May 2007 - MEncoder-users - lists.mplayerhq.hu

Help with MPG video encoding
by Walter Belhaven 20 Apr '13

20 Apr '13

Hi, I've got a question on how to get the best possible MPG encoding of a movie source, "at any price" so to speak -- i.e., as many bits as the encoder needs to get the highest possible video quality. It appears to me that the MPG encoding process is introducing artifacts that are present regardless of the amount of "free reign" I give to the encoder with respect to quantization, bitrate, and the like. My bottom line question is thus: "Are these artifacts inherent in MPG encoding, or am I just not twiddling the knobs and dials on the encoder correctly?" Here's all the details ... My source is a mixed hard- and soft-telecined "movie" at 1080/60i that I'm trying to tanscode to 480/24p, soft telecined to 480/60i, for ultimate burning to an NTSC DVD-Video. When I use mencoder to inverse telecine, crop out the crud, expand back to the original letterbox ratio (it's 2.35:1, not 16/9), and 'encode' the video with HUFFYUV (422P), the resulting video looks stunning, with no 'blocking' artifacts to speak of. This, of course, is as expected. But when I encode the video in MPG2 or MPG4, even with vqscale=1 or vqscale=2, I see sometimes severe blocking in the resulting video. Is this because the chroma is being subsampled vertically and horizontally (4:2:0), or is it because 'vqscale' is (perhaps necessary but) not sufficient to get "maximum quality at any price" encoding? Thanks, WB ____________________________________________________________________________________ Be a PS3 game guru. Get your game face on with the latest PS3 news and previews at Yahoo! Games. http://videogames.yahoo.com/platform?platform=120121

3 9

-mpegopts format
by Nicolas Hesler 05 Jun '07

05 Jun '07

Would someone be able to give me the exact details/differences between -mpegopts format=mpeg2 and -mpegopts format=dvd. (What are the presets set by each of these options and why might it cause the problem below??) When I create an mpeg-2 for DVD using format=mpeg2 with a soft telecine, Adobe Encore will accept as a soft telecine file, when I use format=dvd, it does not and tells me the file needs to be transcoded yet again. Here is the command line I'm using to create my file. If anyone has any suggestions or can point out any problems/issues with the order or parameters included the command line, please pass them along, mencoder -oac lavc -ovc lavc -of mpeg -mpegopts format=mpeg2:tsaf:telecine -vf scale=720:480,harddup -srate 48000 -af lavcresample=48000:0:2 -lavcopts vcodec=mpeg2video:vrc_buf_size=1835:vrc_maxrate=8600:vbitrate=5000:keyint=12:vstrict=0:acodec=ac3:abitrate=192:aspect=16/9 -ofps 24000/1001 -o output.mpg input.mpg Thanks, Nicolas

4 9

raw dvcpro25 to quicktime compatible mp4-container (aac lc+x264)
by Johan Arvidsson 01 Jun '07

