v0.3.11

Fixed

• Remote CPU workers handed an NVENC job no longer pick catastrophic libx265 settings. The old translation mapped nvenc p6 / CQ 20 to libx265 slower / CRF 16 — a near-lossless preset 40× slower than fast. Workers now prefer the server's configured libx265 defaults, fall back to a conservative translation capped at slow, and use CRF = CQ + 3 (matching libx265's higher per-bit efficiency, instead of the inverted sign the old code had).

Docker images

Pinned to this release:

docker pull ghcr.io/i-ial9000/shrinkerr:0.3.11         # portable (amd64 + arm64, CPU)
docker pull ghcr.io/i-ial9000/shrinkerr:0.3.11-nvenc   # x86_64 + NVIDIA GPU (NVENC)
docker pull ghcr.io/i-ial9000/shrinkerr:0.3.11-edge    # as above but ffmpeg master
docker pull ghcr.io/i-ial9000/shrinkerr:0.3.11-edge-nvenc

Floating tags (:latest, :edge, :nvenc, :edge-nvenc) track the most recent release of each lineage.