Denoise Image
Denoise images online free
Image noise shows up as random speckles, grain, or colored dots — particularly in shadow areas and uniform surfaces like skies. It is most visible in photos taken in low light, at high ISO camera settings, or scanned from older prints. Denoising smooths these artifacts so the image looks clean without losing the actual detail in edges and textures.
The tradeoff is fine texture. Aggressive noise reduction can soften details like hair, fabric weave, and surface grain alongside the noise. Good denoising targets the flat, uniform regions — where randomness is clearly noise — while leaving structured edges intact.
What causes image noise?
Digital noise has two main sources. Luminance noise comes from the camera sensor amplifying a weak signal in low light, producing a grainy, film-like texture. Chroma noise is the colored version — random red, green, or blue speckles that appear when the sensor's color channels disagree on what color a pixel should be. Both become more pronounced at higher ISO values and longer exposures.
Compression also introduces noise-like artifacts. JPEG encoding in particular creates blocky patterns called compression artifacts in flat areas and around edges. These look similar to sensor noise but respond differently to denoising — the image compression guide explains how format choice and quality settings affect how many artifacts appear in the first place.
When to denoise and what to do after
- Before sharpening: always denoise first. Sharpening amplifies noise as aggressively as it clarifies edges — applying it to a noisy image makes the grain worse. Use sharpen image only after noise looks under control.
- Before tonal edits: brighten or correct contrast after denoising, since lifting shadows also lifts noise. Use the brightness tool or auto contrast as a follow-up pass.
- After denoising, before export: a light pass with sharpen can restore edge crispness that denoising softens slightly.
When not to denoise
Skip denoising on images where grain is intentional — film scans, stylized portraits, and cinematic edits often use visible grain as part of the aesthetic. Denoising those images removes mood along with texture. Evaluate the visual intent before applying any global smoothing pass.
For images destined for large-format print or high-resolution display, strong denoising can produce a plastic or overly smooth look at full size even if it appears fine in a small preview. Test at 100% zoom before committing to an export.
Denoise image FAQ
What is the difference between median and Gaussian smooth?
Median filtering replaces each pixel with the median of its neighborhood, which reduces salt-and-pepper speckles while preserving some edge structure at moderate strength. Gaussian smooth applies a soft blur that reduces grain uniformly but also softens fine detail and edges.
Why denoise before sharpening?
Sharpening boosts local contrast, which amplifies noise as well as edges. Cleaning noise first keeps sharpening from turning grain into harsh texture.
Will denoise remove intentional film grain?
Yes, global passes treat grain like noise. Skip denoise or use a low amount when grain is part of the look.