On Thu, Oct 21, 2010 at 5:57 PM, Carl Karsten <[email protected]> wrote: > On Thu, Oct 21, 2010 at 7:46 PM, Carl Karsten <[email protected]> wrote: >> On Wed, Oct 13, 2010 at 12:20 PM, Dan Dennedy <[email protected]> wrote: >>> On Tue, Oct 12, 2010 at 10:38 PM, Carl Karsten <[email protected]> >>> wrote: >>>> This example shows something weird. in/out is in frames, but I think >>>> it is getting converted to time, and then converted back to frames, >>> >>> nah >>> >>>> and a different fps used for the 2 conversions. I haven't done the >>>> math yet, and my ntsc math is questionable anyway. >>>> >>>> Also not sure how the auto_profile thing you mentioned will effect >>>> this. It may hide it, but I don't think it will fix it. or something. >>>> >>>> melt -profile dv_ntsc -producer color:red out=1300 meta.attr.titles=1 >>>> meta.attr.titles.markup=#timecode# -attach data_show dynamic=1 >>>> -progress -consumer avformat:bar.dv pix_fmt=yuv411p >>> >>> There was a bug in the frame number to timecode converter with >>> non-integral frame rates. I just committed a fix. I also committed a >>> change to add support for #frame# >>> >>>> melt -profile dv_ntsc bar.dv in=1000 out=1000 >>>> # I see 00:00:34:14 >>> >>> Using #frame# this now shows 1000. >>> >>>> melt bar.dv in=1000 out=1000 >>>> # I see 00:00:41:10 >>> >>> And this obviously does not show 1000 but rather 1199. >>> >> >> Is 1199 what it should be now that you fixed it? >> >> it is what I am seeing now that I updated from >> deb http://ppa.launchpad.net/sunab/kdenlive-svn/ubuntu maverick main >> >> c...@dc10:~/temp/mltbug$ melt --version >> MLT melt 0.5.11 >> >> I have support for #frame# >> >> and it shows 1199 when I do >> $ melt bar.dv in=1000 out=1000 >> >> >>> Thanks for helping to locate a bug here. Good thing that timecode >>> converter was not used for anything critical with MLT's timing. >>> Instead, it was only used for this timecode burn-in filter. >>> -- >>> +-DRD-+ >>> >> >> -- >> Carl K >> > > > I just encoded the stuff I was working on that made me aware of a > problem, and it is fine now. > > But I am still confused by why I get 1199 and not 1000 in the above test.
Yay! You helped me find a bug in the new auto-profile code. If you had used "melt -profile dv_ntsc bar.dv in=1000 out=1000" with an explicit profile, then it would have shown 1000. Check sunab's repo periodically for when mlt is updated post Oct 21. -- +-DRD-+ ------------------------------------------------------------------------------ Nokia and AT&T present the 2010 Calling All Innovators-North America contest Create new apps & games for the Nokia N8 for consumers in U.S. and Canada $10 million total in prizes - $4M cash, 500 devices, nearly $6M in marketing Develop with Nokia Qt SDK, Web Runtime, or Java and Publish to Ovi Store http://p.sf.net/sfu/nokia-dev2dev _______________________________________________ Mlt-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/mlt-devel
