Re: [Mjpeg-users] lavrec from v4l device input selection broken?
> You find here that: > p - PAL through Composite or 1st Bt8x8 input > P - PAL through SVHS- or 2nd Bt8x8 input > t - PAL through TV-tuner- or 3rd Bt8x8 input Yes, I read that, and misunderstood it. I thought it ment that Composite==1st Btx8 input, and that -p selected it. Thanks, Maarten --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
[Mjpeg-users] VCD 2500vbr better quality than SVCD
Hello, I have been tranfering some DV recording to VCD and SVCD, and I am a bit surprised by the result. I first created the VCD, with a non- standard bitrate of 2500, and I was surprised how good it looked (played with a DVD player on a TV). Then I created the SVCD (standard settings). expecting it to be even better, but in fact it contained a lot of artifacts (the wringling on large surfaces around borders, I don't know the slang for) Now, I realize the SVCD bitrate has been a bit lower than the VCD bitrate I used, but only slightly. Maybe the VCD is smoother because I used yuvdenoise in the chain? Or maybe because the VCD resolution is a lot lower, while this doesn't really matter on a TV set? Or maybe I just did something wrong when creating the SVCD? I send you the settings I used. Any suggestions will be greatly appreciated! Maarten cat oulala001.dv | raw2yuv -i 0 | yuvdenoise -F | yuvscaler -O VCD | mpeg2enc -f 2 -b 2500 -s -H -B 224 -o video.mpg cat oulala001.dv | raw2yuv -i 0 | yuvscaler -M BICUBIC -O SVCD | mpeg2enc -q 9 -f 4 -B 224 -I 1 -4 2 -2 1 -o /mnt/winF/oulalaSVCD/svcd.mpg --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] VCD 2500vbr better quality than SVCD
Am Sonntag, 5. Oktober 2003 10:48 schrieb Maarten De Boer: > Hello, > > I have been tranfering some DV recording to VCD and SVCD, and I am > a bit surprised by the result. I first created the VCD, with a non- > standard bitrate of 2500, and I was surprised how good it looked > (played with a DVD player on a TV). Then I created the SVCD (standard > settings). expecting it to be even better, but in fact it contained a > lot of artifacts (the wringling on large surfaces around borders, I > don't know the slang for) > > Now, I realize the SVCD bitrate has been a bit lower than the VCD > bitrate I used, but only slightly. Maybe the VCD is smoother because > I used yuvdenoise in the chain? Or maybe because the VCD resolution > is a lot lower, while this doesn't really matter on a TV set? Or > maybe I just did something wrong when creating the SVCD? The VCD resolution is a quarter of the resolution of SVCD ( half width/half height) and you are using the same bitrate as for SVCD. This means that with the vcd-resolution the encoder has enough room for scaling the bitrate, the second try with svcd resolution most likely hits max bitrate all the time and is resulting in continous artefacts. Depending on your TV Set you may not recognize any horizontal resolution that is bigger than 352. And yes yuvdenoise could have had an effect to. What you want to try is: -K kvcd yuvdenoise for the svcd maybe cvd resolution for svcd (352x576) playing with -q value playing with max bitrate (if your dvd standalone can handle more than standard svcd bitrate you may want to set the max bitrate higher, giving the encoder more room. > I send you the settings I used. Any suggestions will be greatly > appreciated! > > Maarten > > cat oulala001.dv | raw2yuv -i 0 | yuvdenoise -F | yuvscaler -O VCD | > mpeg2enc -f 2 -b 2500 -s -H -B 224 -o video.mpg > > cat oulala001.dv | raw2yuv -i 0 | yuvscaler -M BICUBIC -O SVCD | > mpeg2enc -q 9 -f 4 -B 224 -I 1 -4 2 -2 1 -o > /mnt/winF/oulalaSVCD/svcd.mpg > You have set the -q value for svcd to 9 for your vcd experiment to 4 (standard) , this may have too an effect, beside the resolution and yuvdenoise. I'm no expert either but HTH. Steffen --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] VCD 2500vbr better quality than SVCD
Hallo > I have been tranfering some DV recording to VCD and SVCD, and I am > a bit surprised by the result. I first created the VCD, with a non- > standard bitrate of 2500, and I was surprised how good it looked > (played with a DVD player on a TV). Then I created the SVCD (standard > settings). expecting it to be even better, but in fact it contained a > lot of artifacts (the wringling on large surfaces around borders, I > don't know the slang for) > > Now, I realize the SVCD bitrate has been a bit lower than the VCD bitrate > I used, but only slightly. Maybe the VCD is smoother because I used > yuvdenoise in the chain? Or maybe because the VCD resolution is a lot > lower, while this doesn't really matter on a TV set? Or maybe I just did > something wrong when creating the SVCD? You see if the maximal bitrate was reached when you multiplex the stream. (mjpeg-howto Section: Creating MPEG2 Videos and section: Optimizing the stream) > cat oulala001.dv | raw2yuv -i 0 | yuvscaler -M BICUBIC -O SVCD | mpeg2enc > -q 9 -f 4 -B 224 -I 1 -4 2 -2 1 -o /mnt/winF/oulalaSVCD/svcd.mpg mplex audio and video. If you average bitrate should be about 20% lower than the maximal bitrate. If not, than you'll very likely see artefacts. To lower the bitrate you can set a lower quality factor (10,11), use yuvdenoise maybe yuvmedianfilter, and mpeg2enc -h/--reduce-hf For my experience using a other quantisation matrix might save some bitrate, but often it saves less that 0,5%. BTW: SVCD is by default for PAL 480x576, VCD PAL: 352x288 BTW2: You can save some bitrate when you use a lower (192,160) audio bitrate, and increas the video bitrate. auf hoffentlich bald, Berni the Chaos of Woodquarter Email: [EMAIL PROTECTED] www: http://www.lysator.liu.se/~gz/bernhard --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] VCD 2500vbr better quality than SVCD
