Voiceover gives you the words with the timing worked out, for a person to read — segment by segment, against a duration it was given rather than one it guessed.
Nothing opens in this browser — this one takes a job and does the work on our own graphics cards in Oregon. The free allowance below is how you try it: 60 segments, no card and no payment, on the key you are issued.
There is no speech model on our machines, so it needs a timed caption file or a duration, and it says so before anything else.
The problem
Written copy and spoken copy are different lengths and nobody finds out until the recording session. You book the time, get three takes in, and discover the second section cannot be said in the twenty seconds the edit leaves for it — so it gets rushed, or the edit gets rebuilt, and either way the session costs double what it should have.
The script is worked out segment by segment against the duration each one has. Not a word count, not an average reading speed over the whole thing — the bit that has to fit in eleven seconds is checked against eleven seconds.
The output is words for somebody to read aloud, with the timing marked. No synthetic voice is produced, because there is no speech model here and no plan for one.
It needs a timed caption file or a stated duration. With neither it refuses and says so before anything else, rather than estimating from word counts.
Voiceover cannot listen — there is no speech model on Tower 1, by design, because a speech model would mean your audio on our disks. So the duration has to come from you: a timed caption file, or a stated length per segment. Without one it refuses before anything runs, in those words. The alternative is estimating from words per minute, which is right on average and wrong on every individual segment — and the segments are the whole point.
How it works
A timed caption file, or a duration per segment typed in. Either is fine; neither is optional.
Segment by segment, each written to be sayable in the time it has. Marked with its duration so whoever reads it can see what they are working to.
A person, a microphone, your own room. Voiceover produced a script, not a sound file, and there is nothing of yours on our machines when the job closes.
What it will never do
The limits here are structural rather than unfinished. Each one is the reason a different promise can be kept.
Pricing
A segment is a piece of script with its own duration — usually a line, a beat or a cue. Sixty are free on every key, which is a short script in full.
A short script, timed and written in full.
A regular series or a year of scripts.
A studio scripting continuously.
Per segment written. A refused job — no timing supplied — is never charged, so finding out you need a timed file costs nothing. Your licence key is emailed to you within two working days of payment.
Checkout is our own bitcoin server, not a card processor. The units you buy are added to your key: buy from the Buy more box inside the program and they arrive on their own once the payment confirms, or buy here and paste the invoice number from your receipt into that box.
Questions
Because it would mean a speech model on our machines, and the moment that exists the promise that your audio never comes near us is gone.
There is a second reason and it is not a technical one: a tool that generates a voice is a tool that generates somebody's voice, and we are not going to be the vendor that made that easy in this industry.
Your edit. Any editor will tell you how long a gap is, and most will export a timed caption file in a couple of minutes.
If you have captions already, Dialogue will clean them up and the file it returns is exactly what this wants.
It comes back saying so, with how much over it is.
That is the useful answer. The alternative — words crammed into a slot nobody can read them in — is what the recording session finds out the expensive way.
No. It runs on Tower 1's own graphics cards in Oregon.
A key. One line of text, issued to you and shown exactly once — only its hash is kept here, so it cannot be looked up later and we could not tell you what yours is. Every key arrives with a free allowance already on it, which is the tier at the top of the pricing above, and the work starts the first time you send a job with that key on it.
Each job is metered in the unit named in the pricing, and a unit that fails — a file that will not read, a row the model held back — is not charged and is named in the job report rather than silently skipped. When an allowance runs out the next job is refused before any work starts, with the number it used printed on the refusal.
More units are bought in bitcoin. The checkout is a bitcoin server we run ourselves. Buy from the Buy more box inside the program and the units are added to your key on their own as soon as the payment confirms; buy from the plan buttons above and paste the invoice number from your receipt into that same box. Each payment adds its units once, to one key.
Give it the duration your edit actually left and see whether the script you used fits. The ones that do not are the takes that got rushed.