November 2008 - 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

B-frames and GOP with MPEG-2 for DVD
by Nicolas Hesler 16 Apr '11

16 Apr '11

Can anyone offer any input to the following command line? I've been trying to refine for the highest possible quality of MPEG-2 while staying DVD compliant. The source is 24fps. (Note: format=mpeg2, because format=dvd does not work with Adobe Encore 2.0). From what I understand, keeping GOP open, improves quality. Is this true? Is it 100% DVD compliant? Do I want to introduce B frames in my encoding? (The vb_strategy part of my command line is currently useless, right?) If parts of my video are reaching into 9-10 Mbs bitrate, would it beneficial to introduce B-Frames? Will it improve quality? (This is probably a stupid question) Are the sync issues with bframes only relevant to mpeg4 or lavf? mencoder -oac lavc -ovc lavc -of mpeg -mpegopts format=mpeg2:tsaf:telecine / -vf scale=720:480,harddup -mc 0 -noskip -srate 48000 -af lavcresample=48000:0:2 -lavcopts / vcodec=mpeg2video:vrc_buf_size=1835:vrc_maxrate=7000:vbitrate=6000:keyint=12 / :trell:mbd=2:precmp=2:subcmp=2:cmp=2:dia=-10:predia=-10:cbp:mv0:vqmin=1:lmin=1 / :dc=10:vpass=1:vb_strategy=2:acodec=ac3:abitrate=224:aspect=16/9:vstrict=0 -ofps 24000/1001 -o $output $input

12 43

2 pass encoding failed witg VP6 codec
by guillaume montard 26 Mar '09

26 Mar '09

Hi, I m trying to encode a .mov video to flv using the great VP6 codec. I m able to encode it in 1-pass, but i would like to make a 2-pass. I googled for like day and i m still stuck.... Here is my code line : mencoder testmencoder.mov -o test.flv -ovc vfw -xvfwopts codec=vp6vfw.dll:compdata=1passvodeclic.mcf -nosound -vf yadif,scale,flip -of lavf mencoder testmencoder.mov -o test.flv -ovc vfw -xvfwopts codec=vp6vfw.dll:compdata=2passvodeclic.mcf -nosound -vf yadif,scale,flip -of lavf My problem is that mencoder fail doing the 2 pass. Mencoder simply stop doing nothing : VDec: vo config request - 1024 x 768 (preferred colorspace: Planar YV12) VDec: using Planar YV12 as output csp (no 0) Opening video filter: [flip] Movie-Aspect is undefined - no prescaling applied. [swscaler @ 0x6622f0]SwScaler: using unscaled yuv420p -> bgr24 special converter Opening video filter: [expand] Expand: -1 x -1, -1 ; -1, osd: 0, aspect: 0.000000, round: 1 Starting compression: Input format: biSize 40 biWidth 1024 biHeight 768 biPlanes 1 biBitCount 24 biCompression 0x0 ('') biSizeImage 2359296 Output format: biSize 40 biWidth 1024 biHeight 768 biPlanes 1 biBitCount 24 biCompression 0x32365056 ('VP62') biSizeImage 1576960 Does any one have an hint ? Thanks -- Guillaume MONTARD guillaume.montard(a)gmail.com

