Skip to main content
Blur effects apply Gaussian blur to video clips. MovieLite provides three blur effect classes: Blur, BlurIn, and BlurOut.

​
Blur

Apply Gaussian blur to the clip. Can be static (constant blur) or animated (blur increases over time).

​
Constructor

Blur(intensity: float = 5.0, animated: bool = False, duration: float = None)
​
intensity
float
default:"5.0"
Blur intensity (kernel size). Higher values = more blur. Must be odd number ≥ 1. The effect automatically converts to odd numbers and ensures minimum value of 1.
​
animated
bool
default:"False"
If True, blur increases from 0 to intensity over the specified duration. If False, applies constant blur at the specified intensity.
​
duration
float
default:"None"
Duration of the blur animation in seconds. Required when animated=True.

​
Examples

​
Static Blur

from movielite import VideoClip, vfx

clip = VideoClip("input.mp4")
clip.add_effect(vfx.Blur(intensity=7.0))

​
Animated Blur

clip = VideoClip("input.mp4")
clip.add_effect(vfx.Blur(intensity=15.0, animated=True, duration=3.0))
When animated=True, you must specify the duration parameter, otherwise a ValueError will be raised.

​
How It Works

  • Static blur: Applies OpenCV’s GaussianBlur with constant kernel size
  • Animated blur: Linearly interpolates kernel size from 1 to intensity over duration
  • Kernel sizes are always odd numbers (required by OpenCV)
  • Kernel size of 1 means no blur is applied
Source: src/movielite/vfx/blur.py:6-63

​
BlurIn

Blur-in effect that starts blurred and gradually becomes sharp.

​
Constructor

BlurIn(duration: float, max_intensity: float = 15.0)
​
duration
float
required
Duration of the blur-in effect in seconds, from the start of the clip.
​
max_intensity
float
default:"15.0"
Maximum blur intensity at the start of the effect. The blur decreases to 0 by the end.

​
Example

from movielite import VideoClip, vfx

# Start blurred, become clear over 3 seconds
clip = VideoClip("input.mp4")
clip.add_effect(vfx.BlurIn(duration=3.0, max_intensity=15.0))

​
How It Works

The blur intensity decreases over time:
  • t < duration: kernel_size = max_intensity × (1 - t/duration)
  • t ≥ duration: No blur applied (sharp)
Source: src/movielite/vfx/blur.py:65-105

​
BlurOut

Blur-out effect that starts sharp and gradually becomes blurred.

​
Constructor

BlurOut(duration: float, max_intensity: float = 15.0)
​
duration
float
required
Duration of the blur-out effect in seconds, measured from the end of the clip.
​
max_intensity
float
default:"15.0"
Maximum blur intensity at the end of the effect.

​
Example

from movielite import VideoClip, vfx

# Become blurred over the last 3 seconds
clip = VideoClip("input.mp4")
clip.add_effect(vfx.BlurOut(duration=3.0, max_intensity=15.0))

​
How It Works

The blur is applied at the end of the clip:
  • t < (clip_duration - duration): No blur applied (sharp)
  • t ≥ (clip_duration - duration): kernel_size increases from 1 to max_intensity
Source: src/movielite/vfx/blur.py:107-149

​
Combining Blur Effects

​
Blur In and Out

clip = VideoClip("input.mp4")
clip.add_effect(vfx.BlurIn(duration=2.0, max_intensity=15.0))
clip.add_effect(vfx.BlurOut(duration=2.0, max_intensity=15.0))

​
Blur with Fade

clip = VideoClip("input.mp4")
clip.add_effect(vfx.FadeIn(1.0))
clip.add_effect(vfx.BlurIn(duration=1.0, max_intensity=15.0))
clip.add_effect(vfx.FadeOut(1.5))

​
Performance Considerations

Blur effects use OpenCV’s optimized GaussianBlur function, which is highly efficient even at high intensities. However, higher blur intensities will increase processing time.

​
Intensity Guidelines

  • Light blur: 3-7
  • Medium blur: 9-15
  • Heavy blur: 17-25
  • Extreme blur: 27+
Very high blur intensities (>50) can significantly slow down rendering. For most use cases, intensities between 5-20 provide good visual results with reasonable performance.

​
Technical Details

  • Uses OpenCV’s cv2.GaussianBlur with automatic sigma calculation
  • Kernel size must be odd and ≥ 1 (enforced automatically)
  • For animated blur, kernel size is recalculated for each frame
  • Standard deviation (sigma) is set to 0, letting OpenCV calculate it from kernel size

​
See Also

Build docs developers (and LLMs) love

Get started for freeTalk to us