- Removed open_unlocked() — open() is the only entry
- Removed redundant init() call from open_title()
- init() called once in open(), handles everything
- Each function does one thing: open→init→scan→read
status() (sub_cmd 0x13) returns ILLEGAL REQUEST on some drives.
Was retrying 6× with 30s timeouts = 180s hang during init().
Steps 1-10 all pass on hardware:
unlock: OK, load_firmware: OK, calibrate: OK,
register_a: OK, register_b: OK
Only status fails — not needed for reads.
When MPLS STN table parsing drifts (disc-specific alignment issue),
a PGS subtitle entry (coding_type 0x90/0x91) can appear in the audio
stream section. Previously this showed as garbled "ng PGS 5.1" audio.
Fix: guard in stream builder checks if audio-typed streams have
subtitle codecs and reclassifies them as subtitles.
Also: unknown stream types now filtered out (filter_map) instead of
creating fake Video entries that showed as "?" in output.
Tested on V for Vendetta BD — "ng PGS 5.1" gone, clean output.
A (): single WRITE_BUFFER → verify 0x45 → unlock×2
B (): WRITE handshake → READ 0x3000 → WRITE 16B → verify → unlock×5
9/10 handlers are identical A/B. Only load_firmware has different logic.
Both paths end with do_unlock() — firmware upload is a prerequisite for
unlock, not a substitute. init() tries unlock first, falls back to
load_firmware only on failure (cold boot).
Every handler traced instruction-by-instruction from operation: do_unlock with configurable response size
operation: WRITE_BUFFER + verify buf=0x45 + unlock×2
operation: do_unlock → validate → send pre-built CDB → [4:20]
operation: same with CDB B
operation: init → scan 0x0000-0x5800 → build table → triple speed
operation: ↔x86 VM only (host_write 16B), no SCSI
operation: do_unlock → validate → probe 0x13 → check sig → features
operation: 3 paths by param count (1/5/9), dynamic READ_BUFFER
operation: search 64-entry table → position probe →
set_cd_speed_max → custom SET_CD_SPEED with matched value
operation: ↔x86 VM only (host_read 8B), no SCSI
init() matches x86 dispatch exactly:
Phase 1: unlock → [load_fw] × 6
Phase 2: calibrate × 6
Phase 3: probe (drive info)
Phase 4: register A + B × 5
Phase 5: status × 6
Handlers 5/9 are VM communication (no SCSI equivalent in Rust).
All other handlers send real SCSI commands.
DriveProfile now has every field traced from firmware:
- drive_signature, unlock_init_value, unlock_response_size_minus_init
- ld_microcode (base64, ~1888B firmware payload)
- hardware_register_a_cdb, hardware_register_b_cdb (10B pre-built CDBs)
- drive_nominal_speed_cdb (12B calibration speed)
- speed_zone_table (28B), speed_calc_table (25B)
drive.rs simplified:
- open() calls init() instead of unlock()
- init() is the ONLY entry point — handles full dispatch sequence internally
- Removed read_config, read_register, maintain_speed, read_sectors from public API
- Added set_read_speed() for per-zone speed during content reads
- disc.rs updated to call init() instead of unlock()
Compiles clean, all tests pass.
BD drives reject the MediaTek unlock command (sense 0x05).
Unlock is only needed for UHD raw access. Standard BD reads
work with standard READ(10) without vendor unlock.
Back to basics: open, unlock, SET CD SPEED max, read.
Remove all calibration probes, register reads, maintain_speed calls.
This is closest to the build that hit 17 MB/s earlier.
Also: drive discovery moved to libfreemkv (find_drive, resolve_device),
AACS via UDF only, clean pipeline, sg device support.
- Fixed: extents were relative to m2ts file, not absolute disc LBAs
- Fixed: u16 truncation of remaining sector count (13M → 36!)
