ld decode tools - happycube/ld-decode GitHub Wiki
ld-decode-tools is a suite of applications for processing the .tbc output from ld-decode. The tools enable various processing of the LaserDisc including VBI, 40-bit FM code, White-Flag, dropout detection/correction and comb filtering. The ld-decode-tools use a JSON based metadata file to store and communicate information about the .tbc. Details of the JSON format can be found on the JSON Metadata format wiki pages.
From the tools directory (of the ld-decode repo) type 'qmake' followed by 'make all'. Once all of the applications are compiled use 'sudo make install' to install and 'sudo make uninstall' to remove.
Note: The ld-decode-tools.pro qmake project file is designed only for command line compilation with Ubuntu. The individual application .pro files can be used within the Qt Creator IDE (you should build ld-decode-shared first in order to make a local copy of the required shared libraries).
So if, for example, your .tbc output file is test_video.tbc (and is NTSC):
ld-process-vbi test_video.tbc (add in the VBI data and determine the field order)
ld-dropout-correct test_video_map.tbc test_video_mapdoc.tbc (correct the dropouts and create a new .tbc)
ld-analyse test_video_mapdoc.tbc & (view the corrected .tbc)
ld-chroma-decoder test_video_mapdoc.tbc test_video_mapdoc.rgb (convert the .tbc to RGB 16-16-16 frames)
For a PAL .tbc output file a typical sequence would be:
ld-process-vbi test_video.tbc (add in the VBI data and determine the field order)
ld-discmap test_video.tbc test_video_map.tbc (maps the disc and removes duplicate frames)
ld-dropout-correct test_video.tbc test_video_doc.tbc (correct the dropouts and create a new .tbc)
ld-analyse test_video_doc.tbc & (view the corrected .tbc)
ld-chroma-decoder test_video_doc.tbc test_video_doc.rgb (convert the .tbc to RGB 16-16-16 frames)
Note that the VBI processing, NTSC processing, disc-mapping, drop-out detection and correction are optional.
Note: If compilation fails after pulling a new version of the tools from github, try running the following commands in the tools/ directory before filing a github issue report:
make distclean
qmake
make all
sudo make install
This application examines the input TBC file and determines the available VBI data for each available frame. The VBI data is stored as both the raw data value for the 3 VBI lines as well as a full decode of the VBI according to the IEC specifications. The resulting information is written back into the JSON metadata file for the TBC output.
For NTSC sources this application also processes items from the IEC NTSC specification that are specific to NTSC LaserDiscs such as 40-bit FM codes, white flags, etc.
Syntax:
ld-process-vbi <options> <input TBC file name>
Options:
-h, --help Displays this help.
-v, --version Displays version information.
-d, --debug Show debug
--input-json <filename> Specify the input JSON file (default input.json)
--output-json <filename> Specify the output JSON file (default same as
input)
-n, --nobackup Do not create a backup of the input JSON metadata
-t, --threads <number> Specify the number of concurrent threads (default
is the number of logical CPUs)
Arguments:
input Specify input TBC file
This application reads an ld-decode JSON metadata file, typically as produced by ld-process-vbi, and exports information in standard formats that other tools can read. At present, it can export:
- Per-frame signal quality information from the VITS test signals, as CSV
- Per-frame LaserDisc VBI control signals, as CSV
- LaserDisc navigation information, as FFMETADATA1 (which FFmpeg can use to add chapter navigation to a video file)
Syntax:
ld-export-metadata <options> <input>
Options:
-h, --help Displays this help.
-v, --version Displays version information.
-d, --debug Show debug
-q, --quiet Suppress info and warning messages
--vits-csv <file> Write VITS information as CSV
--vbi-csv <file> Write VBI information as CSV
--ffmetadata <file> Write navigation information as FFMETADATA1
Arguments:
input Specify input JSON file
This application uses the drop-out information in the JSON metadata file to perform dropout correction on the input TBC file and produces a new output file. The current version of the corrector uses framing in order to provide inter-field correction. Note that inter-field correction may not function correctly for NTSC pull-down sources.
Syntax:
ld-dropout-correct <options> <input TBC file name> <output TBC file name>
Options:
-h, --help Displays this help.
-v, --version Displays version information.
-d, --debug Show debug
--input-json <filename> Specify the input JSON file (default input.json)
--output-json <filename> Specify the output JSON file (default output.json)
-r, --reverse Reverse the field order to second/first (default
first/second)
-o, --overcorrect Over correct mode (use on heavily damaged sources)
-i, --intra Force intrafield correction (default interfield)
-t, --threads <number> Specify the number of concurrent threads (default
is the number of logical CPUs)
Arguments:
input Specify input TBC file (- for piped input)
output Specify output TBC file (omit or - for piped
output)
This application takes the PAL or NTSC TBC input video and performs chroma decoding (i.e colourises it). Output is a sequence of RGB 16-16-16 video frames suitable for pipelining to an external application such as ffmpeg.
Syntax:
ld-chroma-decoder <options> <input TBC file name> (<output RGB file name>)
The output RGB file name is optional. If omitted the output will be sent to stdout (for pipe-lining).