2 1

pullup useless?
by Daniel Gnoutcheff 30 Dec '08

30 Dec '08

Hello all, OK, I'm working on making a DVD backup of a VHS tape of an old animated movie. I've already done the capture, stored in a huge .avi, video compressed with the lossless ffvhuff with the hope of making the thing fit on my HD without losing any information. I'm in NTSC country, and the movie is obviously telecined via 2:3 pulldown. I want to do an inverse pulldown on it. Now, the detc filter works - but not perfectly. It seems to miss things when there is only a little bit of isolated movement. So I'd like to use one of the reputedly better pulldown and filmdint. But there's a slight problem. pulldown does nothing. At all. mencoder runs, but the video comes out unchanged and still telecined. Now, I might put this down to noise - the VHS is crappy, the VCR is crappy, and the TV capture card is crappy - but there are some scenes in the source video (e.g. during pans) where the telecining is *really* obvious. But to my horror, pullup doesn't get a clue. I've tried adjusting pullup's parameters to help - I made it ignore some junk on the top and bottom, and I adjusted the sc paramiter. My latest attempt: mencoder sw7d-telcine.avi -vf pullup=2:2:4:4:-1:0,softskip -oac copy -ovc copy -o test.avi sw7d-telcine.avi is a sort clip extracted from the big capture that shows a pan, with very obvious telecine. I understand that this command has pullup ignore 16 pixels on the left, 16 on the right, 8 on the top, and 8 on the bottom. The resulting test.avi is indistinguishable from sw7d-telcine.avi, telecine and all. Any ideas? How should I make pullup deal with noise? How to I give this thing a clue, especially given the detc handles the same clip beautifully? Thanks all, Daniel

8 15

mencoder two pass trouble on linux
by Richard Ji 17 Dec '08

17 Dec '08

Hi all, I am using mencoder to create avi file from DVD. I use the same commands on both Windows and Linux (OpenSUSE). The following worked fine on Windows. However, on my Linux box, the second pass fails with: Opening video decoder: [mpegpes] MPEG 1/2 Video passthrough VDec: vo config request - 720 x 480 (preferred colorspace: Mpeg PES) Could not find matching colorspace - retrying with -vf scale... Opening video filter: [scale] The selected video_out device is incompatible with this codec. Try appending the scale filter to your filter list, e.g. -vf spp,scale instead of -vf spp. VDecoder init failed :( Opening video decoder: [libmpeg2] MPEG 1/2 Video decoder libmpeg2-v0.4.0b Selected video codec: [mpeg12] vfm: libmpeg2 (MPEG-1 or 2 (libmpeg2)) However, mplayer plays the DVD fine under Linux. Here is the commands I used: mencoder dvd://$1 -dvd-device /dev/cdrom -aid 128 -ovc lavc -lavcopts vcodec=mpeg4:vpass=1 -oac copy -o temp.avi mencoder dvd://$1 -dvd-device /dev/cdrom -aid 128 -ovc lavc -lavcopts vcodec=mpeg4:mbd=2:trell:vpass=2 -oac mp3lame -lameopts br=128:vol=9 -o $2 Any helps is appreciated. Thanks Rich

5 9

Cannot go above 640x480 with v4l2 webcam capture
by Kovacs Peter Tamas 09 Dec '08

