Can someone give me a hand? Please do a fresh pull of x264 and ffmpeg, then try to encode any video with ffmpeg using libx264 and verify that you get a segmentation fault. It seems there is a new bug either in ffmpeg or x264. If you search google, there is 1 or 2 threads about this on the ffmpeg mailing list from August 2008. In one of the messages the ffmpeg dev says it 'probably' has something to do with x264.
Can someone also tell me the last SVN of ffmpeg that they successfully got working with x264? I temporairily went back to 13060 which was the last version I was using prior to rebuilding x264 and ffmpeg last night, but I'm sure there's probably a later version that will still work. Thanks.
What is the command that you used to get the segmentation fault? A patch has been released 6 1/2 hours ago regarding a rc_eq default value bug: [PATCH] Fix crash in libx264 when rc_eq is not set. It will probably be applied to SVN, but I don't know how long that will take.
So I downloaded 14787 and compilied, and it doesn't have the seg fault problem. However, this version has a different issue. I use the following command line:
The 14787 version insists on only encoding at 150 kbits/s, if I remove the "-maxrate xxxx" then it encodes at the specified bitrate. Was this yet another change in ffmpeg? Version 13060 doesn't have any problems with the above command line.Code:ffmpeg -threads 2 -i INPUT_FILE -vcodec libx264 -s 320x240 -aspect 4:3 -flags +loop -cmp +chroma -deblockalpha 0 -deblockbeta 0 -b 800k -maxrate 1250k -bufsize 4M -bt 256k -refs 1 -coder 0 -me umh -me_range 16 -subq 7 -partitions +parti4x4+parti8x8+partp8x8 -g 250 -keyint_min 25 -level 30 -qmin 10 -qmax 51 -qcomp 0.6 -trellis 2 -sc_threshold 40 -i_qfactor 0.71 -acodec libfaac -ac 2 -ab 128k OUTPUT_FILE
I updated to 14872 and the latest x264. Your command seg faulted while converting a dv avi, but it worked when I added -rc_eq 'blurCplx^(1-qComp)' and it encoded at the proper bitrate. I don't know if that's helpful since I updated to a different revision.
This is the rate control equation for x264. I don't understand what the equation itself means, but there have been several ffmpeg bugs lately if you include it (5 weeks ago) or don't include it (this bug). The equation is the recommended default for x264, but for some reason ffmpeg doesn't use it by default so it has to be declared.
Ok, well that all kind of makes sense now regarding my original problem with the seg fault. At least I'm up and running with a new version of ffmpeg. Thanks for the help.
What command would output the highest quality x264/aac mp4 possible in a single pass encode? file size isnt too a big issue to me, but if you could explain what each variable did would be nice also
Bookmarks