How to make a video from only your own footage with nothing added
Upload your footage and tell Vizard Agent that no generated shots, stock clips, images or B-roll may be added. Vizard Agent works entirely within what you filmed — improving the pacing, adding a voiceover and captions — and leaves every frame of picture as something you actually shot.
What is the short version?
Plenty of platforms and affiliate programmes require that the footage be genuinely yours, and plenty of viewers can spot a generated shot instantly. Ruling out everything except your own material is a legitimate constraint, and it has to be stated because it is not the default.
- Go to Vizard Agent and upload your original footage.
- Say that nothing generated, stock or borrowed may be added.
- Give the voiceover script word for word if you have one.
What do you need before you start?
Your footage and your script. Vizard Agent will not go looking for anything to fill gaps once the constraint is stated, so what you upload is the entire visual budget for the video and it needs to cover the whole runtime.
- The footage. Everything you filmed, however rough.
- The rule. Say it explicitly: nothing added.
- The script. Word for word, in your language.
- The language. It decides the voice and the typeface.
- The length. Short, if the footage is short.
What do you type into Vizard Agent?
State the prohibition before the request. Vizard Agent otherwise reaches for stock or generated shots wherever your footage runs thin, which is usually helpful and is exactly what you are ruling out here, so the constraint has to come first.
Prompt
Variants worth knowing:
- Several voices compared. Generated as options, chosen by listening.
- Captions in your language. With the correct typeface installed.
- Pacing only. Tightened, with nothing else touched.
What does Vizard Agent actually do?
Here is the order Vizard Agent worked in on a real video built strictly from the uploader's own footage. Two habits stand out: it generated alternative voiceovers and judged between them, and it went and downloaded the typefaces the language actually needed.
- Checks the uploaded video's details and inspects the frames to see what exists.
- Checks the audio levels of the original and analyses the footage's content.
- Reviews the voiceover, music and caption options available to it.
- Searches for natural conversational voices in the right language.
- Generates the voiceover, then generates alternatives for comparison.
- Evaluates the options for tone, naturalness and clarity before picking one.
- Trims the silence from the chosen voiceover and pulls word-level timings.
- Checks the available typefaces and downloads the ones the language needs.
- Searches and evaluates background music, verifies the video's true dimensions after rotation, and cuts to the word timings.
Step 6 is the one worth borrowing. A generated voice that reads fine on the first listen often sounds artificial next to two alternatives, and generating three costs almost nothing compared with delivering the wrong one.
What does the result look like?
From the run this page is written from, probed on the delivered file: 1080x1920, H.264, 30fps, 8.23 seconds, AAC audio. Vertical, eight seconds, every frame from the uploader's own camera, with a conversational voiceover in their language, captions in the correct typeface and music underneath.
Eight seconds is short because the footage was short. That is the honest consequence of the constraint: with nothing added, the runtime is set by what you filmed rather than by what you wanted.
When does this not work well?
Ruling out everything except your own footage puts real limits on what any editor can do, human or otherwise. Vizard Agent will hit those limits and say so rather than quietly working around them by adding the one shot that would fix the problem.
- Short footage means a short video. There is nothing to pad it with.
- A missing shot stays missing. No product close-up, no product close-up.
- Poor footage cannot be replaced. Grading and pacing help; they do not substitute.
- Repetition becomes visible. Reusing your three clips shows quickly.
- Voice and music are still generated. Say so if those are prohibited too.
How do you fix a result that came back wrong?
Say which part is not working. Vizard Agent keeps every voiceover candidate, the word timings, the caption settings and the pacing decisions, so the voice can be swapped or the cut re-timed without ever breaking the rule about the picture.
- "The voice sounds artificial." Swapped for one of the alternatives already generated.
- "It is too fast." Re-paced against the same footage and timings.
- "The captions render as boxes." The correct typeface is installed and re-rendered.
How does Vizard Agent compare to doing it yourself?
By hand this constraint is easy to state and tedious to honour, because the moment a section runs short the obvious fix is to drop in something you did not film. Doing it properly means re-cutting instead, which is slower and better.
| By hand | Vizard Agent | |
|---|---|---|
| The rule | Easy to break under pressure | Held as a hard constraint |
| The voice | Record it, or accept the first take | Several generated, judged, chosen |
| Typefaces | Whatever is installed | Downloaded for the language |
| Pacing | Trim by feel | Cut to word-level timings |
Common questions
Does this really mean nothing is added? Nothing visual. Vizard Agent uses only frames you filmed once the constraint is stated.
Can it still add a voiceover? Yes. Vizard Agent generates the voice and the music; the rule here covers the picture.
What if my footage is too short? Then the video is short. Vizard Agent will not pad it with anything you did not shoot.
Can Vizard Agent improve the pacing? Yes. Tightening, reordering and trimming all stay inside the rule.
Will it compare voices for me? Yes. Vizard Agent generates alternatives and evaluates them for tone and naturalness.
Why state the rule instead of just not asking for stock? Because filling a thin section with stock or a generated shot is the normal, helpful thing to do, and Vizard Agent will do it unless told otherwise. Saying it explicitly is what makes the constraint hold.
Does it work in any language? Yes. Vizard Agent finds a suitable voice and installs the typeface the captions need.
Why check the dimensions after rotation? Because phone footage carries rotation metadata, and the real frame is not always what the file reports.
Can I ban generated audio too? Yes. Say so and Vizard Agent works with the original sound instead.
Does this satisfy affiliate programme rules? Read the programme's own terms, but the constraint helps. Vizard Agent leaves every frame as something you filmed, which is usually what a rule about original footage is asking for.
Can I mix in footage from an older video of mine? Yes, and it still counts as your own. Vizard Agent treats anything you filmed as fair material; the rule is about generated and licensed content, not about which shoot it came from.
Does Vizard Agent check the finished video? Yes. It reviews the cut and confirms nothing outside your footage appears in it.