01 Jun '07

Hi! I´ve created a script that creates a quicktime compatible file from raw dvcpro25 material. The source files are from 600 Mb to ~4000 Mb big (3-20 minutes). If anybody is interested i can post my script but any real programmer that looks at it will probably cringe :) I hit a few bumps before i got it working, the official build for windows from mplayer.hq does'nt work with my files. It only does ten minutes of the source file and then stops processing it but the svn-builds i found here http://oss.netfarm.it/mplayer-win32.php works with files bigger than 2 gb. I tried the approach in the documentation (http://www.mplayerhq.hu/DOCS/HTML/en/menc-feat-quicktime-7.html) but i could not create files that worked well with quicktime that way. What i do now is that i encode audio and video separatly and then i mux them together with mp4box, audio and video is a little out of sync but not so much that it matters. There are a few things that bugs me and suggestions and tips are warmly welcomed: 1. I get a few green frames in the beginning of all video files i create when i play them with quicktime 7.1.6. I wonder if it could be related to the stats for the first encoded 12 frames in the example below. Is it something with my source files that is making that happen or if it is a known bug. I'd be happy to provide samples if anybody is interested. Command: %COMSPEC% /c C:\h264\tools\mencoder-sherpya-070529.exe C:\h264\source\IN-2007-0328-VASTNYTT19.dv -o NUL -ovc x264 -x264encopts pass=1:turbo=2:bitrate=700:bframes=1:me=umh:partitions=all:trellis=1:qp_step=4:qcomp=0.7:direct_pred=auto:keyint=25:threads=auto -nosound -cache 20480 -passlogfile C:\h264\temp\IN-2007-0328-VASTNYTT19.log -vf yadif,harddup -of rawvideo -ofps 25 -fps 25 2>radF8171.tmp 1>rad79DF3.tmp StdOut: MEncoder Sherpya-SVN-r23404-4.2.0 (C) 2000-2007 MPlayer Team CPU: Intel(R) Pentium(R) D CPU 3.00GHz (Family: 15, Model: 6, Stepping: 4) CPUflags: Type: 15 MMX: 1 MMX2: 1 3DNow: 0 3DNow2: 0 SSE: 1 SSE2: 1 Compiled with runtime CPU detection. success: format: 0 data: 0x0 - 0x266e100 libavformat file format detected. [lavf] Video stream found, -vid 0 [lavf] Audio stream found, -aid 1 VIDEO: [dvsd] 720x576 0bpp 25.000 fps 28800.0 kbps (3515.6 kbyte/s) [V] filefmt:35 fourcc:0x64737664 size:720x576 fps:25.00 ftime:=0.0400 Input fps will be interpreted as 25.00 instead. Opening video filter: [expand osd=1] Expand: -1 x -1, -1 ; -1, osd: 1, aspect: 0.000000, round: 1 Opening video filter: [harddup] Opening video filter: [yadif] ========================================================================== Opening video decoder: [dshow] DirectShow video codecs Decoder supports the following YUV formats: YUY2 UYVY Decoder is capable of YUV output (flags 0x9) VDec: vo config request - 720 x 576 (preferred colorspace: Packed YUY2) [PP] Using codec's postprocessing, max q = 4. Could not find matching colorspace - retrying with -vf scale... Opening video filter: [scale] VDec: using Packed YUY2 as output csp (no 0) Movie-Aspect is 1.37:1 - prescaling to correct movie aspect. VideoDecoder::SetExtAttr: registry failure Selected video codec: [qdv] vfm: dshow (Sony Digital Video (DV)) ========================================================================== Pos: 0.0s 1f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.0s 2f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.0s 3f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.0s 4f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.0s 5f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.1s 6f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.1s 7f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.2s 8f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.2s 9f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.2s 10f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.3s 11f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.3s 12f ( 0%) 0.00fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.4s 13f ( 0%) 12.88fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.4s 14f ( 0%) 13.55fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.4s 15f ( 0%) 14.27fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.5s 16f ( 0%) 14.86fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.5s 17f ( 0%) 15.53fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.6s 18f ( 0%) 16.07fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.6s 19f ( 0%) 16.73fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.6s 20f ( 0%) 17.20fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.7s 21f ( 0%) 17.81fps Trem: 0min 0mb A-V:0.000 [0:0] Pos: 0.7s 22f ( 0%) 18.24fps Trem: 0min 0mb A-V:0.000 [0:0] -------------------------------------SNIP-------------------------------------------------------- 2. Two pass encoding. In this example the first pass seems to process the whole file and the last frame is number 30106. But the log-file is still named divx2pass.log.temp (actually i use another name for the logfile but i figured if somebody else googles for divx2pass.log.temp they will find this in the archives and hopefully a good answer too) so i added a check in my script after the first pass where it looks for divx2pass.log.temp and if it finds that file it renames it to divx2pass.log. And it sort of works. It seems like the last chunk of logs is not written to the logfile before mencoder quits when it is done with the first pass. With the renamed file i get some errors at the end of the second pass but the resulting h264 file i can mux with the audio anyway. I lose about a second of video at the end but it does'nt matter much. Still it would be nice if it worked :) >From the logs of the second pass: x264 [error]: 2nd pass has more frames than 1st pass (30080) x264 [error]: continuing anyway, at constant QP=28 x264 [error]: disabling adaptive B-frames x264 [error]: 2nd pass has more frames than 1st pass (30080) x264 [error]: continuing anyway, at constant QP=28 x264 [error]: disabling adaptive B-frames x264 [error]: 2nd pass has more frames than 1st pass (30080) x264 [error]: continuing anyway, at constant QP=28 x264 [error]: disabling adaptive B-frames Assertion failed: h->frames.b_have_lowres, file encoder/slicetype.c, line 376 This application has requested the Runtime to terminate it in an unusual way. Please contact the application's support team for more information. -------------------SNIP----------------------------------------------------------------------- 3. What is this about: ty:WARNING - Would have blown my audio buffer 4. If there is a way to keep the interlacing intact and still have files that is quicktime compatible that would be wonderful. Has anybody managed to do something like that? x264 is supposed to be able to encode interlaced material but i haven't figured out how to do it. Johan Arvidsson

4 6

mipsel-linux build of mencoder?
by Robert Prince 01 Jun '07

01 Jun '07

Hello folks, I've looked through the list archives into early last year, and all over with Google, but I can't seem to find the answer I'm after. I'm looking for a build of mencoder that will run on little-endian MIPS (mipsel), specifically linksys hardware, e.g. the various WRT models that can run OpenWRT. I've found an OpenWRT package of MPlayer without mencoder, but no mencoder so far. I've been trying to get a cross-compiler environment going, and while I've made progress there and I think I could have it solved, it could also take me days longer. Does anyone have, or know of, a build of mencoder that'll run on mipsel? Thanks, Robert

2 1

Audio out of sync if -oac is not lavc
by Timo Wiren 31 May '07

31 May '07

I'm trying to encode an avi and mp3 file together with libavcodec. If I use something else than lavc as audio codec, resulting video's audio is out of sync. This works and is in sync: mencoder trigun_clips.avi -ovc lavc -lavcopts vqscale=2:autoaspect:mbd=2:trell -oac lavc -audiofile 09_Big_Man_With_a_Gun.mp3 -o output.avi I don't want to use lavc for audio because it breaks seeking and doesn't work on my friends' Windows machines (choppy video in MPlayer, no audio in other players). I tried to use lame: mencoder trigun_clips.avi -ovc lavc -lavcopts vqscale=2:autoaspect:mbd=2:trell -oac mp3lame -lameopts cbr:br=192 -audiofile 09_Big_Man_With_a_Gun.mp3 -o output.avi Audio is not in sync with video. Any ideas how to fix this? Same problem occurs in -oac pcm and copy. MPlayer/MEncoder version is a subversion snapshot from 2007-05-27.

