OBS Studio Scale Filtering Guide

When resizing video sources in OBS Studio, the Scale Filtering option determines how the software recalculates pixels to maintain image clarity. This article explains how Scale Filtering affects your source quality, breaks down the different filtering algorithms available—such as Point, Bilinear, Bicubic, and Lanczos—and helps you choose the best setting for your streaming or recording setup.

What is Scale Filtering?

Scale Filtering in OBS Studio is the rendering method used when you scale a source (like a webcam, game capture, or image) away from its native resolution. If you stretch a 1080p source to fit a 4K canvas, or shrink a 1080p webcam to fit into a small corner of your screen, OBS must add or remove pixels. Scale Filtering dictates the mathematical algorithm used to perform this adjustment, directly impacting both visual sharpness and system performance.

By default, OBS applies a basic scaling method, but you can manually change this by right-clicking any source, navigating to Scale Filtering, and selecting one of the options.

The Scale Filtering Options and Their Quality Impact

1. Point (Nearest Neighbor)

• How it works: This method simply duplicates or deletes pixels to match the new size without blending them.
• Quality Impact: It produces a blocky, pixelated look. While it is terrible for high-definition video, webcams, or text, it is the absolute best option for retro games (like 8-bit or 16-bit titles) because it preserves the sharp, clean edges of pixel art without blurring them.
• Performance: Extremely low resource usage.

2. Bilinear

• How it works: This algorithm averages the colors of neighboring pixels to smooth out the transition.
• Quality Impact: Bilinear filtering results in a softer, slightly blurry image. It is particularly noticeable when scaling down high-resolution sources, as details like small text can become hard to read.
• Performance: Very low resource usage. It is ideal for low-end computers.

3. Bicubic (16 Samples)

• How it works: Bicubic scaling looks at a larger grid of neighboring pixels (16 samples) to calculate the color of the scaled pixels.
• Quality Impact: This is the recommended “sweet spot” for most users. It provides a sharp, clean image with minimal blurriness and handles both upscaling and downscaling very well.
• Performance: Moderate resource usage. Most modern graphics cards handle Bicubic scaling with no noticeable performance drop.

4. Lanczos (36 Samples)

• How it works: This is a highly complex algorithm that samples 36 surrounding pixels to reconstruct the scaled image.
• Quality Impact: Lanczos provides the sharpest, most detailed image quality possible in OBS, making it excellent for downscaling highly detailed gameplay or text-heavy sources. However, because it is so sharp, it can occasionally introduce “ringing” artifacts (faint halos around high-contrast edges).
• Performance: High resource usage. It should only be used if your GPU has headroom to spare.

5. Area

• How it works: Area filtering averages pixels in a specific frame area, scaling down by grouping pixels together.
• Quality Impact: This option is highly effective at reducing “shimmering” or temporal aliasing, which often occurs when high-detail textures (like chain-link fences or thin lines in games) are scaled down and motion is introduced.
• Performance: Low to moderate resource usage.

Summary Recommendation

To get the best visual quality out of your OBS sources, use this quick cheat sheet:

• For webcams, capture cards, and general gameplay: Choose Bicubic for a balanced, sharp image, or upgrade to Lanczos if you have a powerful PC and want maximum sharpness.
• For retro games and pixel art: Choose Point to keep the pixels sharp and blocky.
• For sources experiencing shimmering/flickering when shrunk: Choose Area to smooth out the jagged moving edges.