You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
New custom EBML/Matroska parser (mkv.py) reads the MKV Cues index to locate only the clusters containing PGS subtitle data, then seeks directly to them — bypassing FFmpeg entirely for extraction
For UHD Blu-ray remuxes (40 GB+), total I/O drops from tens of gigabytes to a few megabytes, making extraction dramatically faster
Only applies to MKV/MK3D files with subtitle Cues present (typically Matroska version 4, which includes Cue entries for non-video tracks). Older muxes that lack subtitle Cues fall back to FFmpeg transparently
Handles both zlib-compressed and header-stripped subtitle blocks
Parallel I/O with thread-local file handles for multi-track extraction
.mk3d Support
.mk3d (3D Matroska) is now a recognized container format throughout — file input, context menu, and direct cue parsing all support it
Windows Explorer context menu updated with proper submenu entries for .mk3d
Subtitle Count in Track Listing
MKV/MK3D track listing now shows the total subtitle count per track (read from the Cues index), giving a quick sense of track density before extraction
Other Improvements
Shell context menu restructured: .sup files get a simple "Validate" entry, while containers get both "Validate" and "Validate fast" options