Re: [Mjpeg-users] Mjpegtools 1.6.2 rc1 (Upgrading!)
On Tuesday 26 August 2003 20:14, Bernhard Praschinger wrote: > Hallo > > I remember that I have changed that loop option. At least I applied the > patch from a user to have that feature. > > I trie to answer all mails about that topic in one mail. Might aswell. :) > > Uh. That would mean cvs access, right? Guess I'll have to wait, then. :) > > You can get it if you want ... just tell you your SF user ;) Didn't have any at the time. Now I do, and CVS access too. I committed the --max-file-frames change of lavrec a few minutes ago; I hope it worked out ok on the server... > If you want only one pass you do not need the -l option, that is the > default behavior of jpeg2yuv. Because of that -l 1 was used to for "loop > forever". Strange logic I know. A bit tricky to handle in the code. > If you really want a differnet behavior of the option please send me a > manpage that describes how the programm should behave after the next > change. ;) > > I have read the different solution, I really don't know which I sould > use. I chose the "-l -1 is forever, -l n=>1 is number of loops" because this is consistent with the way the already existing -n works. It felt somewhat odd to have two flags with very similar functions and completely different ways of specifying essentially the same thing. And it was easier to handle in the code. :) How this is resolved in the end isn't something I'm religious about as long as I can use it to convert single images from single files, but I feel that it might be a good thing to keep backwards compatibility, at least until a more significant version bump than a 0.0.1 one, and to keep the argument handling consistent with similar arguments within the same tool. Sounds like you're already on top of this loop thing, and will handle it? /Sam --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Mjpegtools 1.6.2 rc1 (Upgrading!)
Hallo I remember that I have changed that loop option. At least I applied the patch from a user to have that feature. I trie to answer all mails about that topic in one mail. > > Perfect, I would have proposed the same thing. I don't see any advantage > > for the current behaviour as default over the old behaviour. > Cool. > > > I'll apply ... ehm... damn, I'm gone all week ('till monday). Please > > apply yourself or wait for me to apply it next monday. :). > > Uh. That would mean cvs access, right? Guess I'll have to wait, then. :) You can get it if you want ... just tell you your SF user ;) > Anyway, I found two more suspicious-looking lines in jpeg2yuv.c: > > in main() > > mjpeg_info("Wp1\n"); I removed that in my version. > and in parse_commandline() > > mjpeg_info("Parsing char %c\n", c); That prints out the current option parsed. I removed that too. > Given the fact that they output information before the v flag is parsed, and > thus cannot honor it, they look like forgotten debug code. Very likely. > Wouldn't this be so much simpler? > > -l num > Specifies the number of loops (default: 1 loop) > The ranges of images will be looped through `num' times, by > default just once. The value of 0 will loop indefinitely. > > > The -n option in ppmtoy4m and lavpipe use 0 to mean all. For lav2yuv it's -f > 0 to mean all frames. jpeg2yuv and png2yuv use -1 for all. There are different version of how to do it. I guess that on I inserted was on the the less common. > Interesting code in line 464 of jpeg2yuv.c: > > if (param->loop != 1) >loops--; > > Somehow, I believe this is a typo and should read '-1' instead of '1'. > Could you re-try with that change? I'm rather soure that 1 is correct, and it is the same way as it is described in the manpage. The only bug I found if if you did not want to loop. It still looped forever that change fixed that a few line after the if statment aboth: } while(loops >=1 ); Now I'd guess I have a proper working version. With a unusual command syntax. ;) If you want only one pass you do not need the -l option, that is the default behavior of jpeg2yuv. Because of that -l 1 was used to for "loop forever". Strange logic I know. If you really want a differnet behavior of the option please send me a manpage that describes how the programm should behave after the next change. ;) I have read the different solution, I really don't know which I sould use. auf hoffentlich bald, Berni the Chaos of Woodquarter Email: [EMAIL PROTECTED] [EMAIL PROTECTED] www: http://www.lysator.liu.se/~gz/bernhard --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: Fw: [Mjpeg-users] Mjpegtools 1.6.2 release candidate 1 (1.6.1.90)
