---|---|---|
| Move across the row, then on to the next | Tab, ⇧Tab back | Tab past the last row makes a new one. It never leaves the grid. |
| New row, in the column you are in | Enter | |
| Cut the row where the picture changes | ⌘Enter, or Visual starts here | "Put the cursor in the AUDIO column first". It needs the caret partway along the spoken line. |
| Lift words out into a full-width section | Make section | "Select the words that should become the section". With nothing selected it promotes the whole cell. |
| Turn a section heading back into a row | Make row | "Put the cursor in a section heading first". The text lands in VIDEO, where a misread cue such as RUN TITLE CARD belongs. |
| Move a row up or down | ⌘⇧↑ / ⌘⇧↓ | "Already the first row of this section". A row will not hop a section heading. |
| Merge with the row above or below | ⌘⌫ / ⇧⌘⌫, or ⌫ at the very start of a row | "Nothing above to merge into, or a section heading is in the way". |
| Duplicate or delete a row | The last two row buttons | "Put the cursor in a row first". |
A merge joins video to video and audio to audio, so the picture notes stay beside the words that cue them, and the caret lands on the seam in the AUDIO cell, the one spot worth reading back. Every key is in Chapter 10.
Pasting depends on where the caret is. Inside a cell, what you paste arrives as text in that cell. Outside one, each paragraph becomes a row with the words in AUDIO, except short all-caps lines opening like a production cue (RUN, ROLL, CUT, SUPER, LOWER THIRD), which go to VIDEO. Any other short all-caps line is read as a section heading, and that guess is sometimes wrong, which is why Make row and Make section exist.
The running clock
On the Timing tab, Read at sets the reading rate: 120, 135, 150, 165 or 180 words per minute. It starts at 150, roughly where broadcast voice-over sits.
Every estimate comes from the AUDIO column. Shot notes add no time, since what is on screen is on screen for as long as the words beside it last, and text struck out as a tracked change does not count either. An estimated row is never shorter than one second.
Total running time sits at the right of the Timing tab and moves as you type. Tick Timecode on the View tab to put the clock down the left of the page too: each row shows where it starts, with its own length under that in smaller type. The status bar carries the same figure in minutes, from the same model. Chapter 13 covers the rest of that bar.
When the estimate is wrong
A title card, a slow drone shot or a deliberate pause has a length the words cannot tell you.
Put the cursor in the row and press Row duration, or click the clock in the row's
left gutter. The dialog takes 12, 0:12, 1:05 or
1:02:03, and an empty answer hands the row back to the estimate.
Anything else is refused with the row untouched, and the status bar says so. Away from a row, Row duration answers "Put the cursor in a row first".
A row you timed yourself draws its clock in blue, so a number somebody decided never looks like a guessed one. Zero is allowed: a lower third running under the row before it adds nothing to the running time. Splitting a row keeps the time on the first half, the one you were looking at, and the tail goes back to the estimate. Merging adds the two times together.
The teleprompter
Teleprompter on the Timing tab opens a reading surface built from the AUDIO column only. Shot notes are instructions for somebody else and have no business in front of a presenter.
It scrolls to the script's own clock, so a row timed at eight seconds takes eight seconds to pass the eyeline. Space plays and pauses, ↑ and ↓ change the speed, ← and → jump five seconds, the wheel scrubs, and Esc closes it. Mirror flips the text for a beam-splitter rig. A row with nothing to read holds its place as a gap of the right length, so a presenter is warned of a pause instead of finding the next line has moved.
What comes out, and in which file
| Export | What it carries |
|---|---|
| Word (.docx) | A real two-column table, VIDEO in capitals on the left, header row repeating. With Timecode on it gains a third TIME column carrying the same numbers as the screen, and the total under the title. |
| Plain text (.txt) | Each row as two lines: the shot note prefixed
[VISUAL], then the voiceover. |
| Rich text (.rtf), OpenDocument (.odt), Web page (.html) | Two paragraphs per row: the shot note in capitals, then the spoken line indented under it. Linear formats, so there are no column rules. |
| Typeset PDF (.pdf) | The prose writer, so a split page comes out as paragraphs. Print / PDF on the View tab prints the page as you see it, grid, timecode and all. |
Notes stay out of all of them. The neighbouring format is the audio drama, written for the ear alone: Chapter 20. Getting finished work out to other applications is Part VIII.
Updated 2026-08-28 · Full contents · Something wrong or missing?