- Added: UdfFs::file_start_lba() for m2ts LBA lookup
- Added: DriveSession::read_content() with 30s timeout for bulk reads
- Added: SET CD SPEED 0xFFFF on title open
- Added: adaptive batch reading (96→48→3 on error, ramp back up)
- Rip working end-to-end: scan → AACS → decrypt → write
Architecture:
- Each BD-J format in own file: paramount.rs, criterion.rs, pixelogic.rs, ctrm.rs
- Standard interface: detect() → bool, parse() → Option<Vec<StreamLabel>>
- PARSERS array in mod.rs — drop in a new parser with one line
- Shared vocab.rs for BD spec codec names only (MLP→TrueHD, AC3→Dolby Digital)
- All other label data passes through raw from disc — no guessing
Changes:
- New: paramount.rs (playlists.xml — Paramount/onQ format)
- Renamed: bluray_project.rs → pixelogic.rs
- Renamed: stream_properties.rs → criterion.rs
- Merged: language_streams.rs + menu_base.rs → ctrm.rs
- Removed: jar module (superseded by labels), dead apply functions
- Added: DriveSession::eject() with PREVENT ALLOW MEDIUM REMOVAL
- Added: DiscRegion enum (Free/BluRay/Dvd)
- Fixed: capture sector ranges now include all files (only skip STREAM/)
- Renamed: StreamLabel.region → variant (not a BD spec field)
labels::apply(session, udf, titles) does everything internally.
Disc::scan() is one line: crate::labels::apply().
No intermediate variables, no fallback logic in disc.rs.
App reads disc.titles[].streams — labels already applied if
disc had config files, empty otherwise. MPLS data always there.
src/labels/ with 4 parsers tried in order:
1. language_streams.txt (Warner CTRM CSV)
2. menu_base.prop (Warner CTRM properties)
3. streamproperties.xml + playbackconfig.xml (Criterion XML)
4. bluray_project.bin (Pixelogic binary tokens)
Disc::scan() calls labels::extract() → apply_disc_labels().
If no disc files found, streams keep MPLS data as-is.
No JAR fallback — disc files or nothing.
Covers 4/8 discs with JARs (Dunkirk UHD, V for Vendetta BD,
Being There, Barbie). Remaining 3 (Civil War, Dune, V for
Vendetta UHD) have no disc config files.
Traces <clinit> bytecode to find ldc/putstatic pairs that map
enum field names to display name strings. Pattern: new X, dup,
ldc "English", invokespecial X.<init>, putstatic X.a.
Proven on Dune UHD JAR: aw.a="English", aw.b="French", etc.
Next: trace mapping class (ISO code → enum field) to complete the chain.
Labels like eng_MLP_ now match to the correct stream by checking language
and codec hint against the stream's properties. Fixes Barbie where labels
were swapped (TrueHD label was on DD stream). TextField format (Civil War)
still uses index matching.
TrackLabel now carries language + codec_hint for structured matching.
- Extract all class strings once, try format parsers in chain
- Format 1: TextField,Audio{N} (A24/Lionsgate) — fixes Civil War
- Format 2: eng_MLP_ label strings (Warner UHD) — Barbie, Dune
- Format 3: playlist-only (MAIN_FEATURE etc.)
- Simplified TrackLabel to just description + raw
- 5 unit tests for both formats
- Tested on 12 disc captures: Civil War now gets 3 audio + 2 sub labels
Known issues: Barbie labels swapped, Dune returns 0 labels (different format?)
apply_jar_labels() matches JAR audio/subtitle labels to streams by
position and sets AudioStream.label. Apps read labels directly from
streams instead of doing their own JAR matching.
Each stream type has only its relevant fields. No more HDR on audio
or channels on video. Added SubtitleStream.forced field (TODO: parse).
Removed display helpers from lib (belongs in CLI).