Hallo > I also noticed that using either -f8 or -f9 that something wasn't quite right with > some sort of timestamps. I did a 'mplayer test.m2v -ss 15:00' for example and that > really took me about 45 minutes into the movie (maybe a little further). This > worked when I encoded using the older version of mpeg2enc, doesn't with the new. That seems to be a problem of every VBR stream and mplayer. mplayer seeks always the maximal bitrate for the file forward. If you have set a bitrate of 5000k, and the bitrate for the next time is about 2500k it will jump ahead the double amout of time you wanted. That's no problem of mpeg2enc or mplex. auf hoffentlich bald, Berni the Chaos of Woodquarter Email: [EMAIL PROTECTED] [EMAIL PROTECTED] www: http://www.lysator.liu.se/~gz/bernhard --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Re: Mjpegtools 1.6.2 release candidate 1(1.6.1.90) (Ronald Bultje)
Hey Steven, On Tue, 2003-08-26 at 18:00, Steven M. Schultz wrote: > It'll probably take 24 hours for SF to catch up but perhaps Ronald > will be creating a new release candidate tarball in a couple days > or so. I'll wait for Martin to commit some of his proposed fixes for other issues in jpeg2yuv. After that, I'll do a new RC (.91), let it float around for a week and finally release 1.6.2 or so. Ronald -- Ronald Bultje <[EMAIL PROTECTED]> --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Re: Mjpegtools 1.6.2 release candidate 1 (1.6.1.90)(Ronald Bultje)
On Mon, 25 Aug 2003, Steven M. Schultz wrote: > Typo. The default limit for both formats (DVD and DVD-like) > is 7500. For now rely on the usage summary (mpeg2enc -h) until > the manpage is repaired. I've fixed the manpage for mpeg2enc and the -f 8 and -f 9 description reversal has been taken care of (as has the mention of the default bitrate for -f 9 which is 7500, the same as for -f 8). It'll probably take 24 hours for SF to catch up but perhaps Ronald will be creating a new release candidate tarball in a couple days or so. Cheers, Steven Schultz --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Weird interlacing -- 3:2 pulldown?
On Tue, 26 Aug 2003, Scott Bigham wrote: > [Hmm, this hasn't shown up on the list for twelve hours, so I'm assuming > it got lost in the shuffle and resending it. We Apologize for the > Inconvenience.(TM)] SF is just backlogged - sometimes worse than other times. I've seen stuff I sent out delayed enough I had forgotten I sent it ;) > > :pserver:[EMAIL PROTECTED]:/cvsroot/kino Yep - smilutils is off of the kino project, should be straightforward to check out the latest smilutils > > How are you doing the capture? With 'dvgrab' perhaps? > > Yep, --format raw. Should I be using something else? (I'm hoping to dvgrab with the raw format is a very good way to do it - I use that myself and then kino to do the trimming/cutting. > > If you play the raw DV data what does it look like? > > The combing effect is visible with playdv, most clearly on changes of U, I thought the problem was 'juddering' which would indicate a field order problem of some type. "combing" is just an unfortunate side effect of displaying interlaced video on a computer/progressive display. > camera angle or objects moving against a contrasting background. I can > also see the effect in Kino, particularly when stepping frame-by-frame. You might try deinterlacing on playback to see if that makes a difference. I know MPlayer+ffmpeg offer a number of deinterlacing filters that can be tried. Steven Schultz --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: Fw: [Mjpeg-users] Mjpegtools 1.6.2 release candidate 1 (1.6.1.90)
