Gaussian Blur
What is Gaussian Blur?
Gaussian blur softens an image by mixing each pixel's colour with those around it, weighted so nearby pixels matter more than distant ones, producing a smooth, hazy result.
At a glance
- Also known as
- Gaussian smoothingSoft blurLens blur (in some applications)
- Used for
- Softening images and videoCreating depth-of-field effects in post-productionObscuring sensitive contentCompositing and visual effects
- Common tools
- Adobe photoshopAdobe after effectsDaVinci resolveGIMPOpenCV
- Related terms
- Depth of fieldBokehMotion blurCompositingNoise reduction
Ready to create?
Direct scenes, design characters, and ship full films
All-in-one AI creative platform with simple, transparent pricing, no speed throttles, and an infinite Canvas for max creativity.
How it compares
Gaussian blur softens an image uniformly in all directions by averaging neighbouring pixel values, creating a hazy, defocused look, whereas motion blur applies directional smearing along a defined axis to simulate the visual effect of fast movement across the frame.
Think of it like…
Imagine you took a photo and then pressed your finger gently onto the wet ink: the colours would spread out a little into each other, and everything would look softer than before. That is what Gaussian blur does to a digital image, but in a very carefully calculated way so the result looks natural, not smudged. When audiences see Gaussian blur in a film, it often helps focus attention on what is sharp in the frame and makes the scene feel more like what human eyes actually see.
Pro tip
When using Gaussian blur to create a depth-of-field effect in post-production, apply the blur to a duplicate layer and use a gradient mask to keep the foreground sharp while softening the background. Varying the blur radius across depth zones produces a more convincing result than applying a single uniform radius, because real lenses blur more as distance increases.
Types and variations
- While Gaussian blur is the most widely used form of image softening, several related blur types serve different purposes.
- Motion blur simulates the streaking caused by fast movement or a slow shutter, while radial blur creates a circular smearing effect from a central point.
- Box blur, a simpler cousin of Gaussian blur, averages pixels equally without weighting, producing a less natural, more angular result.
- Lens blur in applications like Photoshop attempts to simulate the optical bokeh created by a real camera lens, adding specular highlights and iris shape into the blurred areas for a more cinematic appearance.
- Smart blur and surface blur are variations that preserve edge detail while softening flat areas, making them particularly useful for skin retouching.
Ready to make your first scene in Morphic?
Try MorphicCommon use cases
- Portrait photographers and video editors use Gaussian blur to soften backgrounds and create the impression of shallow depth of field when the original capture lacked sufficient lens separation.
- Motion graphics artists apply it to text and graphic elements to create soft glow effects or to simulate depth within a layered composition.
- Privacy-focused content uses Gaussian blur to obscure faces, number plates, and personal information in documentary and news footage.
- In AI workflows, creators apply Gaussian blur selectively to adjust the perceived sharpness of AI-generated images before combining them with other elements.
Ready to create?
Direct scenes, design characters, and ship full films
All-in-one AI creative platform with simple, transparent pricing, no speed throttles, and an infinite Canvas for max creativity.
FAQs
Gaussian blur is a digital image processing technique that softens an image by averaging each pixel's value with its neighbours, using a bell-shaped weighting curve so that nearby pixels have more influence than distant ones. The result is a smooth, natural-looking softening similar to an out-of-focus photographic lens.
The technique takes its name from the Gaussian function, a mathematical formula for the normal distribution developed by the German mathematician Carl Friedrich Gauss. The bell-curve shape of this function determines the weighting pattern applied to surrounding pixels during the blur calculation.
The radius controls how many surrounding pixels are included in the averaging calculation. A small radius creates subtle softening, while a large radius produces a stronger, more diffuse blur. Higher radius values also increase processing time because more pixels are involved in each calculation.
Gaussian blur softens an image in all directions uniformly, simulating defocus or haze, while motion blur applies directional smearing to simulate the appearance of movement. They serve different creative purposes — Gaussian blur creates stillness and softness, while motion blur communicates speed and dynamic action.
Gaussian blur is widely used to obscure faces, number plates, and other identifying information in video and photography. Most editing applications include a tool that applies a blur to a defined region, and some platforms apply automatic face blurring using AI detection before publishing content.
Gaussian blur is referenced in AI workflows both as a prompt descriptor and as a post-processing tool. Creators use it to soften AI-generated images, to blend AI elements with live footage, and to describe desired atmospheric or defocused qualities when writing image or video generation prompts.
Gaussian blur removes fine detail and edge sharpness from an image, and this information cannot be recovered once the blur is applied to a flattened file. Working non-destructively: for example, using smart filters in Photoshop or applying blur to a separate layer: preserves the original data and allows adjustments later in the workflow.
Gaussian blur applies a uniform, mathematically smooth softening based on the Gaussian function, while lens blur simulates the optical characteristics of a real camera lens, including the shape of the aperture and the way specular highlights render as bokeh. Lens blur is more computationally complex but typically produces a more cinematic result.
