Vizard Agent

How to add fog, rain or atmosphere to a video

Last updated 2026-08-16 · 5 min read

Upload the footage to Vizard Agent and say what atmosphere you want and where it should sit. Vizard Agent studies the camera movement in your shot, finds real elements that match it, and composites them into the frame. Your original picture stays exactly as filmed underneath the effect.

What is the short version?

Adding weather to a shot is not the same as generating a shot with weather in it. The second one gives you a different video; the first one gives you your video, with fog in it. If you filmed the shot for a reason, you want the first, and you should say so.

  1. Go to Vizard Agent and upload the footage.
  2. Say what to add and exactly where it belongs in the frame.
  3. Say that the underlying video must not change.

What do you need before you start?

The shot as filmed, and a placement. Vizard Agent will study how your camera moves before it chooses anything, because an element that sits still while your camera pans reads as a sticker on the lens — but it cannot know which part of the frame the atmosphere belongs in unless you tell it in plain words.

What do you type into Vizard Agent?

Placement and restraint in one sentence. The second half of the prompt does as much work as the first: it tells Vizard Agent to treat your footage as fixed and the effect as a layer, which is the difference between compositing and regenerating.

Prompt

Add thick fog around [the hills]. Do not change the video itself. Make it look as natural as possible.

Variants worth knowing:

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real fog shot. The two steps that decide whether the result is convincing come before anything is added: it looks at the frames, then works out what the camera is doing, because the element has to move the way your shot moves.

  1. Inspects the video and extracts sample frames.
  2. Builds those frames into a montage and looks at it to see the composition and where the hills sit.
  3. Analyses the video's technical details.
  4. Downloads the source into the workspace to work on locally.
  5. Analyses the camera movement and composition — the step that decides how the fog has to behave.
  6. Reviews its own capabilities, checking what the stock library and the video editor can do before committing to an approach.
  7. Searches the stock library for fog, then searches again more specifically for fog and cloud flowing over hills.
  8. Downloads the chosen elements in parallel and composites them into the shot.

Step 7 is worth noticing. Vizard Agent does not settle for generic fog; it searches again for fog behaving the way your shot needs it to behave.

What does the result look like?

From the run this page is written from, probed on the delivered file: 1080x1920, H.264, 30fps, 16.3 seconds, audio preserved. The original audio was untouched, which is what you want — atmosphere is a picture operation, and a shot that suddenly gains ambient wind noise announces itself.

Ask for a longer or differently framed version and Vizard Agent re-composites from elements it has already chosen, rather than starting the search again.

How long it takes was not measured separately here. Comparable work runs a median of 28 to 38 minutes end to end, and the upper quartile is several times the median. Across all projects the median cost by tier is Flash 47, Pro 55, Max 242, Ultra 263 credits.

When does this not work well?

Compositing an element into a shot means matching light, movement and depth all at once. Vizard Agent handles ordinary camera moves well, and the failures cluster around shots where those three things are hard to read from the footage itself.

How do you fix a result that came back wrong?

Say where and how much. Vizard Agent keeps the downloaded elements, the camera analysis and the composite, so adjusting density or placement is a change to the existing layer rather than a fresh search — which is why "less fog on the left" is a quick instruction.

How does Vizard Agent compare to doing it yourself?

By hand this is tracking, masking and compositing in After Effects: track the camera, roto the foreground, find or generate an element, match its movement, grade it to sit in the shot. It is skilled work, and most of the hours go into the tracking and the roto rather than the creative choice.

By hand Vizard Agent
Camera match Track the shot manually Analyses the movement first
Finding elements Browse a stock library Searches, then re-searches specifically
Placement Roto and mask by hand Say where it goes
Adjusting density Re-render the comp One instruction

Common questions

Will it change my footage? Not if you say not to. Say "do not change the video itself" and Vizard Agent composites over it rather than regenerating it.

Can it add rain or snow instead? Yes. Same instruction shape: what, where, and how heavy.

Can it replace the sky? Yes. Vizard Agent masks the horizon and keeps the rest of the frame as filmed.

Can it add the sound too? Yes, but ask for it. Vizard Agent treats atmosphere as a picture operation by default, so wind, rain or room tone under the shot is a separate instruction rather than something that arrives with the fog.

Does the audio change? No. Atmosphere is added to the picture only, so any wind or rain sound has to be asked for separately.