On Mon, 25 Aug 2003, Ray Cole wrote: > It went from 6 hours encoding a 2 hour movie to 24 hours (total time > includes filters, which I did not change filters between tests). > top shows mpeg2enc using 99% of the CPU (which I would expect). I > don't see any dip in CPU usage so I don't believe I'm seeing a stall but am willing to try. 200 char lines are a bit much... Any how, that kind of performance dropoff would indicate that Bernhard's suggestion that the MMX optimizations are not being used. When mpeg2enc starts up do you see lines like the following? INFO: [mpeg2enc] SETTING 3DNOW and EXTENDED MMX for QUANTIZER! INFO: [mpeg2enc] SETTING EXTENDED MMX for MOTION! INFO: [mpeg2enc] SETTING MMX for TRANSFORM! INFO: [mpeg2enc] SETTING EXTENDED MMX for PREDICTION! If not then the MMX/3DNOW instructions are not being used and that will cause a major slowdown. It sounds like the cpu detection isn't working right on your system or 'nasm' isn't installed so that the assembly portions can be built. Steven Schultz --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: Fw: [Mjpeg-users] Mjpegtools 1.6.2 release candidate 1 (1.6.1.90)
On Mon, 25 Aug 2003, Ray Cole wrote: > Here's the command line I used on the old one vs the new: > > Old Command: > nice mpeg2enc -f 8 -b ${aRate} -V 230 -n n -s -a 2 -g 6 -G 18 -I 0 \ >-r 24 -4 2 -2 2 -F 1 -p -v 0 -o ${aName}.m2v > > New Command: > nice mpeg2enc -f 9 -b ${aRate} -V 230 -n n -s -a 2 -g 6 -G 18 -I 0 \ >-r 24 -4 2 -2 2 -N 1.50 -Q 1.00 -v 0 -p -F 1 -o ${aName}.m2v -N 1.5 is a bit aggressive (but is the same as the old -N though), 1.0 would achieve most of the bitrate reduction but better looking output. > I also noticed that using either -f8 or -f9 that something wasn't quite > right with some sort of timestamps. I did a 'mplayer test.m2v > -ss 15:00' for example and that really took me about 45 minutes into > themovie (maybe a little further). This worked when I encoded using > the older version of mpeg2enc, doesn't with the new. Possible to set the line length to something closer to 80 or so? That paragraph came out as a single very long line... The problem's with mplayer's seeking. Mplayer probably isn't using the timestamps but is instead assuming CBR and a fixed GOP size and positioning accordingly. With VBR and variable GOP sizes seeking is a bit trickier - it's (almost) necessary to sequentially scan the file looking for the correct timestamp. DVDs produced with the new encoder work flawlessly in standalone players whereas the older mplex/mpeg2enc did have issues in that area. Sounds like bugs in the older version interacted with mplayer to accidentally get the desired result ;) I'd have to check what the maximum GOP size is for NTSC DVDs - thought it was 15 intead of 18 but I might be misremembering that though. What's the speed of the cpu being used? The command looks ok and I'd expect somewhere around 6 frames/sec or so on a 2.xGHz system. Steven Schultz --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Mjpegtools 1.6.2 rc1 (Upgrading!)
On Tuesday 26 August 2003 12:07, Ronald Bultje wrote: > Perfect, I would have proposed the same thing. I don't see any advantage > for the current behaviour as default over the old behaviour. Cool. > I'll apply ... ehm... damn, I'm gone all week ('till monday). Please > apply yourself or wait for me to apply it next monday. :). Uh. That would mean cvs access, right? Guess I'll have to wait, then. :) Anyway, I found two more suspicious-looking lines in jpeg2yuv.c: in main() mjpeg_info("Wp1\n"); and in parse_commandline() mjpeg_info("Parsing char %c\n", c); Given the fact that they output information before the v flag is parsed, and thus cannot honor it, they look like forgotten debug code. /Sam --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Weird interlacing -- 3:2 pulldown?
