How to land on-screen words exactly on the vocal that says them
Tell Vizard Agent which words land on which line and to time them to the vocal rather than the beat. It measures where each phrase actually starts by analysing the signal, subtracts the animation's own pop-in delay, and checks on real frames that the word is readable at the moment it is sung.
What do you type into Vizard Agent?
Say "on the vocal" explicitly. Timing to the beat grid is the sensible default for a montage and the wrong default here, and the phrase that overrides it is worth saying plainly rather than implying with "make it tight". Vizard Agent also needs to know which occurrences of a repeated word get text.
Prompt
What does Vizard Agent actually do?
It finds two separate causes of late text and fixes them independently. An animated word doesn't appear, it arrives over several frames, so a trigger timed to the vocal makes the word legible after the syllable has gone. And two choruses that sound identical rarely are. This was a music video whose on-screen text was landing consistently a fraction late. In the session behind this article it:
- Compared the current text times against the beat grid and measured the exact vocal onsets in every chorus and chant section, by signal analysis where the transcript was unclear.
- Tried word-level transcription on the chorus slices as a cross-check.
- Measured the pop-in delay of every text graphic, re-timed all the text hits and verified on frames that the text was up at each vocal hit.
- Checked when one word actually settled into readability, viewed the settle frame by frame, re-rendered its timing and verified it at all four of its appearances.
- Correlated the first chorus against the confirmed second one, listened closely to where the first really entered, re-cut it to the true vocal entry and checked the frames at each vocal moment.
When does this not work well?
Onset measurement needs a vocal that can actually be found in the signal. Where the voice sits deep inside a dense mix, or the phrasing is deliberately loose and behind the beat, the measurement gets less certain and the timing becomes a judgement call rather than a number. Long held or slurred entries have no single moment.
Common questions
Is the beat grid useless? No. Vizard Agent still builds one. It just isn't what the words are timed to.
Do I need stems? No, but they help. Vizard Agent measures from the mix if that is all there is.
Does the animation style change the timing? Yes, and Vizard Agent measures each style's own delay.