ResizeInterpolationMode

Balances speed and quality. Each destination pixel is computed from a weighted average of the 4 nearest source pixels. Produces smoother results than NearestNeighbor, but is faster and less accurate than Cubic. Recommended for general resizing when performance is important and high precision is not required.

Produces the highest quality results by considering a 16-pixel neighborhood for each destination pixel. This method is slower than Linear, but yields smoother gradients and fewer artifacts. Recommended for enlarging images or when maximum quality is desired.

The fastest and simplest method. Each destination pixel is assigned the value of the closest source pixel. This can result in jagged edges and blocky artifacts, but preserves hard edges. Recommended for binary or indexed images, or when speed is critical and quality is less important.

Computes each destination pixel as a weighted average of all source pixels that map to it. Provides the best results when significantly reducing image size, minimizing aliasing and moiré patterns. Recommended for downscaling images by large factors.

Uses a high-quality sinc-based filter to compute each destination pixel, providing sharp results with minimal artifacts. Slower than Linear and NearestNeighbor, but often superior for both upscaling and downscaling. Recommended for photographic images or when preserving detail is important.

Chooses the most appropriate interpolation method based on the scale factor: typically uses Cubic for upscaling and SuperSampling for significant downscaling. Recommended when unsure which mode to use, or for general-purpose resizing.