[Hmm, this hasn't shown up on the list for twelve hours, so I'm assuming it got lost in the shuffle and resending it. We Apologize for the Inconvenience.(TM)] On Sun, Aug 24, 2003 at 12:09:03PM -0700, Steven M. Schultz wrote: > On Sun, 24 Aug 2003, Scott Bigham wrote: > > Actually, I'm probably using the 0.2.0 release; I couldn't find any > > mention of an associated CVS repository. > It is a bit hard to track down - try this: > > :pserver:[EMAIL PROTECTED]:/cvsroot/kino Oh, Sourceforge. Okay, I know how to get stuff from there. > How are you doing the capture? With 'dvgrab' perhaps? Yep, --format raw. Should I be using something else? (I'm hoping to automate this, so Kino is probably out.) > If you play the raw DV data what does it look like? The combing effect is visible with playdv, most clearly on changes of camera angle or objects moving against a contrasting background. I can also see the effect in Kino, particularly when stepping frame-by-frame. -sbigham --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Mjpegtools 1.6.2 rc1 (Upgrading!)
On Tue, 26 Aug 2003, Martin Samuelsson wrote: > The man page has this to say: > >-l num > Specifies the nummber of loops (default: 0 loops ) > When this option is not used the given range of images is only > processed once. If you use this option and as number 1, jpeg2yv > will loop forever writing the image to stdout. When you use n > 1 > it will loop. n-time till it finishes. Wouldn't this be so much simpler? -l num Specifies the number of loops (default: 1 loop) The ranges of images will be looped through `num' times, by default just once. The value of 0 will loop indefinitely. The -n option in ppmtoy4m and lavpipe use 0 to mean all. For lav2yuv it's -f 0 to mean all frames. jpeg2yuv and png2yuv use -1 for all. I suggest using -n for frame count. head and tail use -n for line for number of lines. -f seems like it should take a filename as an argument, like tar and grep. Why is it, jpeg2yuv and ppmtoy4m? How about jpegtoy4m? YUV also refers a colorspace, while y4m is probably only used to refer to a stream format also known as YUV4MPEG, read with functions like y4m_read_stream_header(). --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Mjpegtools 1.6.2 rc1 (Upgrading!)
On Tuesday 26 August 2003 10:25, Ronald Bultje wrote: > Interesting code in line 464 of jpeg2yuv.c: > > if (param->loop != 1) >loops--; > > Somehow, I believe this is a typo and should read '-1' instead of '1'. > Could you re-try with that change? Could be. Just changing that line didn't make any difference, though. In generate_YUV4MPEG(), there is a loops = param->loop;. Param->loop seem to be zero, regardless of -n, loops-- decrements loops and the first comparison while(loops) does is to a negative loops. It all seems to boil down to a section of parse_commandline(): case 'l': param->loop = atoi(optarg); if (param->loop == 0) mjpeg_error_exit1( "-l option requires a number greater than 0"); break; Earlier, param->loop is initialized to 0, and if one doesn't supply -l (which I didn't do), the zero check isn't performed. The man page has this to say: -l num Specifies the nummber of loops (default: 0 loops ) When this option is not used the given range of images is only processed once. If you use this option and as number 1, jpeg2yv will loop forever writing the image to stdout. When you use n > 1 it will loop. n-time till it finishes. which almost is consistent with 1.6.1.90's behaviour. However, I'd like to propose a couple of changes: Line 464: 1: if (param->loop != -1) loops--; This will break the previous meaning of -l: In 1.60.1.90 "-l 1" means "loop forever". This would change "loop forever" to "-l -1". With this change, "-l 1" means "one copy", "-l 2" "two copies", and so on. 2: static void parse_commandline(int argc, char ** argv, parameters_t *param) { param->loop = 1; } Initialize loop to 1, meaning one copy, instead of the illegal and possibly-never-catched 0. 3: In the usage text: " -l numloop -1=forever, l >= 1 n-times\n" 4: In the man page: -l num Specifies the number of loops (default: 1 loop ) When this option is not used the given range of images is only processed once. If you use -l -1, jpeg2yuv will loop forever writing the image to stdout. When you use l >= 1 it will loop n-times until it finishes. (The man page mentions n in the l instructions. Correct or a typo?) With these changes, -n 1 -l 1 gives one copy of one file, -n 2 -l 1 gives one copy each of two files, and -n 1 -l 2 gives two copies of one file. Guess what -n 2 -l 2 does? Does this make sense at all? /Sam --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
[Mjpeg-users] mjpeg 1.6.1.90 libjpeg-mmx configure change
The help from the configure script no longer lists the --with-jpeg-mmx option, even though it appears to still exist. If the script doesn't guess where jpeg-mmx is, there is no error most users (especially those who don't know to look) are going to notice. The result is no MMX jpeg, which is a pretty big slowdown. Why does the configure script check to see that g77 works? Is someone planning to create a codec in FORTRAN? The horror, the horror... --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Mjpegtools 1.6.2 rc1 (Upgrading!)
