Decks and narration
On this page
A deck is a set of slides written from your material, with a spoken script for each one. You can watch it like a lecture or read it like notes — the transcript is beside the slides, and clicking a word jumps the audio there.
How one gets made
Your material is read, the deck is written, then it is rendered and narrated. Rendering happens before narration on purpose: a deck that fails to render should not have cost you a voice reading it first.
Generating a deck again makes a new one rather than replacing what you had. The lecture keeps both, so a version you liked does not disappear because you were curious.
The viewer
- Slides and transcript are two tabs on the same rail. The transcript is grouped by slide and searchable. Once the narration has word timings, the word being spoken is marked in the transcript too, and the transcript keeps it in view as the lecture plays. Scroll the transcript yourself and it waits for you: it follows again a few seconds after you stop, or at once with Back to now.
- Progress is shown per slide, not across the whole lecture. A lecture's length is the sum of its audio; a single bar across an hour tells you nothing useful.
- Captions run in a thin bar of their own directly under the slide — never over it, so they never hide anything on the slide. One line at a time, on a phone too: white text with a dark outline, as subtitles are on TV, on the slide's own background, and the word being spoken turns gold once the narration has word timings. They stay while you pause. The bar takes a line's room from the slide; to give it back, turn the captions off with CC: the viewer remembers that in this browser. (The S key turns them off for this visit only.)
- When you can open the deck's course, the bar above the deck's title names it, and the lecture's number where there is room; tap the course name to open the course page. In your own course the lecture takes you to its place there.
- Speed is set with the − and + buttons in the player: each press changes it by 0.05, anywhere
from half speed to double, and holding one down keeps changing it until you let go. Tap the
number between them to go back to normal speed. The speed you
set with the buttons is remembered in this browser, so the next lecture starts at it. On a
keyboard,
<and>change it too, for the lecture you are watching only. On a phone held upright, or a small phone turned on its side, the buttons are in the settings: tap the speed shown in the player to open them right there. - Points appear as the narration reaches them, so the slide fills in as it is explained rather than showing you the ending first.
- Parts of a slide: the marks on the progress bar divide a slide's narration into parts, one for each point it makes. The two part buttons between the slide arrows and play jump to the previous or next part, so a point that went past too quickly is one press away. If you are already a moment into a part, the previous-part button starts it again; at the very start of a slide it goes back to the last part of the slide before. On a keyboard, Shift+← and Shift+→ do the same. On a phone this button has taken the place of the old 10-second replay.
- When a slide's narration ends, there is a short pause before the next slide, so the last point has a moment to sink in. Press pause during it and you stay on that slide; press play again and the lecture carries on with the next one.
- Up next, under the rail, is the lecture's quiz — it lights up as you reach the end. When the quiz is still being prepared (right after you finished the previous lecture, Autopilot usually has not made it yet), it says Quiz being prepared instead — the same thing the lecture's row on the course page says — with a way back to the course. Once the quiz is ready it takes that place by itself, while the deck keeps playing; there is no need to reload.
- A recap (the presentation made from a quiz's misses) leads back to that quiz's results page — where the gap quiz lives — and, once the gap quiz is ready, Up next offers it under the retake.
Voices
A deck is narrated in one voice. You pick it when you generate, and the choice is fixed for that deck — a deck narrated halfway in one voice and halfway in another is two people talking.
You can add a second voice to a finished deck afterwards. That gives the same deck a second narration track and a picker in the player; it does not replace the first. Each voice speaks one language, so only voices for the deck's own language are offered.
The player remembers the voice you pick, for each language: the next deck in that language opens in the same voice when it has one, instead of going back to its default. This browser remembers it. Signed in, the player inside the app also keeps it on your account, so it follows you to your other devices. The public pages — the sample course you can watch without an account — remember it only in this browser, even when you are signed in. A link that opens a deck in a particular voice still plays that voice. The decks Autopilot or the Prepare button make for you next are narrated in that voice too, unless your course's settings name a voice of their own.
Now and then a voice is withdrawn. Decks already narrated in it keep it: it still plays, and you can still pick it in their player. It is no longer offered for anything new, though — new decks, an added voice — and if your course's settings named it, your next decks get the voice you last picked in the player, else the language's default; the settings say so.
If a deck looks wrong
Regenerate it. If the second attempt has the same problem, the material is usually the reason — a transcript with the important part inaudible cannot produce slides that explain it.
All help pages (14)
- Help
- What this does
- Starting a course
- Your material
- The browser extension
- Decks and narration
- Quizzes and recaps
- Results, exams and progress
- Autopilot and the course page
- Sharing and visibility
- Your account and settings
- Ulu
- Using LetUlu on your phone
- Waiting, and when something breaks
- Contact and support
Still stuck?
Write to us — we reply in a conversation you can come back to.