Glossary
Masking: showing, hiding, and blending parts of the frame
Masking selects part of the frame so your edits apply only there — blur one face, darken one sky, reveal text behind one person. It is how edits look surgical instead of smeared across everything. This guide explains what video masking is and how beginners can make clean, convincing masks.
From the Mellius product team · Updated September 28, 2026
Conspiracy theorists, you may hate this
AI Director
›
Just type what you need
- ›Remove all silent parts
- ›Add bold kinetic captions
- ›Only have Bill Ackman speaking
- ›Cut to highlights only
- ›Add cinematic transitions
- ›Match cuts to the beat
The short answer
Video masking means selecting part of the frame so an effect applies only there — blurring a face, darkening a sky, or revealing text behind a subject. A mask is just a shape, often feathered and animated to follow movement. Masks turn global effects into local ones, which is how edits look precise instead of painted over everything.
The shape follows the subject
A convincing mask tracks its target through movement — static shapes on moving subjects break the illusion within seconds.
Edges blend believably
Feathering and edge refinement make the masked area melt into surrounding pixels instead of showing a cardboard cutout line.
The effect stays motivated
Every mask serves a visible purpose — privacy, focus, or style — rather than decorating footage that needed nothing.
What masking is, and the shapes it uses
At its simplest, a mask is a drawn shape — rectangle, ellipse, or freeform path — that defines where an effect happens. Everything inside the shape gets the treatment; everything outside stays untouched; feathering softens the boundary between them. Stack a blur inside an ellipse over a face and you have the most common mask in online video.
Advanced masks get intelligent: garbage masks roughly isolate regions for compositing, difference and luma masks select by brightness, and AI-powered tools select people, skies, or objects with a click and track them automatically. Underneath every variant sits the same idea — a grayscale map where white means apply fully, black means not at all, and gray means partially.
Everyday uses worth learning first
Privacy and cleanup come first: blurring faces, license plates, and screens, or hiding a mic, logo, or boom shadow. Then selective correction: darkening a blown-out sky without touching the landscape, brightening a face without washing the scene, or saturating one object while the world stays muted. These targeted fixes solve problems global adjustments cannot touch.
Creative masking opens the next level: text that slides behind a subject, transitions that wipe along an object's edge, clones and split-screens built from frozen halves, and focus pulls simulated with masked blur. Each effect is the same recipe — shape, feather, track, apply — dressed in different intentions.
Why beginner masks look fake
Hard edges are the prime suspect: an unfeathered mask draws a visible boundary no real lens would produce. Feather generously, zoom to 100 percent to inspect edges, and remember that moving subjects need wider feathering than still ones. A mask that looks perfect paused can still flicker in motion, so always judge edges during playback.
The second giveaway is drift — a mask drawn once and left behind while the subject walks away from it. Animate the mask path with keyframes on simple moves, or use planar and object tracking for complex ones. In Mellius, you can ask to blur a face through a scene in plain language, then check the result on the editable timeline frame by frame.
A practical sequence
Put it into practice
- 01
Draw the shape around your target
Pick an ellipse, rectangle, or pen mask and outline the face, object, or region. Keep the shape slightly generous — feathering will shrink its effective core.
- 02
Feather, then track the movement
Soften the edge until it melts into surrounding pixels, then keyframe or track the shape to follow the subject. Check the track at full speed, not just frame by frame.
- 03
Apply the effect and inspect in motion
Add the blur, correction, or graphic inside the mask and watch the whole shot through. Fix any frame where the edge shows or the shape lags behind.
Frequently asked questions
What is the difference between masking and green screen?
Masking selects regions of an existing shot with drawn or tracked shapes, while green screen (chroma key) removes a colored backdrop to composite a new background. Masking works on any footage with no special setup; green screen needs a lit colored backdrop during filming. They often combine — key the screen, then mask away what the key missed.
Can beginners do motion tracking for masks?
Yes — most editors include point or planar tracking that follows a face or object once you draw the initial shape. For simple predictable movement, hand-keyframing the mask every few frames works fine and teaches the fundamentals. Reserve manual tracking for shots where auto-trackers lose their target.
Why does my blur look like a floating box?
That floating-box look comes from a rectangular mask with too little feathering that is not tracking tightly. Switch to an ellipse that hugs the subject, raise feathering until edges dissolve, and make the track stick. Matching the blur's shape to the subject matters as much as the blur strength.
Continue reading
Primary sources
Product capabilities change. These official references support the comparisons above.
Describe the edit. Keep the timeline.
Mellius turns conversational direction into cuts, captions, graphics, motion, color, and sound you can still inspect and change.
Start editing free