Hey Martin, On Tue, 2003-08-26 at 10:11, Martin Samuelsson wrote: [..] Interesting code in line 464 of jpeg2yuv.c: if (param->loop != 1) loops--; Somehow, I believe this is a typo and should read '-1' instead of '1'. Could you re-try with that change? Ronald -- Ronald Bultje <[EMAIL PROTECTED]> --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Mjpegtools 1.6.2 rc1 (Upgrading!)
On Tuesday 26 August 2003 08:22, Ronald Bultje wrote: > Hey Martin, > > On Tue, 2003-08-26 at 01:33, Martin Samuelsson wrote: > > With 1.6.0, I could give jpeg2yuv an absolute filename, and get the > > expected output. With 1.6.1.90, it will go into a tight loop. Is there > > any change I've overlooked? > > -n 1, I suppose? I'm not sure I understand what you mean. :) Some output: If I run the new jpeg2yuv I get this: [EMAIL PROTECTED] sam]$ jpeg2yuv -v 2 -n 1 -f 25 -I p -j snap-00.jpg >frame.yuv INFO: [jpeg2yuv] Wp1 INFO: [jpeg2yuv] Parsing char v INFO: [jpeg2yuv] Parsing char n INFO: [jpeg2yuv] Parsing char f INFO: [jpeg2yuv] Parsing char I INFO: [jpeg2yuv] Parsing char j INFO: [jpeg2yuv] Parsing & checking input files. --DEBUG: [jpeg2yuv] Analyzing snap-00.jpg to get the right pic params INFO: [jpeg2yuv] YUV colorspace detected. INFO: [jpeg2yuv] Starting decompression INFO: [jpeg2yuv] Image dimensions are 320x240 INFO: [jpeg2yuv] Movie frame rate is: 25.00 frames/second INFO: [jpeg2yuv] Non-interlaced/progressive frames. INFO: [jpeg2yuv] Frame size: 320 x 240 INFO: [jpeg2yuv] Number of Loops 0 INFO: [jpeg2yuv] Now generating YUV4MPEG stream. --DEBUG: [jpeg2yuv] Preparing frame INFO: [jpeg2yuv] Processing non-interlaced/interleaved snap-00.jpg, size 8001l. INFO: [jpeg2yuv] Rescaling color values. --DEBUG: [jpeg2yuv] Frame decoded, now writing to output stream. --DEBUG: [jpeg2yuv] Preparing frame INFO: [jpeg2yuv] Processing non-interlaced/interleaved snap-00.jpg, size 8001l. INFO: [jpeg2yuv] Rescaling color values. --DEBUG: [jpeg2yuv] Frame decoded, now writing to output stream. ...and those last four lines will repeat eternally, while jpeg2yuv happily writes an ever increasing file. With the old jpeg2yuv, I get this: [EMAIL PROTECTED] sam]$ 160jpeg2yuv -v 2 -n 1 -f 25 -I p -j snap-00.jpg >frame.yuv INFO: [160jpeg2yuv] Parsing & checking input files. --DEBUG: [160jpeg2yuv] Analyzing snap-00.jpg to get the right pic params INFO: [160jpeg2yuv] Image dimensions are 320x240 INFO: [160jpeg2yuv] Movie frame rate is: 25.00 frames/second INFO: [160jpeg2yuv] Non-interlaced/progressive frames. INFO: [160jpeg2yuv] Frame size: 320 x 240 INFO: [160jpeg2yuv] Now generating YUV4MPEG stream. --DEBUG: [160jpeg2yuv] Preparing frame INFO: [160jpeg2yuv] Processing non-interlaced/interleaved snap-00.jpg, size 8001l. --DEBUG: [160jpeg2yuv] Frame decoded, now writing to output stream. And then it exits, with one yuv frame in the resulting file. With -n 0: [EMAIL PROTECTED] sam]$ jpeg2yuv -v 2 -n 0 -f 25 -I p -j snap-00.jpg >frame.yuv INFO: [jpeg2yuv] Wp1 INFO: [jpeg2yuv] Parsing char v INFO: [jpeg2yuv] Parsing char n INFO: [jpeg2yuv] Parsing char f INFO: [jpeg2yuv] Parsing char I INFO: [jpeg2yuv] Parsing char j INFO: [jpeg2yuv] Parsing & checking input files. --DEBUG: [jpeg2yuv] Analyzing snap-00.jpg to get the right pic params INFO: [jpeg2yuv] YUV colorspace detected. INFO: [jpeg2yuv] Starting decompression INFO: [jpeg2yuv] Image dimensions are 320x240 INFO: [jpeg2yuv] Movie frame rate is: 25.00 frames/second INFO: [jpeg2yuv] Non-interlaced/progressive frames. INFO: [jpeg2yuv] Frame size: 320 x 240 INFO: [jpeg2yuv] Number of Loops 