: Raw DV files require both an audio stream and a video stream; if you do not provide audio, you will not get any output (empty file). If your source material has no audio, force a silent audio stream to be added by using the options -x ...,null -y dvraw (fill in the appropriate input video codec in the -x option). Do not add a ,null to the -y option!
Raw DV files require both an audio stream and a video stream; if you do not provide audio, you will not get any output (empty file). If your source material has no audio, force a silent audio stream to be added by using the options -x ...,null -y dvraw (fill in the appropriate input video codec in the -x option). Do not add a ,null to the -y option!
Dynamic quantisation leads to the best picture quality. Requesting static quantisation speeds up the encoder; use it if your system is too slow to capture in realtime: -y dvraw="qno=2", with a value of 2 for eg. noisy satelite television signal, and 1 for sharp sources.
The following text is a redacted version from docs/README.dv in old transcode releases.
By Peter Chiocchetti <pch14 at myzel dot net>
DV only supports two resolutions: 720x576 for PAL, and 720x480 for NTSC. possibly, there is also 1920x1080.
transcodes' encoder allows other frame sizes too; on playback, these will show as split images, or rolling seemingly random colours.
transcode may also encode the planes of the video in a nonstandard way: this shows as a green screen on playback. if the installed version of libdv is compiled with its defaults, you need to add --uyvy or --dv_yuy2_mode on the transcode line - on older versions, only if you use -V, which is always advised.
if you capture from a v4l source, there may be a green bar on the top or bottom edges of the picture. you should be able to crop the image, and uncrop it again, adding black borders, instead of green ones - be sure, the frame stays the above sizes. such a bar may be the result of a sampling error in the capture card or even in signal transmission.
if you want to show your video on tv, you need not care of few green pixels at the bottom at all, as between 5 and 10% of the image will be offscreen anyways - tv set and beamer displays are closer to 640x480 than to 720x576 for PAL systems - see eg. the title-safe region in editing software.
This line captures from a pal source to pal dv and crops 64 pixels from the bottom and adds it again as solid black - obviously the clipping params are non-intuitively, shouldnt the sequence be t,l,b,r?
transcode -q 2 -x v4l2 -i /dev/video0 \ -p /dev/dsp -g 720x576 -j 64,0,0,0 \ -Y -64,0,0,0 -y dvraw="qno=2" -o rec.dv \ -H 0 -n 0x1 -N 0x1 --uyvy --print_status 25