Re: NHW Image codec - improvement of precision
Raphael Canut <[email protected]> Thu, 16 May 2013 18:20:22 +0200
| Newsgroups | gmane.comp.multimedia.ogg.theora.general |
|---|---|
| Message-ID | <CAKE58qEu16MwF_Se6iJDZ4Z7yBZvkQwMZZfVew2ESdBAAyo9SQ@mail.gmail.com> |
Hello, Just a quick message to let you know that I have improved the compression scheme of UV DC part, of 200octet in average... And so I use this space to better select the wavelet coefficients in the higher bands (for Y comp).This new version has then a little more precision. Many thanks, Raphael 2013/4/8 Raphael Canut <[email protected]> > Hello, > > Just a quick message, I have improved precision on U and V components, and > so this version is better if it may be of interest.Yes, for now the NHW > codec is just an experimental test tool, that would be optimized for visual > evaluation (more neatness-sharpness) rather than PSNR. > > If you could find time to test it, would be really interested in any > feedback (about for example the overall approach, results,...). > > Many thanks again, > Raphael > > > 2013/3/25 Raphael Canut <[email protected]> > >> Hello, >> >> Just a very quick message, for those interested, I improved a >> little precision of my codec.This version also has a better compression, >> especially for some images.This version seems a little better.Any comment >> is welcome. >> >> Many thanks, >> Raphael >> > > _______________________________________________ theora mailing list [email protected] http://lists.xiph.org/mailman/listinfo/theora