> Also that different information should possibly be optional > as in version 2 format, and with that the first line could specify > the version of the format
Given that this would be a backwards incompatible change (anything assuming every line is a stat line would break on the header), should I take "optional" here to mean "controlled by a flag"? How about I add: - A 'stats_version' option, integer, defaults to '1', which outputs the existing format as is, without a version header. - A second version '2', which outputs a version header, perhaps a list of included fields? - A 'log_raw' flag, which, when enabled, adds the raw error and max numbers for each channel and averaged over all channels, from which MSE and PSNR can be calculated. Alternatively version 2 could just include all the raw numbers. I'm concerned though that for large videos the logs could become quite large with the raw numbers included, which is why I think they should be optional. On 20 July 2016 at 17:10, Michael Niedermayer <mich...@niedermayer.cc> wrote: > On Tue, Jul 19, 2016 at 10:49:20AM -0700, Lucas Cooper wrote: > > This allows retroactive aggregation of the PSNR statistics (e.g. > > calculating average PSNR for entire video or a segment of the video). > > To do this accurate it would be better to list the actual integer > differences instead of float rounded to 2 digits afer the decimal > point. > Also that different information should possibly be optional > as in version 2 format, and with that the first line could specify > the version of the format > > [...] > -- > Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB > > What does censorship reveal? It reveals fear. -- Julian Assange > > _______________________________________________ > ffmpeg-devel mailing list > ffmpeg-devel@ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-devel > > _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
