Home
Jellyfin-FFmpeg is a maintained FFmpeg fork with targeted improvements for Jellyfin media server workloads. It focuses on better HDR handling, more robust hardware-accelerated pipelines, and practical fixes that improve streaming reliability and playback compatibility.
- Better Dolby Vision handling for remux and streaming (HLS fMP4 and MPEG-TS), including side-data preservation and improved RPU workflows.
- Faster, higher-quality HDR to SDR tone mapping with multiple backends (CUDA, OpenCL, Vulkan, Videotoolbox/Metal, and an SIMD-optimized software filter).
- Stronger hardware-acceleration pipelines across vendors (NVIDIA/Intel/AMD/Apple/Rockchip) with interop glue, bug workarounds, and performance tuning.
- Streaming and muxing fixes that reduce client playback issues (HLS/segment muxer improvements, Safari fMP4 fixes, Dolby Atmos pass‑through in mov, etc.).
- Extra codec and filter support relevant to home media (e.g., AC‑4 decoding for ATSC 3.0; practical overlays/deinterlacing/transpose in hardware).
For more info about additional features, see Features
- Users streaming HDR and Dolby Vision libraries to mixed clients.
- Systems relying on hardware acceleration (QSV/VAAPI/NVENC/Videotoolbox/AMF/RKMPP).
- Servers with Apple Silicon, Intel iGPUs, NVIDIA GPUs, AMD GPUs, and Rockchip SoCs.
- Devices sensitive to streaming edge cases (HLS, fMP4, MPEG‑TS, subtitles).
We upstream bug fixes when feasible, but many feature patches are intentionally maintained here due to different priorities and constraints in upstream FFmpeg. See: Upstreaming.
- Simple direct play or basic software transcodes without HDR, DV, or HW accel.
If you already use upstream FFmpeg and everything works, that’s fine. Jellyfin‑FFmpeg exists to reduce playback issues and improve efficiency in workloads common to Jellyfin users.