But FFmpeg is not accurate and it started the video from a nearby point instead (from 00:24:46~). I tried to add 2 seconds to my starting point and it took another frame (not what I wanted).
I'm having trouble to find the right ffmpeg options to encode a video that can be read on a htc G1 cell phone. I have used several codecs and formats but none is working.
I have followed these instruction to install ffmpeg and x264 [URL]
Here is my ffmpeg config :
Code: FFmpeg version SVN-r24953, Copyright (c) 2000-2010 the FFmpeg developers built on Aug 27 2010 22:44:01 with gcc 4.4.1 configuration: --enable-gpl --enable-version3 --enable-nonfree --enable-postproc --enable-libfaac --enable-libmp3lame --enable-libopencore-amrnb --enable-
Anybody had any success in getting ffmpeg to work as advertised with video capture from a webcam? I really want to convert the webcam output to VP8 or H264, but apparently ffmpeg can't even capture the webcam with a video4linux device.
I have some videos in an mkv container that are 1920 pixels wide, but less than 1080 pixels high. This causes problems when playing the videos on a PS3 (after converting to an AVCHD system), because the PS3 won't centre the video, leaving a very large black bar at the bottom but none at the top. Is there a way to use mencoder or ffmpeg to losslessly add padding to the top and bottom to make the video 1920x1080?
I spent about a half hour wrestling with different website tutorials about how to convert a file with ffmpeg and figuring out how to get all the video quality options right. Then I discovered you can just use the -sameq option and it figures it all out for you if you don't want to change the vid quality but just want it in another format. Thought I'd leave this on the site in case anyone else finds himself in the same boat.
I'm trying to write a bash script for gpodder to automatically convert video podcasts to play on my media player. I'm using ffmpeg for the conversions (compiled myself with all codecs enabled). I'd like to avoid resampling the video or audio whenever it's unnecessary but ffmpeg seems to want to resample my video even if I only give it audio parameters to change.For example I have a test video with the following parameters:
I am told I need the output file to comply with this
Video Resolution: 480*320 Video Bitrate: 768 kbps Audio Bitrate: 128 kbps Video Format: MPEG-4 (be sure not to use H264, as it�s not supported in the current firmware)
i am using Ubuntu 11.04 on my computer system. I urgently need a good video converter for converting videos.I have already installed FFmpeg and men-coder,Winff etc. The problem is each has its own drawback.For instance ffmpeg cannot convert a .avi to .3gp with audio working. My preferences are the converter should be user friendly, should support all popular video formats.
I am having problems with ffmpeg. My goal is to capture a video stream from my webcam and feed that into a webcam-capturing program. But to get that to work, I will need ffmpeg to work. I need the following command to work, but I get an error:
Code: $ ffmpeg -b 100K -an -f video4linux2 -s 320x240 -r 10 -i $device -b 100K -f image2pipe -vcodec mjpeg - | perl -pi -e 's/\xFF\xD8/KIRSLESEP\xFF\xD8/ig' ffmpeg: relocation error: /usr/lib/libavfilter.so.2: symbol avformat_find_stream_info, version LIBAVFORMAT_53 not defined in file libavformat.so.53 with link time reference
I've got a 1920x1080 video I've edited and rendered with Cinelerra, and I'm trying to use ffmpeg to transcode it to something smaller. However, when I use a command like this, for instance:
Code:
I inevitably get some weird green band at the bottom of the frame in the converted video. I know that there's some weird pixel stretching going on here, because the NTSC standard for 16:9 is 720x480 with rectangular pixels, and the 1080 version has square pixels, so I'm guessing the green band is an artifact of that process?
I am using ffmpeg for merge wav files to a mov video. My doing is below
1. First extract audio (wav file) from video 2. Create wav file from mp3 track 1 3. Create wav file from mp3 track 2 4 Merge extract audio from video with track 1 and track2. Now finally create a new video with original video's video stream and merged audio stream.
