ctrlsw_decoder --input <bitstream> --output <raw> --<input format> [options]
ctrlsw_decoder –-help
ctrlsw_decoder –h
Following is an
example:
ctrlsw_decoder -avc -in bitstream.264 -out decoded.yuv -bd 8
ctrlsw_decoder -hevc -in bitstream.265 -out decoded.yuv -bd 10
ctrlsw_decoder -jpeg -in bitstream.jpg -out decoded.yuv
The options are as follows:
| Command | Description |
|---|---|
--help,-h
|
Shows this help |
--help-json
|
Show this help (json) |
--version
|
Show version |
--input,--in,-in,--i,-i
<string>
|
Input bitstream |
--output,--out,-out,--o,-o
<string>
|
Output YUV |
--avc,-avc
|
Specify the input bitstream codec (default: HEVC) |
--hevc,-hevc
|
Specify the input bitstream codec (default: HEVC) |
--jpeg,-jpeg
|
Specify the input bitstream codec (default: HEVC) |
--framerate,--fps,-fps
<number>
|
force framerate |
--clock,--clk,-clk
<number>
|
Set clock ratio, (0 for 1000, 1 for 1001) |
--bitdepth,--bd,-bd
<string>
|
Output YUV bitdepth (8, 10, 12, alloc (auto), stream, first) |
--output-format
<string>
|
Output format FourCC (default: auto) |
--sync-i-frames
|
Allow decoder to sync on I frames if configurations's nals are presents |
--wavefront-parallel-processing,--wpp,-wpp
|
Wavefront parallelization processing activation |
--low-latency-decoding,--lowlat,-lowlat
|
Low latency decoding activation |
--slice-latency,--slicelat,-slicelat
|
Specify decoder latency (default: Frame Latency) |
--frame-latency,--framelat,-framelat
|
Specify decoder latency (default: Frame Latency) |
--no-reordering
|
Indicates to decoder that the stream doesn't contain B-frame & reference must be at best 1 |
--fbuf-format,--ff,-ff
|
Specify the format of the decoded frame buffers (raster, tile, tile64 (64x4 tiling), raster-msb (16-bit per sample, with MSB left-aligned)) |
--fbuf-format-out,--ffo,-ffo
|
Specify the format of the decoded frame buffers returned to the user (raster, raster-msb (16-bit per sample, with MSB left-aligned)). Default value aligned on --fbuf-format. |
--split-input
|
Send stream by decoding unit |
--split-from-sizes
<string>
|
Send stream by decoding unit |
--sei-file
<string>
|
File in which the SEI decoded by the decoder will be dumped |
--hdr-file
<string>
|
Parse and dump HDR data in the specified file |
--prealloc-args
<string>
|
Specify stream's parameters: 'widthxheight:video-mode:chroma-mode:bit-depth:profile:level' for example'1920x1080:progr:422:10:HEVC_MAIN:5'. video-mode values are: unkwn, progr or inter. Be careful cast is important. |
--output-position
|
Specify the position of the decoded frame in frame buffer |
--decode-intraonly
|
Decode Only I Frames |
| [Run] | |
--max-frames
<number>
|
Abort after max number of decoded frames (approximative abort) |
--loop,-loop
<number>
|
Number of Decoding loop (optional) |
--timeout
<number>
|
Specify timeout in seconds |
--embedded,-embedded
|
Force usage of embedded mcu |
--next-chan
|
Start the configuration of a new decoding channel. The options that are applied on all channels must be specified in the first channel. |
--exit-on
|
Specify early exit condition (e/error, w/warning) |
| [Trace & Debug] | |
--multi-chunk
|
Allocate luma and chroma of decoded frames on different memory chunks |
--input-buffer-count,--nbuf,-nbuf <number>
|
Specify the number of input feeder buffer |
--input-buffer-size,--nsize,-nsize <number>
|
Specify the size (in bytes) of input feeder buffer |
--circular-buffer-size,-stream-buf-size
<number>
|
Specify the size (in bytes) of internal circular buffer size (0 = default) |
--crc_ip,-crc_ip
<string>
|
Output crc file |
--first-frame-to-trace,--t,-t <number>
|
First frame to trace (optional) |
--frame-to-trace-count,--num,-num <number>
|
Number of frames to trace |
--use-early-callback
|
Low latency phase 2. Call end decoding at decoding launch. This only makes sense with special support for hardware synchronization. |
--core,-core
<number>
|
Number of decoder cores |
--non-realtime
|
Specifies that the channel is a non-realtime channel |
--ddrwidth,-ddrwidth
<number>
|
Width of DDR requests (16, 32, 64) (default: 32) |
--nocache,-nocache
|
Inactivate the cache |
--device
|
Path of the driver device(s) file(s) used to talk with the IP. Default(s) are: /dev/allegroDecodeIP |
--noyuv,-noyuv
|
Disable writing output YUV file |
--md5
<string>
|
Filename to the output MD5 of the YUV file |
--log
<string>
|
A file where logged events will be dumped |
| [Misc] | |
--color
|
Enable color (Default: Auto) |
--no-color
|
Disable color |
--quiet,-q
|
Do not print anything |
--verbosity
<number>
|
Choose the verbosity level (-q is equivalent to --verbosity 0) |