Most of the options here correspond to settings in ld-analyse's 'Chroma decoder configuration' window, so you can experiment with them interactively using ld-analyse. For more information on what the settings mean, see the ld-analyse User Guide.
The mono filter isn't available in ld-analyse. It treats the whole signal as being luma, so it's the best choice when you know that the input video doesn't contain any colour information (e.g. black-and-white films).
Options:
-h, --help Displays this help.
-v, --version Displays version information.
-d, --debug Show debug
-q, --quiet Suppress info and warning messages
--input-json <filename> Specify the input JSON file (default
input.json)
-s, --start <number> Specify the start frame number
-l, --length <number> Specify the length (number of frames to
process)
-r, --reverse Reverse the field order to second/first
(default first/second)
--chroma-gain <number> Gain factor applied to chroma components
(default 1.0)
-b, --blackandwhite Output in black and white
-f, --decoder <decoder> Decoder to use (pal2d, transform2d,
transform3d, ntsc1d, ntsc2d, ntsc3d,
ntsc3dnoadapt, mono; default automatic)
-t, --threads <number> Specify the number of concurrent threads
(default number of logical CPUs)
-o, --oftest NTSC: Overlay the adaptive filter map (only
used for testing)
-w, --white NTSC: Use 75% white-point (default 100%)
--chroma-nr <number> NTSC: Chroma noise reduction level in dB
(default 0.0)
--luma-nr <number> NTSC: Luma noise reduction level in dB
(default 1.0)
--simple-pal Transform: Use 1D UV filter (default 2D)
--transform-mode <mode> Transform: Filter mode to use (level,
threshold; default threshold)
--transform-threshold <number> Transform: Uniform similarity threshold in
'threshold' mode (default 0.4)
--transform-thresholds <file> Transform: File containing per-bin similarity
thresholds in 'threshold' mode
--show-ffts Transform: Overlay the input and output FFTs
Arguments:
input Specify input TBC file (- for piped input)
output Specify output RGB file (omit or - for piped
output)
This application does the opposite of ld-chroma-decoder: it reads a stream of RGB 16-16-16 frames, and encodes them into PAL or NTSC composite video, writing a TBC file as output. It's mostly useful for testing and development purposes -- for example, you can use it to generate TBC files from standard test videos, then look at the effects of decoding them with different options in ld-analyse or ld-chroma-decoder.
Syntax:
ld-chroma-encoder <options> <input RGB file name> <output TBC file name>
Specify the input filename as - to make it read from standard input.
Options:
-h, --help Displays help on commandline options.
--help-all Displays help including Qt specific options.
-v, --version Displays version information.
-d, --debug Show debug
-q, --quiet Suppress info and warning messages
-c, --sc-locked Output samples are subcarrier-locked. PAL only.
(default: line-locked)
-f, --system <system> Select color system, PAL or NTSC. (default PAL)
--chroma-mode <chroma-mode> NTSC only. Chroma encoder mode to use
(wideband-yuv, wideband-yiq; default:
wideband-yuv)
--no-setup NTSC only. Do not add 7.5 IRE setup (as in
NTSC-J)
Arguments:
input Specify input RGB file (- for piped input)
output Specify output TBC file
This GUI application provides a range of features for examining TBC output including drop-out detection, video extent, line scope and VBI data. The application works with NTSC and PAL TBC output files.
Note: The input file name is optional (you can either specify it from the command line or using the GUI once the application is running).
For more information about ld-analyse please see the ld-analyse User Guide
Syntax:
ld-analyse <options> <input TBC file name>
Options:
-h, --help Displays this help.
-v, --version Displays version information.
-d, --debug Show debug
Arguments:
input Specify input TBC file
This application processes the .efm output from ld-decode into either digital audio or data. The application implements both audio and data error detection and correction. Audio output is a stereo 44.1KHz 16-bit PCM file.
Syntax:
ld-process-efm <options> <input EFM file name> <output data file name>
Options:
-h, --help Displays help on commandline options.
--help-all Displays help including Qt specific options.
-v, --version Displays version information.
-d, --debug Show debug
-q, --quiet Suppress info and warning messages
-c, --conceal Conceal corrupt audio data (default)
-s, --silence Silence corrupt audio data
-g, --pass-through Pass-through corrupt audio data
-p, --pad Pad start of audio from 00:00 to match initial disc
time
-b, --data Decode F1 frames as data instead of audio
-t, --time Non-standard audio decode (no time-stamp information)
--debug-efmtof3frames Show EFM To F3 frame decode detailed debug
--debug-syncf3frames Show F3 frame synchronisation detailed debug
--debug-f3tof2frames Show F3 To F2 frame decode detailed debug
--debug-f2tof1frame Show F2 to F1 frame detailed debug
--debug-f1toaudio Show F1 to audio detailed debug
--debug-f1todata Show F1 to data detailed debug
Arguments:
input Specify input EFM file
output Specify output file
CLI options:
- Audio
- Conceal corrupt audio data - Attempts to conceal corrupt audio data using the concealment type selected in the following section
- Silence corrupt audio data - Silences any corrupt samples
- Pass-through corrupt audio data - Passes any corrupt samples through to the output as-is
- Concealment type
- Linear interpolation - Interpolates between last-good and next-good samples to conceal errors
- Options
- Pad start of audio from 00:00 to match initial disc time - Pads the start of the input sample to ensure the output starts from the 00:00 time of the disc. Useful for assembling partial captures.
