← Back to Gallery

Error Diffusion Dithering Comparison

Stucki vs Floyd-Steinberg vs Atkinson vs Sierra

Original
Floyd-Steinberg
* 7/16
3/16 5/16 1/16
Stucki
* 8/42 4/42
2/42 4/42 8/42 4/42 2/42
1/42 2/42 4/42 2/42 1/42
Atkinson
* 1/8 1/8
1/8 1/8 1/8
    1/8
Bayer 4×4 (Ordered)
Threshold matrix
(not error diffusion)
Sierra
* 5/32 3/32
2/32 4/32 5/32 4/32 2/32
    2/32 3/32 2/32
Sierra Lite
* 2/4
1/4 1/4
Jarvis-Judice-Ninke
* 7/48 5/48
3/48 5/48 7/48 5/48 3/48
1/48 3/48 5/48 3/48 1/48

Controls

Error Diffusion Dithering

Error diffusion works by quantizing each pixel to black or white, then spreading the quantization error to neighboring pixels according to a diffusion kernel. This creates smooth gradients and preserves image detail.

Floyd-Steinberg (1976): Classic algorithm, good balance of quality and speed.
Stucki (1981): Larger kernel, smoother gradients, less dithering noise.
Atkinson: Only diffuses 75% of error, preserving more contrast.
Sierra/JJN: Even larger kernels for maximum smoothness.

Serpentine scanning (alternating left-to-right and right-to-left) reduces directional artifacts common in error diffusion algorithms.