On Sun, 5 Oct 2003, Maarten De Boer wrote: > (played with a DVD player on a TV). Then I created the SVCD (standard > settings). expecting it to be even better, but in fact it contained a > lot of artifacts (the wringling on large surfaces around borders, I > don't know the slang for) The technical term is "not enough bits" :) > Now, I realize the SVCD bitrate has been a bit lower than the VCD bitrate > I used, but only slightly. Maybe the VCD is smoother because I used > yuvdenoise in the chain? Or maybe because the VCD resolution is a lot If yuvdenoise was used for VCD then it is even more important to use yuvdenoise with SVCD because proportionally there are fewer bits/pixel. I'll use NTSC dimensions - doing the calculations for PAL is left to the reader ;) VCDs have a standard bitrate of ~1150kb/s. You used the SVCD rate of 2500 which is ok if your player can handle it - but it is non-standard. VCD is 352x240 so there are 84480 pixels which gives 1152000/84480 bits/pixel or 13.6 SVCD is 480x480 or 230400 pixels/frame. To maintain the same number of bits/pixel as VCD one would need (480*480)/(352*240) * 1150kb/s or 2.727 the bitrate or 2.727*1152kb/s = 3136kb/s You can see that SVCD is on the edge of not having enough bits at 2500. > something wrong when creating the SVCD? First thing is that you didn't use yuvdenoise ;) > cat oulala001.dv | raw2yuv -i 0 | yuvdenoise -F | yuvscaler -O VCD | > mpeg2enc -f 2 -b 2500 -s -H -B 224 -o video.mpg Using ~2500kb/s for VCD will give great quality because there are so many bits available. > > cat oulala001.dv | raw2yuv -i 0 | yuvscaler -M BICUBIC -O SVCD | mpeg2enc > -q 9 -f 4 -B 224 -I 1 -4 2 -2 1 -o /mnt/winF/oulalaSVCD/svcd.mpg There are several things that can be done to reduce the average bitrate: 1) Add 'yuvdenoise -l 1' (a mild setting) to the pipeline 2) Add "-N 0.5" (or perhaps a little higher but values over 1.0 are not recommended) to the mpeg2enc parameters 3) Try the KVCD matrices - add a "-K kvcd" to the mpeg2enc options. 1 and 3 can be used at the same time. It might not hurt anything to use all three but I think using -N and "-K kvcd" at the same time will degrade the picture. -N and "-K tmpgenc" would probably work together ok though. There has been a report (I forget who did the tests) of "-K kvcd" producing very good looking SVCDs. I'd definitely add a yuvdenoise step. Good Luck! Steven Schultz --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
[Mjpeg-users] cut AC3 file
Any idea how to cut the first 10 frames of an AC3 file and save them into another file, without modifying any of the audio stream parameters? (bitrate, # of channels...) -- Florin Andrei http://florin.myip.org/ --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Linux Video Studio: Effects do not work with interlaced material
On Saturday 04 October 2003 14.34, Bernhard Praschinger wrote: > > http://www.nada.kth.se/~hanke/files/README > > Very fine too :) Updated, too. Font handling should now work. > > > Here is the patch: > > http://www.nada.kth.se/~hanke/files/lvs.diff > > Automatich patching worked. Here is an updated patch: http://www.nada.kth.se/~hanke/files/lvs_1.diff > > > Any suggestions and comments are welcome. > > Do you want CVS access to upload you changes direct to CVS ? > Else I can upload the changes to the CVS. Since I am not so deeply involved in the development, I would prefer to have somebody (you or Ronald) to critically evaluating my patches before being submitted to the CVS. > > A description how the filter works would be fine, Something with a help > button, that shows some basic steps. > I know that a documentation would be fine for everything in LVS. Yes. I spent much time to understand the inner working. Moreover, I am not very sure if I am using LVS the way the authors thought. Where can I find some information about the scene detection algorithm used in lavplay? Yours, Michael --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
[Mjpeg-users] lav2wav error
Hi, I met a problem with lav2wav. After a long editing process I prepared a rather complex *.eli file containing references to a number of *.avi files. Some of the avi files do not contain any sound track (but most do). lav2wav gives an error that there is now sound in the material. It would be very inconvenient for me to add 'silence' to all files without sound. Would it be possible to modify lav2wav such that, instead of generating an error, the output becomes a wav file with absolute silence? Thanks, Michael --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
