2017-08-31 21:42 GMT+02:00 Michael Niedermayer <mich...@niedermayer.cc>:
> On Wed, Aug 30, 2017 at 03:54:08AM +0200, Thomas Mundt wrote: > > Hi, > > > > we did a transcoding cascade test at work were over-sharpening became > > visible with the complex low-pass filter. This patch rectifies the > > behaviour. > > > > Please comment... > > did you perform subjective and or objective tests ? > > objective being some metric like PSNR vs correctly sampled data > subjective being double blind tests with humans about which they prefer > The tests have been subjective with a team of video engineers and technicians with several test files. I did a quick SSIM/PSNR check with the first generation of an HD->SD file as a reference against the 6th generation. ffmpeg -i test.mxf -i ref.mxf -lavfi "ssim;[0:v][1:v]psnr" -f null - 2>test.txt Results without the patch: [Parsed_ssim_0 @ 041bef00] SSIM Y:0.956508 (13.615881) U:0.991601 (20.757750) V:0.993004 (21.551382) All:0.974405 (15.918463) [Parsed_psnr_1 @ 03fc8440] PSNR y:31.838009 u:48.424280 v:48.962711 average:34.759466 min:31.699297 max:40.857847 Results with the patch: [Parsed_ssim_0 @ 0424ed00] SSIM Y:0.970051 (15.236232) U:0.991883 (20.905857) V:0.993174 (21.658049) All:0.981290 (17.279202) [Parsed_psnr_1 @ 0424f780] PSNR y:34.412108 u:48.504454 v:48.969496 average:37.264644 min:34.310637 max:42.373392 _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
