[FFmpeg-user] fps in March 13 git: -r 29.97 = 2997 tbn ???
seandarcy2 at gmail.com
Sat Apr 9 07:21:51 CEST 2011
On 04/05/2011 12:20 PM, Baptiste Coudurier wrote:
> On 04/05/2011 06:33 AM, Mike Scheutzow wrote:
>> Baptiste Coudurier wrote:
>>>> Also, be careful believing the tbn / tbc "Output #0" stats - they have
>>>> been broken for months. Use 'ffmpeg -i infile' to get accurate values.
>>> No, the values are correct and accurate, you can trust them.
>> It was still broken in Michael's git as of 2011-mar-16.
>> The test case is very simple: h264 in .ts transcoded to h264 in .ts.
>> During encode run, for "Output #0" ffmpeg prints:
>> Stream #0.0: Video: libx264, yuv420p, 720x480 [PAR 10:11 DAR 15:11],
>> q=10-51, 200 kb/s, 90k tbn, 29.97 tbc
>> Yet, ffmpeg -i on the resulting file says:
>> Stream #0.0[0x100]: Video: h264 (High), yuv420p, 720x480 [PAR 10:11
>> DAR 15:11], 29.97 fps, 29.97 tbr, 90k tbn, 59.94 tbc
>> Do the tbc look the same to you?
> And which one do you think are wrong anyway ?
> Information printed is accurate, the interpretation is different.
How is the interpretation different? If it were interlaced, I guess one
would be fields, the other frames. But it's not interlaced. And which of
these fields does "-r" affect?
More information about the ffmpeg-user