09 Dec '08

Dear All, I'm trying to capture a video stream from a Logitech Quickcam Pro for Notebooks webcam. I'm using the latest uvcvideo kernel module (from SVN), latest libwebcam (from SVN), and mencoder (from SVN). If I try to capture a video stream with resolution 640x480 or below, everything is working fine, eg: mencoder tv:// -tv driver=v4l2:width=640:height=480:device=/dev/video0 -nosound -ovc lavc -o capture.avi However, when I try to go further, it always gives me a 640x480 stream, like: -------------------------------------- pc05 ~# mencoder tv:// -tv driver=v4l2:width=800:height=600:device=/dev/video0 -nosound -ovc lavc -o capture.avi MEncoder SVN-r27899 (C) 2000-2008 MPlayer Team CPU: AMD Athlon(tm) 64 X2 Dual Core Processor 6400+ (Family: 15, Model: 67, Stepping: 3) 3DNow supported but disabled 3DNowExt supported but disabled CPUflags: Type: 15 MMX: 1 MMX2: 1 3DNow: 0 3DNow2: 0 SSE: 1 SSE2: 1 Compiled for x86 CPU with extensions: MMX MMX2 SSE SSE2 success: format: 9 data: 0x0 - 0x0 TV file format detected. Selected driver: v4l2 name: Video 4 Linux 2 input author: Martin Olschewski <olschewski(a)zpr.uni-koeln.de> comment: first try, more to come ;-) v4l2: your device driver does not support VIDIOC_G_STD ioctl, VIDIOC_G_PARM was used instead. Selected device: UVC Camera (046d:0991) Capabilites: video capture streaming supported norms: inputs: 0 = Camera 1; Current input: 0 Current format: YUYV v4l2: ioctl set format failed: Invalid argument v4l2: ioctl set format failed: Invalid argument v4l2: ioctl set format failed: Invalid argument tv.c: norm_from_string(pal): Bogus norm parameter, setting default. v4l2: ioctl enum norm failed: Invalid argument Error: Cannot set norm! Selected input hasn't got a tuner! v4l2: ioctl set mute failed: Invalid argument v4l2: ioctl query control failed: Invalid argument [V] filefmt:9 fourcc:0x32595559 size:640x480 fps:15.000 ftime:=0.0667 Opening video filter: [expand osd=1] Expand: -1 x -1, -1 ; -1, osd: 1, aspect: 0.000000, round: 1 ========================================================================== Opening video decoder: [raw] RAW Uncompressed Video VDec: vo config request - 640 x 480 (preferred colorspace: Packed YUY2) 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 undefined - no prescaling applied. SwScaler: reducing / aligning filtersize 1 -> 4 SwScaler: reducing / aligning filtersize 1 -> 4 SwScaler: reducing / aligning filtersize 1 -> 1 SwScaler: reducing / aligning filtersize 9 -> 8 [swscaler @ 0x1104800]BICUBIC scaler, from yuyv422 to yuv420p using MMX2 [swscaler @ 0x1104800]using 4-tap MMX scaler for horizontal luminance scaling [swscaler @ 0x1104800]using 4-tap MMX scaler for horizontal chrominance scaling [swscaler @ 0x1104800]using 1-tap MMX "scaler" for vertical scaling (YV12 like) [swscaler @ 0x1104800]640x480 -> 640x480 videocodec: libavcodec (640x480 fourcc=34504d46 [FMP4]) Selected video codec: [rawyuy2] vfm: raw (RAW YUY2) ========================================================================== Forcing audio preload to 0, max pts correction to 0. New_Face failed. Maybe the font path is wrong. Please supply the text font file (~/.mplayer/subfont.ttf). subtitle font: load_sub_face failed. New_Face failed. Maybe the font path is wrong. Please supply the text font file (~/.mplayer/subfont.ttf). subtitle font: load_sub_face failed. Writing header... ODML: Aspect information not (yet?) available or unspecified, not writing vprp header. Writing header... ODML: Aspect information not (yet?) available or unspecified, not writing vprp header. Pos: 2.3s 35f ( 0%) 13.01fps Trem: 0min 0mb A-V:0.000 [543:0] Flushing video frames. Writing index... Writing header... ODML: Aspect information not (yet?) available or unspecified, not writing vprp header. Video stream: 543.651 kbit/s (67956 B/s) size: 158565 bytes 2.333 secs 35 frames v4l2: ioctl set mute failed: Invalid argument v4l2: 37 frames successfully processed, -36 frames dropped. -------------------------------------------- >From this log, it doesn't look like it tried to go for 800x600. However, when I insert the uvcvideo kernel module, it can be seen that it supports this resolution (in YUV and MJPEG too): -------------------------------------------------------- Nov 20 01:08:11 pc05 kernel: uvcvideo: UVC device initialized. Nov 20 01:08:11 pc05 kernel: uvcvideo: Probing generic UVC device 4 Nov 20 01:08:11 pc05 kernel: uvcvideo: Found format MJPEG. Nov 20 01:08:11 pc05 kernel: uvcvideo: - 160x120 (30.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 176x144 (30.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 320x240 (15.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 352x288 (15.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 640x480 (15.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 800x600 (15.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 960x720 (10.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: Found format YUV 4:2:2 (YUYV). Nov 20 01:08:11 pc05 kernel: uvcvideo: - 160x120 (30.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 176x144 (30.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 320x240 (15.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 352x288 (15.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 640x480 (15.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 800x600 (15.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 960x720 (10.0 fps) Nov 20 01:08:11 pc05 kernel: uvcvideo: - 1600x1200 (5.0 fps) --------------------------------------------- I've also tried using outfmt:mjpeg, and it works, but still cannot get a higher resolution video stream. Using guvcview, I can use the camera in 800x600 and even 960x720, both in YUV and MJPEG. I'm running a stripped down Fedora 8 on x86_64 with 2.6.24 kernel: pc05 ~# uname -a Linux pc05 2.6.24.5-fedora #7 SMP Fri Nov 7 13:42:49 CET 2008 x86_64 x86_64 x86_64 GNU/Linux Do you have an idea what could be the reason for this behaviour? Thank you very much for your kind help in advance, Peter