Process is working. However final video is 3-4 times greater than original one. I want that final video should be near about size of original video. As I understand, all three wav files (created from ) make video larger.
I have some video clip and I want to add a title in its buttom center. For exemple, My clip is video.mp4 and I want to add the title "hello world".Does ffmpeg (or mencoder) enables me to do this?
I've created some time-lapse videos from photos, using this command: ffmpeg -i IMG_%03d.JPG -s 1440x1080 -sameq video.MP4
And it worked great. Now I want to join several of these time-lapse videos to make a single, longer video (all the input videos have the exactly same format). I already tried using: cat video1.MP4 video2.MP4 > stitch.MP4
but the output ends up being equal to video1.MP4, I don't want to transcode nor changing any parameter of the video, I just want a end-to-end stitching, as if those videos were on a playlist.
I'm trying to find out how to add watermark to a video using a ffmpeg via console, because I have to do it quick. There is tutorial on ..... (other forums use also this video tutorial) for this but there is a small problem - in ffmpeg 6.1 there is no more vhook subsystem which is used in tutorial...
I'm trying to encode a wmv file to flv with ffmpeg. The video codec is WMV3 and the audio codec is wmap (Windows Media Audio Professional). The command I use is:
As you can probably see the audio codec is not supported. Is there a way to encode WMV3+wmap with ffmpeg or any other tool in Linux? Windows Media Encoder in Windows is able to encode such files to a supported codec. For example: WMV3+WMA2/WMV2+WMA2, I could then encode it in Linux. I'm trying to find a way to directly encode WMV3+wmap in Linux.
i am using ubuntu 8.04.I wanted to clip from a movie so I ran
Code: mencoder -ss 2453 -endpos 34 InpuTFile -acodec copy -vcodec copy -o Output I changed the ss value still it cut from same frame of the video.It seems mencoder works faster than ffmpeg but jumps to keyframe hence cannot jump to accurately specified second.
I have an Asus EEE PC900. Just installed 10.04 netbook remix and everything works fine straight out of the box. it works great and is a vast improvement on windows xp which was previously installed on the asus. Just one small thing - the battery meter is never accurate as far as time remaining goes - it currently says i have 19 hours 55 minutes to go - unfortunately this is not quite true i suspect. It was the same with 9.10, Does anyone have a solution for this?
I've got a bunch of HD files in .mkv format that I want to convert to a format that the 360 will read (currently just trying .avi files) so I can stream the media to my xbox360. I'd rather avoid transcoding the entire file, so I tried unpacking the .mkv file into audio and video (and subtitle if necessary) and repacking said file into avi using ffmpeg. The repacking is where I am having the problem. Here's the commands I tried at first (it might look familiar):
Error while opening codec for output stream #0.1 - maybe incorrect parameters such as bit_rate, rate, width or height Where stream #0.1 is the output audio stream to the .avi file. I think I have narrowed it down to the fact that the input streams are all 6 channel (5.1 surround sound) which mp2 doesn't support (the ffmpeg audio default) so I did this instead:
to change the sound to stereo it works. Problem is I'd like to keep my media in surround sound. Any thoughts on how to get all 6 channels in the new mp3 file? I'm fairly certain that mp3s support 5.1. And after some messing around, I haven't found that my ffmpeg can transcode audio format from 5.1 back to 5.1. So since (I think) the 360 can play .ac3 files I decided to just copy both the video and audio file to the .avi file sans transcoding using
however now about halfway through, I run into this error:
Code:
NULL @ 0x22f3a00]error, non monotone timestamps 62562 >= 62562
av_interleaved_write_frame(): Error while opening file I can't seem to find any solution to this problem anywhere online, does anyone know how to get around this? I'd prefer to use ffmpeg for this if possible. (so I can write a script and just attack an entire directory at once) For what it's worth, I'm using the latest ffmpeg in the ubuntu (karmic) default repositories.