0 INFO: [jpeg2yuv] Now generating YUV4MPEG stream. Here, it hangs in a tight loop, producing no output. It feels like there's been a change in an exit condition somewhere, as: [EMAIL PROTECTED] sam]$ 160jpeg2yuv -v 2 -n 0 -f 25 -I p -j snap-00.jpg >frame.yuv INFO: [160jpeg2yuv] Parsing & checking input files. --DEBUG: [160jpeg2yuv] Analyzing snap-00.jpg to get the right pic params INFO: [160jpeg2yuv] Image dimensions are 320x240 INFO: [160jpeg2yuv] Movie frame rate is: 25.00 frames/second INFO: [160jpeg2yuv] Non-interlaced/progressive frames. INFO: [160jpeg2yuv] Frame size: 320 x 240 INFO: [160jpeg2yuv] Now generating YUV4MPEG stream. will write 0 bytes of output, and then exit. With -n set to -1, both versions will write that single frame to the output file forever. /Sam --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: Fw: [Mjpeg-users] Mjpegtools 1.6.2 release candidate 1 (1.6.1.90)
It went from 6 hours encoding a 2 hour movie to 24 hours (total time includes filters, which I did not change filters between tests). top shows mpeg2enc using 99% of the CPU (which I would expect). I don't see any dip in CPU usage so I don't believe I'm seeing a stall but am willing to try. -- Ray On Sun, 24 Aug 2003 22:23:03 -0700 (PDT) "Steven M. Schultz" <[EMAIL PROTECTED]> wrote: > > On Sun, 24 Aug 2003, Ray Cole wrote: > > > I downloaded the source and built it. It seems to run about 4x slower > > than 1.6.1. Any ideas? > > Get a new watch? ;-)With out some numbers I find it very > difficult to believe "4x slower". > > For full sized DVD frames around 5 to 6 frames/sec or so is > normal with ~2.4GHz systems. Of course the time will depend on > if you're using any of the filtering programs such as yuvdenoise. > With a dual cpu system the filters will use up one cpu while the > encoder uses the other (and a bit more since it's threaded). > > That would mean you were seeing 20 frames/sec with 1.6.1? I > seriously doubt that - I'd like faster encoding but mpeg2enc > has never gotten up to that range on 720x480 or 720x756(PAL) > sized data. > > A couple things to make sure are being used are "-M 2" with the > encoder, and put a buffering stage in front of the encoder. I > use the 'bfr' program > > http://www.glines.org:8000/software/buffer.html > > something like " ... | bfr -10m | mpeg2enc ..." is the typical > usage. If the encoder is stalling for input due to latency in > earlier stages of the pipeline then 'bfr' will be of good value. > > Some timings of the same pipeline would be helpful of course ;) > > Cheers, > Steven Schultz > > > > > --- > This SF.net email is sponsored by: VM Ware > With VMware you can run multiple operating systems on a single machine. > WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines > at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 > ___ > Mjpeg-users mailing list > [EMAIL PROTECTED] > https://lists.sourceforge.net/lists/listinfo/mjpeg-users --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Re: Mjpegtools 1.6.2 release candidate 1 (1.6.1.90)(Ronald Bultje)