2 2

settings to convert ATSC HDTV 1080i30 to 720p24
by Justin Randall 31 May '07

31 May '07

Does anybody have a good approach to convert 1080i content to progressive? I'm looking for something that is robust and can work well with filmed or video content that is broadcast and captured as 1080i. I've seen a lot of mencoder -vf pullup,softskip to get 1080i ~30fps down to a progressive 24fps... won't that crate a lot of jitter on motion scenes unless the original source really was filmed at 24fps? For example I could see something like that working for "Law & Order" which is recorded on film, but not so much for "The Tonight Show" or something that was recorded live. Any thoughts? -- Justin Randall

6 12

spp, uspp, fspp, pp7??
by Nicolas Hesler 30 May '07

30 May '07

Nico Sabbi wrote: > Nicolas Hesler wrote: > > >> Could someone explain to me when to use either spp, uspp or fspp?? >> >> >> > > RTFM and don't hijack threads > > 1) Don't hijack threads (even your own) 2) With mencoder use pp7, with mplayer use pp7 if you have powerful cpu, otherwise use fspp=5/fspp=4/fspp=3 - depending on your cpu and video resolution. -- Vladimir I Read TFM. I apologize for accidentally hijacking my own thread (mailing list newb). This time I'll be more specific. I understand they are video filters, but I don't understand what they do exactly and in what situations they are useful/not useful (are they meant to offer better compression/better quality encoding/both??)... I haven't been able to find a comparison or command line example using these filters for creating DVD compliant MPEG-2. I was wondering if they offered any quality benefits for creating DVD compliant MPEG-2.

1 0

Is possible unvdpart video w/o reencoding
by Jiří Heryán 30 May '07

30 May '07

Hi, I was used to make my encodes with vdpart option enabled, but my new standalone DVD/MPEG-4 player does not play them. Is it possible unvdpart these encodes without reencoding from scratch? Thanks J.Heryan

1 0

Autosplit a grab at 1 or 2 GB boundaries?
by Levente Novák 29 May '07

29 May '07

Maybe I haven't searched deeply enough, but I am wondering whether it is possible to split a v4l2 grab (raw DV) into 1 GB or 2 GB chunks as separate files on-the-fly? I tried to omit the --enable-largefiles flag at compile time but MEncoder still writes a large monolithic file out of an 1 hour-long footage. This would not be a problem per se, but unfortunately audio and video gradually loses the sync, which becomes very annoying after 20-30 minutes. A possible workaround would be to grab only short clips at once (from an Hi8 source, so no problems with this) but it would be easier for me not to stop and start the camcorder manually so often. Levente PS. Is there a way to force MEncoder to grab audio and video in sync (frame by frame) even if the frame rate changes slightly (eg. sometimes 24.9 fps instead of 25 fps because of the CPU load), which I suspect being the source of the loss of synchronisation. I tried both -mc 0, -noskip and -vf harddup alone and in combination, none of them really helped. I don't mind if the resulting DV file is 24.9 fps of 25 fps (neither the video nor the audio speedup is noticeable when playing the file back at 25 fps) but the async is very annoying.

1 0

Copy h.264 video into avi
by aslan＠gmx.at 28 May '07

28 May '07

Hello. I have a (actually many) transport stream file that contain h.264 video and ac3 audio. It is hard to work with these files as there is hardly any software for it - and the software that exists isn't mature. I figured my best option to work (that is: edit) with such .ts files is to convert them to avi. There are many ways to do this, but none I found yet is really good. In essence it is possible to grab the original stream data and put it into an avi container, however this is a lengthy and error-prone process. I ended up not being able to get the ac3 back in properly, maybe due to the reason that the format changes between stereo and 5.1 within the same stream. So I thought that mencoder might help me here. With a command like mencoder "source.ts" -oac copy -ovc copy -of avi -ffourcc h264 -o "dest.avi" it would just copy the streams. The problem is that the resolution is left out and left at 0x0. It would even leave out the fourcc, but as you see I managed to find out about the -ffourcc option Is there a way to tell mencoder to set those? I could ofcourse just re-encode the video (finally the latest x264 builds manage to decode the video properly and without too many glitches, but I had to use an unofficual build of emncoder) but that takes ages for a full movie and reduces the quality. Hope you have tips for me! Morke

5 12