- Size from source_packet_count * 192 (matches libbluray approach)
- metadata_sector_ranges() uses actual metadata partition size from ICB
instead of arbitrary +256 margin (fixes Top Gun missing CLIPINF)
- Store metadata_sectors in UdfFs
- Fix CLPI CPI bit-packed field parsing (partial, EP map still needs work)
- Disc.volume_id: UDF Volume ID from PVD (always present)
- Disc.meta_title: from META/DL/bdmt_eng.xml (falls back to other languages)
- Disc.format: UHD/BluRay/DVD detected from video codec
- Disc.capacity_bytes, Disc.layers
- Disc.jar_labels: extracted from BDMV/JAR
- Fixed MPLS STN parsing: 16-byte header (was 8), proper stream entry offsets
- Streams now include: HDR, color space, Dolby Vision EL, secondary audio/video
- parse_dstring() for UDF d-string fields
- wait_ready() polls TEST UNIT READY before unlock
Tested on 12 disc captures — all return correct titles, streams, format.
Walk UDF tree to discover sector ranges for all metadata files
(MPLS, CLPI, JAR, AACS certs, etc). Skips STREAM/, BACKUP/,
DUPLICATE/, MKB_RO.inf, ContentHash. Exposes partition_start()
and metadata_start() getters. Used by bdemu smart capture.
- Error codes as public constants (E_DEVICE_NOT_FOUND etc.) — single source of truth
- SCSI opcodes as named constants (SCSI_READ_10, SCSI_REPORT_KEY, etc.)
- AACS key class constant, MKB format constant
- No magic numbers in CDB construction
- docs/disc-to-rip.md — end-to-end flow from disc insert to decrypted content
- Links all module docs together as a starting point
File data extents are in the physical partition (partition_start + lba),
not the metadata partition. ICBs (directory/file entries) are in metadata,
but the data they point to is in physical. This one-line fix makes all
MPLS playlists readable — previously only 1/18 had correct MPLS magic.
Tested on Civil War UHD: 18 playlists, main movie at 00400.mpls (57min, 6 clips).
- AacsState now exposes: version (1/2), bus_encryption, mkb_version,
disc_hash, key_source (KeyDb/KeyDbDerived/ProcessingKey/DeviceKey),
vuk, unit_keys, read_data_key, volume_id
- All keys are public — apps that contribute to KEYDB ecosystem can
read VUK, unit keys, Volume ID, disc hash for database submission
- resolve_keys reports which path succeeded (key_source 1-4)
- Cleaned up resolve_keys with shared builder to reduce duplication
- Device key subset-difference tree: aesg3 key derivation, v_mask calc,
tree traversal from device key to processing key to media key
- MKB SCSI read: REPORT DISC STRUCTURE 0x83 with multi-pack support
- resolve_keys now has 4 paths:
1. disc hash → KEYDB → VUK
2. KEYDB media key + VID → VUK
3. MKB + processing keys → media key → VUK
4. MKB + device keys → processing key → media key → VUK
- setup_aacs reads MKB from drive (not just from file)
- AACS 2.0 detection: drive cert type 0x11 detected, falls back to
AACS 1.0 handshake (P-256 crypto path prepared but not yet built)
- STN table parsing in mpls.rs: video format/rate, audio format/rate/lang,
subtitle lang, coding type — all streamed into Disc title streams
- 31 tests passing
Parses device keys, processing keys, host cert, per-disc VUKs.
Full KEYDB.cfg (60MB, 180K entries) parsed in 1.85 seconds.
All test discs found with VUKs.
Next: disc hash computation, title key decryption, content decryption.
Prevents indefinite hangs on encrypted/unreadable sectors.
Some discs put MPLS/CLPI/JAR data in AACS-encrypted regions
that block reads without proper AACS authentication.
Scans BDMV/JAR/*.jar class files for track label patterns:
eng_MLP_ → TrueHD, eng_ADES_US_ → Descriptive Audio (US)
dan_PGStream4 → Danish subtitles
Solves the "3x AC-3 English" problem — we identify Descriptive Audio, commentary, compatibility tracks.
Future: set MKV "visual impaired" flag automatically during mux.
Local only — needs more disc samples to validate before publishing.
- scsi::open() returns Box<dyn ScsiTransport> on any platform
- Linux SG_IO behind #[cfg(target_os = "linux")]
- macOS IOKit and Windows SPTI stubs with TODO
- drive.rs uses scsi::open() — no platform-specific imports
- Compiles on macOS (returns "platform not yet supported")
MattJackson