2 6

encoded ASF out of AV sync
by DimitryASuplatov 30 Nov '08

30 Nov '08

Hello, I have grabbed and encoded a streaming video with the following command mencoder $1 -o $2.avi -oac mp3lame -lameopts vbr=2:aq=3:q=3:abr:br=192 -ovc lavc -ffourcc DX50 -lavcopts mbd=0:vbitrate=1800:aspect=16/9 -vf harddup=1 -channels 2 -srate 32000 -ofps 29.97 -vf scale=720:404 I got a file with 0,5 seconds desync so I reencoded it with mencoder $2.avi -o $3.avi -ovc copy -oac copy -delay 0.5 Then I got a perfectly good file when playing with mplayer though all other players (vlc, windows media and my pioneer dvd player) refused to recognize the delay header. I`ve read in the internet that -mc 0 option could be helpful when encoding in the first place. But I do not want to redownload the movie. 1/ Is there any way to cut 0.5 seconds from the start of the audio track? 2/ What options should I`ve used in the first place to avoid such a problem? Thank you for your answers. I appreciate your time. SDA

1 0

sstep with cropdetect
by Micah Chambers 28 Nov '08

28 Nov '08

For some reason I can't get cropdetect to work with sstep. I found a previous message where someone had an issue, but that wasn't really resolved. I am using the latest version of mplayer in svn: here is my configuration flags: Install prefix: ~/bin/mplayer Data directory: ~/bin/mplayer/share/mplayer Config direct.: ~/bin/mplayer/etc/mplayer Byte order: little-endian Optimizing for: Languages: Messages/GUI: en Manual pages: en Enabled optional drivers: Input: ftp pvr tv-teletext tv-v4l2 tv-v4l tv cddb cdda libdvdcss(internal) dvdread(internal) vcd dvb smb network Codecs: x264 xvid libavcodec real xanim faad2 faac libmpeg2 liba52 mp3lib libtheora tremor(internal) libmad liblzo gif Audio output: alsa oss v4l2 sdl mpegpes(dvb) Video output: v4l2 sdl gif89a pnm jpeg png mpegpes(dvb) fbdev aa xvidix cvidix opengl dga xv x11 xover md5sum tga Disabled optional drivers: Input: dvdnav vstream radio tv-dshow live555 nemesi Codecs: libdv libamr_wb libamr_nb qtx win32 musepack libdca speex toolame twolame Audio output: sun openal jack pulse esd arts ivtv dxr2 nas Video output: zr zr2 ivtv dxr3 dxr2 vesa svga caca ggi xmga mga winvidix 3dfx xvmc dfbmga directfb bl xvr100 tdfx_vid s3fb tdfxfb so here is what happens: $ ~/mplayer/mplayer -vf cropdetect -vo null -nosound -sstep 100 -frames 10 Judge\ Dredd.vob MPlayer dev-SVN-r26916-4.2.4 (C) 2000-2008 MPlayer Team CPU: Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz (Family: 6, Model: 15, Stepping: 11) CPUflags: MMX: 1 MMX2: 1 3DNow: 0 3DNow2: 0 SSE: 1 SSE2: 1 Compiled for x86 CPU with extensions: MMX MMX2 SSE SSE2 Cannot load bitmap font: /home/osmo/.fonts/microsoft-vista/calibri.ttf Playing Judge Dredd.vob. MPEG-PS file format detected. VIDEO: MPEG2 720x480 (aspect 2) 29.970 fps 9800.0 kbps (1225.0 kbyte/s) dvdsublang...judge dredd fi fin en eng dvdsublang...judge dredd fi fin en eng Opening video filter: [cropdetect] ========================================================================== Opening