On 6/25/2016 4:52 PM, Martin Vignali wrote: > 2016-06-25 21:47 GMT+02:00 James Almer <jamr...@gmail.com>: > >> On 6/25/2016 4:43 PM, Martin Vignali wrote: >>> Hello, >>> >>> In attach a patch to add a fate test for WebP lossless rgb >>> >>> Sample can be found here : https://we.tl/dmPsitbwXK >>> >>> and need to be put inside ./fate-suite/webp (the webp folder doesn't >> exists) >>> >>> to test : >>> make fate-webp SAMPLES=fate-suite/ >>> >>> >>> I doesn't make fate test, for lossy rgb/rgba, because the actual default >>> output of ffmpeg >>> have a lower image quality than the official decoder (google dwebp) on my >>> samples >>> >>> for lossless rgba, i doesn't make fate test, because the alpha value >> seems >>> to not be the right value >> >> This is a bug and should be addressed, then. Lossless should mean lossless. >> Can you fix it, or open a ticket on trac with a testcase? >> >> > Sorry, in fact rgba lossless is ok (my mistake) > > i will open a ticket on trac for lossy
Not sure how it is with VP8/Webp, but unless the output is defined like with hevc and VP9, different decoders may give different results, so it may not be a bug. You should still make a test for it anyway. It will reflect changes and improvements to the decoder if they are added in the future and help detect regressions. And it can be in a separate patch than the rgb lossless ones. > > and i will make another patch for fate for rgb and rgba lossless webP > > Martin > _______________________________________________ > 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