Fixed
- Embedded VobSub OCR now works. mkvextract writes the
.idxpalette:line with a bare,(no space) after every 4th colour, which subtile-ocr's strict parser rejected ("error during palette parsing" → exit 1 → track stayedund). The.idxpalette is now normalized to a uniform,before OCR. Also surface subtile-ocr's stderr on failure instead of just "exit status 1."
Docker images
Pinned to this release:
docker pull ghcr.io/i-ial9000/shrinkerr:0.9.23 # portable (amd64 + arm64, CPU)
docker pull ghcr.io/i-ial9000/shrinkerr:0.9.23-nvenc # x86_64 + NVIDIA GPU (NVENC)
docker pull ghcr.io/i-ial9000/shrinkerr:0.9.23-edge # as above but ffmpeg master
docker pull ghcr.io/i-ial9000/shrinkerr:0.9.23-edge-nvencFloating tags (:latest, :edge, :nvenc, :edge-nvenc) track the most recent release of each lineage.