lav2wav(1) [debian man page]

lav2wav(1)							MJPEG tools manual							lav2wav(1)

NAME

       lav2wav - Extract the audio out of MJPEG container files to stdout

SYNOPSIS

       lav2wav [-s num] [-c num] [-v num] [-I] [-R] [-r samplerate, bitesize, channels] lavfile1 [lavfile2 ... lavfileN]

DESCRIPTION

       lav2wav	can  be  used to extract the audio to stdout.  This output goes to stdout and can be saved as a wav file or piped to another sound
       processing tool that is able to handle the wav format.  This can be mp2enc and toolame for mpeg layer 2 audio, or for example lame for mpeg
       layer 3 audio.

       The  input  files  may be any combination of AVI (.avi), Quicktime (.qt) or editlist files so long as they are all lavtools- readable (e.g.
       MJPEG-encoded AVI/Quicktime or DV type 2 AVI).

OPTIONS

       lav2wav accepts the following options:

       -s num
	    Start extracting at video frame (num)

       -c num
	    Extract (num) frames of audio

       -v num
	    Verbosity level (0, 1 or 2)

       -I   Ignore unsupported bitrates/bits per sample

       -R   If the file does not contain any sound. lav2wav will create silence with 44100kHz Sampelrate, 16 Bit audio bitsize and 2 Chanels

       -r sr,bs,ch
	    If the file does not contain any sound lav2wav will generate silence with the values you supply  the  samplerate  (sr),  audio-bitsize
	    (bs) and channel (ch).

BUGS

       The  "WAV"  file  format (technically: RIFF) is really very much less than ideal for a tool intended to be used in pipelines as lav2wav is.
       The problem is that the header includes a field specifying the length of the file.  This can't be filled in except by seeking back  to  the
       begining  and  over-writing.  If the output is unseekable (e.g. pipe) lav2wav simply writes a large length into the header and leaves it at
       that.  Most tools like sox(1) or mp2enc(1) either ignore the length field anyway or only give a warning.
       The audio length is inacurate calculated when lav2wav generates silence.  This happens only if you have NTSC framerate and than it  creates
       for every hour of video 1.1498sec too less of silence.

AUTHOR

       This man page was written by Bernhard Praschinger.
       If you have questions, remarks, problems or you just want to contact the developers, the main mailing list for the MJPEG-tools is:
	 mjpeg-users@lists.sourceforge.net

       For more info, see our website at
	      http://mjpeg.sourceforge.net

SEE ALSO

       mjpegtools(1), mp2enc(1), sox(1)

MJPEG Linux Square						    2 June 2001 							lav2wav(1)

lav2yuv(1)							MJPEG tools manual							lav2yuv(1)

NAME

       lav2yuv - Convert a MJPEG file to raw yuv

SYNOPSIS

       lav2yuv [options] lavfile1 [lavfile2 ... lavfileN]

DESCRIPTION

       lav2yuv	converts  an  MJPEG  video sequence described by a sequence of MJPEG video files and/or edit lists pointing to such files into the
       simple uncompressed planar 4:2:0 Y'CbCr format, as used by mpeg2enc(1) MPEG encoder and	image  processing  filters  like  yuvscaler(1)	or
       yuv2dfilter(1) or yuvdenoise(1).

       Output is to stdout so that by piping the output of lav2yuv into a suitable pipeline it is possible to process and then encode or play back
       video recorded in any of the mjpegtools MJPEG container formats: AVI, quicktime or edit lists describing editted versions of video held	in
       such files.

       Mixing different files with different video formats is currently not possible.

       The -S -T -D options are used for scene detection which is used by linux video studio.

OPTIONS

       lav2yuv accepts the following options:

       -m   Force mono-chrome

       -c   Conceal frames containing corrupt MJPEG data by repeating the preceeding good frame.

       -S list.el
	    Output a scene list with scene detection

       -T num
	    Set scene detection threshold to num (default: 4)

       -D num
	    Width decimation to use for scene detection (default: 2)

       -o num
	    Frame offset - skip num frames at the beginning.  if num is negative all but the last num frames are skipped

       -f num
	    Only num frames are written to stdout. 0 means that all frames are processed.

       -A width:height
	    Set/override  the  sample  aspect  ratio  (SAR)  emitted in the output stream header.  Currently, the SAR is only auto-detected for DV
	    files, and guessed otherwise.

       -P width:height
	    Set the intended/approximate display aspect ratio (DAR) for the stream.  This value is only used to guess the SAR when  it	cannot	be
	    detected  from  the  input	stream.   The default DAR is '4:3' (corresponding to a standard TV screen); another useful value is '16:9'
	    (corresponding to widescreen TV).

       -C chroma
	    Set output chroma (default: '420jpeg').  Currently, '420jpeg', '420mpeg2', '420paldv', '422' are available.

       -x   Exchange fields.  Useful if the field order was reversed during editing.

AUTHOR

       This man page was originally written by Bernhard Praschinger.
       If you have questions, remarks, problems or you just want to contact the developers, the main mailing list for the MJPEG-tools is:
	 mjpeg-users@lists.sourceforge.net

       For more info, see our website at
	      http://mjpeg.sourceforge.net

SEE ALSO

       mjpegtools(1), mpeg2enc(1), yuvscaler(1), yuv2lav(1), yuvdenoise(1), yuvplay(1)

MJPEG Linux Square						   6 March 2002 							lav2yuv(1)