On 25 Aug 2003, Florin Andrei wrote: Not sure if/when this'll make it out - SF's mail system apparently has its knickers twisted. Sigh. > So, when looking at the man pages for the release candidate, i noticed > that "-f 9" is the new thing. Ai. If you're making DVDs then '-f 9' is absolutely the wrong thing to use because it leaves out the NAV packets that are needed. > In the mpeg2enc man page, there's a mention about -f 9 being more > appropriate for dvdauthor. The man page is wrong ;( Looks like the mpeg2enc manpage hasn't been updated sufficiently. > So, which way is recommended? Use -f 9 on both mpeg2enc and mplex when > authoring with dvdauthor? Or use -f 8? > Does it depend on the dvdauthor version? Use -f 8. There was a (thankfully brief) period when the new format (what is now -f 9) was added but the roles were reversed (which broke everyone's scripts of course). That was fixed (by leaving -f 8 as it had been and using 9 for the new format) but the manpage was apparently not reverted. At least we know folks are reading the manpages ;) > Also, why the bitrate for -f 9 is set to 10200 kbps? Isn't that too > high? (i thought there's a limit imposed by the DVD standard that's > quite close to 10 Mbit) Typo. The default limit for both formats (DVD and DVD-like) is 7500. For now rely on the usage summary (mpeg2enc -h) until the manpage is repaired. Steven Schultz --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Mjpegtools 1.6.2 rc1 (Upgrading!)
Hey Martin, On Tue, 2003-08-26 at 01:33, Martin Samuelsson wrote: > With 1.6.0, I could give jpeg2yuv an absolute filename, and get the expected > output. With 1.6.1.90, it will go into a tight loop. Is there any change I've > overlooked? -n 1, I suppose? Ronald -- Ronald Bultje <[EMAIL PROTECTED]> --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Re: Mjpegtools 1.6.2 release candidate 1(1.6.1.90) (Ronald Bultje)
On Mon, 2003-08-25 at 09:58, Steven M. Schultz wrote: > There were issues with mplex and mpeg2enc that caused problems > (2GB limit, timestamps, etc) with some DVD players. Also some > fixes were made to interoperate with dvdauthor - dvdauthor had > some misassumptions about the MPEG2 header format and mplex > knows (now) how to work around those. So, when looking at the man pages for the release candidate, i noticed that "-f 9" is the new thing. In the mpeg2enc man page, there's a mention about -f 9 being more appropriate for dvdauthor. But in the mplex man page there's no such mention, and the explanations are not quite clear (at least to me). I made quite a few DVDs with -f 8 using the release candidate and then authored them with dvdauthor-0.5.3 and so far i had no issues with that. So, which way is recommended? Use -f 9 on both mpeg2enc and mplex when authoring with dvdauthor? Or use -f 8? Does it depend on the dvdauthor version? Also, why the bitrate for -f 9 is set to 10200 kbps? Isn't that too high? (i thought there's a limit imposed by the DVD standard that's quite close to 10 Mbit) -- Florin Andrei http://florin.myip.org/ --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Re: Weird interlacing -- 3:2 pulldown? (Steven M.Schultz)
On Mon, 25 Aug 2003, William R Sherman wrote: > Hmmm, I'd say that red&white are component audio, and yellow is composite > video. (a bit off the topic of the subject I'll admit, but just thought > I'd clarify.) ARGH! Wow, did I ever get that reversed/wrong. Sigh, fingers weren't just a little ahead of the brain - they were on a different planet. You're right of course. That'll teach me to try to type and author DVDs at the same time ;) Steven Schultz --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Re: Mjpegtools 1.6.2 release candidate 1 (1.6.1.90)(Ronald Bultje)