- Decode F1 frames as audio - Decode F1 frames as audio data
- Decode F1 frames as data - Decode F1 frames as data
- Non-standard audio decode (no time-stamp information in input EFM)
- Detailed debugging for development
- EFM to F3 Frame debug
- F3 and section debug
- F3 to F2 Frame decoding debug
- F2 to F1 Frame decoding debug
- F1 Frame to audio samples debug
- F1 Frame to data sector debug
This application can accept 10-bit packed format .lds files and convert them to 16-bit signed files or vice-versa (useful for expanding .lds files and piping the output to a compression utility such as FLAC).
Syntax:
ld-lds-converter <options>
Options:
-h, --help Displays this help.
-v, --version Displays version information.
-d, --debug Show debug
-i, --input <file> Specify input laserdisc sample file (default is stdin)
-o, --output <file> Specify output laserdisc sample file (default is stdout)
-u, --unpack Unpack 10-bit data into 16-bit (default)
-r, --riff Unpack 10-bit data into 16-bit with RIFF WAV headers (use with -u)
-p, --pack Pack 16-bit data into 10-bit
This application scans through the metadata for a .tbc file and attempts to correct metadata and video data to ensure the .tbc contains the correct amount of frames, in the right order, without duplication. The application also pads the .tbc with dummy (black) frames when one or more frames are missing.
It is recommended to run ld-process-vbi on the input .tbc before running this tool.
Options:
-h, --help Displays this help.
-v, --version Displays version information.
-d, --debug Show debug
-r, --reverse Reverse the field order to second/first (default first/second)
-m, --maponly Only perform mapping, but do not save to target (for testing
purposes)
Arguments:
input Specify input TBC file
output Specify output TBC file
This application performs 'stacking' of multiple versions of the same disc (ideally different copies of discs containing the same mastered contents). Disc stacking requires a minimum of two input sources in order to work (although 3 or more are strongly recommended).
Stacking is performed by taking all available input sources and, pixel by pixel, determining the number of available sources for the pixel (by removing outliers marked as dropouts in the source files). If more than 3 sources are available for the pixel the tool will use the median value of the available sources as the output (when there are an even number available of pixel sources, the two centre values of the median are averaged to the output). If only two sources are available an average is used. If only one source is available it is passed directly as the output.
If no sources are available for a pixel the tool will attempt to use differential dropout detection (diffDOD) to recover a value (in the case where ld-decode marked the pixel as a false-negative) unless the --no-diffdod option is specified.
If the --passthrough option is specified the tool will, when all pixel sources are marked as a dropout, mark the resulting output pixel as a dropout (regardless of the diffDOD result). The --passthrough is useful in non-preservation cases where it is desirable for master plate errors (which cause dropouts in the same place on all resulting LaserDisc copies) to be marked for dropout concealment (where-as diffDOD would correctly identify the plate error as present on all discs (and therefore not an error)).
Use the tool by specifying the available input .tbc files followed by the desired .tbc output file.
Options:
-h, --help Displays this help.
-v, --version Displays version information.
-d, --debug Show debug
-q, --quiet Suppress info and warning messages
--input-json <filename> Specify the input JSON file for the first input
file (default input.json)
--output-json <filename> Specify the output JSON file (default output.json)
-r, --reverse Reverse the field order to second/first (default
first/second)
-t, --threads <number> Specify the number of concurrent threads (default
is the number of logical CPUs)
--no-diffdod Do not use differential dropout detection on low
source pixels
--passthrough Pass-through dropouts present on every source
Arguments:
inputs Specify input TBC files (- as first source for
piped input)
output Specify output TBC file (omit or - for piped
output)
This application performs an analysis of the VITS (Vertical Interval Test Signals) and recalculates the white and black SNR values in the metadata. Note that any TBC file will already have this metadata provided by ld-decode. This tool is generally used along with dropout correction and stacking as the video content of the TBC is modified - running ld-process-vits therefore allows you update the SNR metadata in order to analyse the result (using ld-analyse).
Use the tool by specifying the required input .tbc file. The tool will backup the previous JSON file (to .vbup) before proceeding.
Options:
-h, --help Displays this help.
-v, --version Displays version information.
-d, --debug Show debug
-q, --quiet Suppress info and warning messages
--input-json <filename> Specify the input JSON file (default input.json)
--output-json <filename> Specify the output JSON file (default same as
input)
-n, --nobackup Do not create a backup of the input JSON metadata
-t, --threads <number> Specify the number of concurrent threads (default
is the number of logical CPUs)
Arguments:
input Specify input TBC file