video decoder: [mpegpes] MPEG 1/2 Video passthrough VDec: vo config request - 720 x 480 (preferred colorspace: Mpeg PES) Could not find matching colorspace - retrying with -vf scale... Opening video filter: [scale] Opening video filter: [lavc] Could not find matching colorspace - retrying with -vf scale... Opening video filter: [scale] The selected video_out device is incompatible with this codec. Try appending the scale filter to your filter list, e.g. -vf spp,scale instead of -vf spp. VDecoder init failed :( Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family Selected video codec: [ffmpeg2] vfm: ffmpeg (FFmpeg MPEG-2) ========================================================================== Audio: no sound Starting playback... VDec: vo config request - 720 x 480 (preferred colorspace: Planar YV12) VDec: using Planar YV12 as output csp (no 0) Movie-Aspect is 1.33:1 - prescaling to correct movie aspect. [mpeg1video @ 0xb50d60]removing common factors from framerate VO: [null] 720x480 => 720x540 Mpeg PES V: 208.9 9/ 7 ??% ??% ??,?% 0 0 demux_mpg: 24000/1001fps progressive NTSC content detected, switching framerate. V: 0.0 51/ 48 ??% ??% ??,?% 0 0 Could it be a compile flag I have left out or something? Thanks

3 2

Is there a best site for Mencoder binaries for windows?
by Peter 26 Nov '08

26 Nov '08

I realise that I'm going to get lambasted for wanting binaries, but realistically I'm not up to getting the latest "SVN". I'd like a copy of mencoder that is sufficiently up to date that I can use the "threads" parameter of xvidenc. Does anyone have any suggestions (other than compiling from source)? Actually, if there's a step-by-step to compiling, I'll consider that, too, it's just that in my experience, these things explode into complexity where each step (such as "get minGW") turns out to need several sub-steps (and so on...). -- Peter

2 1

mencoder audio is faster than video
by Emmanuel Guarda 26 Nov '08

26 Nov '08

i'm triing to encode tv programs with a bt878 based card that works prefectly with tvtime. te code that use to encode is: *mencoder tv:// -tv forceaudio:alsa:driver=v4l2:input=0:norm=pal-nc:chanlist=us-cable:channel=$CHAN:width=640:height=480 -ovc lavc -lavcopts vcodec=mpeg4:vbitrate=3000 -ofps 25 -oac mp3lame -lavcopts acodec=mp3:abitrate=128 -srate 48000 -noskip -mc 0 -vf scale=640:480,pp=hb/vb/dr/al/lb -endpos $DURATION -o "$FNAME"_`date +%m%d`.avi -quiet* the image and audio hears perfect, but when i record a long time, the audio is before the video. the option *-noskip -mc 0 *i used because if i not, the video swich between faster a slower speed, and whit this two options record fine (only whith the two. only one don't work) i try to change srate to 44100 and don't work. Changing mp3lame to lavc don't work I use gentoo linux 32bits in a E2180 @ 3200Mhz - 2gb ddr2 - msi p35 neof - and 2 sata2 so the pc is so faster to do this without problems. sory for my poor english, I speak spanish and thanks EMMA

2 2