How to Use -svtav1-params in FFmpeg

The -svtav1-params flag in FFmpeg is used to pass native, fine-grained configuration options directly to the Scalable Video Technology AV1 (SVT-AV1) encoder. When encoding video with the libsvtav1 library, standard FFmpeg flags only cover basic settings like bitrate, CRF, and presets. Using -svtav1-params allows you to customize advanced encoder features—such as film grain synthesis, tiling, and specific tuning modes—that are not exposed by standard FFmpeg options.

Syntax and Usage

The -svtav1-params flag accepts a list of key=value pairs separated by colons. The basic syntax is:

-svtav1-params key1=value1:key2=value2:key3=value3

When integrated into an FFmpeg command, it looks like this:

ffmpeg -i input.mp4 -c:v libsvtav1 -crf 26 -preset 5 -svtav1-params tune=0:film-grain=8:tile-rows=1:tile-columns=2 output.mkv

Common Parameters You Can Configure

By using -svtav1-params, you gain access to a wide range of native SVT-AV1 parameters. Some of the most commonly used options include:

• tune: Determines the quality optimization metric.
  • 0: Visual Quality (VQ) / Subjective tuning (optimized for human eyes).
  • 1: High Quality (HQ) / Objective tuning (optimized for PSNR/SSIM metrics).
• film-grain: Enables film grain synthesis. This removes heavy grain from the source to save bitrate and synthetically adds it back during playback. It accepts a value from 1 to 50 to represent the strength of the grain.
• film-grain-denoise: Controls whether the encoder denoises the input video before applying film grain synthesis (0 for disabled, 1 for enabled).
• tile-columns and tile-rows: Controls video tiling, which divides the video frames into a grid. This is highly useful for multi-threading. Setting tile-columns=1 splits the frame into two columns, allowing parallel encoding and decoding.
• scd: Enables or disables scene change detection (1 to enable, 0 to disable). Disabling this can force a strict keyframe interval.
• scd-mode: Selects the scene change detection algorithm mode (0 to 2), allowing you to balance speed and accuracy in detecting scene cuts.

Why Use -svtav1-params?

Standard FFmpeg mappings for libsvtav1 are limited. If you want to achieve the absolute best visual fidelity, optimize your encoding speed using specific thread/tile configurations, or utilize AV1’s advanced film grain technology to compress grainy movies efficiently, the -svtav1-params flag is the primary tool to unlock these capabilities.