When your script is longer than one take of the video model, the ad is made in parts. Nothing is refused for its length: each part is a clip of its own, and the parts are joined into one video, captioned as a whole when you pick a caption style. You see the parts before you press Generate.
How long your ad is
Length starts on Fit the words, which works the length out from your script:
- From your words. The lines in double quotes, or all your words if nothing is quoted, at an unhurried presenter's pace of about two and a half words a second, with a beat of air at the end.
- From your timings. If your script times itself outside the quoted lines, such as "0:00–0:05", "[3-10s]" or "10 to 16 seconds", the timings set the length, as long as the words can be said in that time. A time inside a quoted line, such as "it charges in 30 seconds", is something they say, not a timing.
When one timed stretch has more words than it can hold at a brisk pace, about three words a second, the bar says which stretch it is, for example "Your 0–3 s beat says 12 words, which take about 5 s". Those words may sound rushed or run into the next stretch, so cut some or give it more time. It's a heads-up: Generate stays on. The Creative Director's script cards say the same.
The count under the words shows the result, for example "your script runs 40 s · 3 parts".
If you pick a length instead, the ad is one take of exactly that length, however long the words are. To have a long script made in parts, leave Length on Fit the words.
Where it's cut
A script is made in parts when it's longer than the model's longest take, or longer than the model can read at once. It's cut where your words show:
- between your timed stretches, if the script has timings;
- otherwise between quoted lines;
- and a line too long for one part, between its sentences.
Each part says only its own lines and opens on the last frame of the part before, which is taken at no charge. Each part is as long as the shortest clip the model makes that holds it.
Before you press Generate, every cut is marked in your words with its part and its time, such as "Part 2 · 0:15–0:30 · 15 s", and the line under the bar says why: for example, that the model makes up to 15 seconds a take, so your script is made in 3 parts.
Limits
- 60 seconds, 8 parts. An ad joins up to 60 seconds of clips, in up to 8 parts. Words past that are left out, and the bar says where the ad ends and how many words are dropped.
- 20,480 characters. That's the most the words box keeps. A longer paste is cut, and the bar says how much was left out.
- What a model reads. If one stretch of your words is more than the model can read in one take, even on its own, the bar says its end will be cut. Shorten it, or pick a Seedance model.
Keeping one voice
With a voice picked under Voices, every part speaks in it. In the model's own voice, Seedance carries the first part's voice into the parts after it when there's at most one presenter and no real person. Otherwise each part may sound a little different, and the bar says so: pick a voice under Voices to keep it the same.
Lip-synced ads
On a model that lip-syncs a voice, the length is the voice's. If the voice read from your words could run past what the model moves to in one take, the words are read part by part in the same voice, each part is lip-synced on the same face, and the parts are joined. Every other part is framed a little closer, so each join reads as a jump cut.
A recording you upload goes in whole, so it has to be no longer than the model takes. A longer one stops Generate, and the bar names the limit.
What it costs
Each part is a clip, priced at its length, so a script in three parts is three clips. The line under the bar lists them, with each part's length and the clips' price together. The opening picture and the finish, which joins the parts and adds any captions, are charged once a take, for the whole joined video. A video to look like is followed by every part, and each part is charged for the seconds of it that it uses.
While it's made
In My ads, the tile says which part is being made, such as "Part 2 of 3", with a bar showing each part made, being made, or to come.
If a part fails, its credits come back automatically, and the captions, and any part that picks up from it, aren't made or charged. Parts already made were charged as they were made. See My video failed. Did I lose my credits?
Related
Still stuck?
A person reads every message.
Usually within 4 hours on a working day.
Have an account? Log in and write to us from Help in the app.