> I've found that doing as much as possible in 4:1:1 and converting > to 4:2:0 at the last moment produces better looking output at > > If you have a recent ffmpeg on the system (do a 'make' but skip > the 'make clean') then build the smilutils with something like: > > ./configure --with-avcodec \ >--with-avcodec-include=/usr/local/src/ffmpeg/libavcodec A bit of followup (and hopefully clarification ;)). In addition to the benefits of processing keeping the data in DV (4:1:1) format as long as possible using ffmpeg's DV decoder produces (from what I have seen) better "YUV4MPEG2 411" output than libdv does. You can see the effect by converting a frame using both methods to PPM and viewing the PPM files with the GIMP or other utility. The exact "why" libavcodec's output is apparently better than libdv's is unknown so far (if some one figures it out it'd be nice to hear about it ;)). Cheers, Steven Schultz --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] mimimum muxrate acceptable
Thanks for the reply (Stevens too :-) > > I am too long of a good 10 minutes and I'd like to reduce the muxrate. > D'oh! How long is that movie anyway? (timewise) 1h25... > While i suspect that increasing -q will produce a similar decrease in > space requirements, i'd rather try -N first. I am trying it right now. Will see the result tomorrow. Thanks a lot Edouard --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] mimimum muxrate acceptable
On Mon, 2003-08-25 at 16:50, Edouard Chalaron wrote: > Hi > There > Trying to fit the unfitable What is the minimum muxrate acceptable for a > DVD ? > I am too long of a good 10 minutes and I'd like to reduce the muxrate. D'oh! How long is that movie anyway? (timewise) > So far I have been using the -f 8 standard but could a > mpeg2enc -f 3 -b 8000 -B 96 with mplex -r 8250 be Ok ? > Or should I go to mpeg2enc -b 7500 ? > Knowing that the source is Hi8 how low can I go in -b in order to keep it > "watchable". Any experience with this ??? While i didn't explore exactly that kind of change, i had good results with using -N (or -N 1 with the recent Release Candidate version). It lowers the actual size/bitrate requirements without the need to change -b. It also has a good effect on the image quality, but that's probably because i'm using an amateur-grade CCD, not a professional camera. While i suspect that increasing -q will produce a similar decrease in space requirements, i'd rather try -N first. -- Florin Andrei http://florin.myip.org/ --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
Re: [Mjpeg-users] Mjpegtools 1.6.2 rc1 (Upgrading!)
Hi! On Monday 25 August 2003 15:01, Ronald Bultje wrote: > > abandoned that course when I was asked for libquicktime-devel, and found > > out that the libquicktime people didn't have any official rpms for > > download. The mjpegtools project can't be blamed for that, though. > > If I'm correct, using a --nodeps should work. We don't specifically need > it. I just added it so that normal people will understand that they can > get quicktime support by this. And libquicktime rpms are available on freshrpms.net... Alexei --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users
[Mjpeg-users] mimimum muxrate acceptable
Hi There Trying to fit the unfitable What is the minimum muxrate acceptable for a DVD ? I am too long of a good 10 minutes and I'd like to reduce the muxrate. So far I have been using the -f 8 standard but could a mpeg2enc -f 3 -b 8000 -B 96 with mplex -r 8250 be Ok ? Or should I go to mpeg2enc -b 7500 ? Knowing that the source is Hi8 how low can I go in -b in order to keep it "watchable". Any experience with this ??? Thanks a lot Edouard --- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here:http://www.vmware.com/wl/offer/358/0 ___ Mjpeg-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mjpeg-users