Re: NHW Project - some update
Raphael Canut <[email protected]> Tue, 10 Jul 2018 11:50:37 +0200
| Newsgroups | gmane.comp.multimedia.ogg.theora.general |
|---|---|
| Message-ID | <CAKE58qFh9-chY7x5+wYbZ3G_LjX=0TcagyG_oAkgDjCLgeNsZA@mail.gmail.com> |
Hello Dave, Thank you very much for your help! Cheers, Raphael 2018-07-10 9:39 GMT+02:00 Dave Johnson Games <[email protected]>: > Let me do some looking. > > On Mon, Jul 9, 2018, 4:55 PM Raphael Canut <[email protected]> wrote: > >> Hello Dave! >> >> Thank you very much for your answer! >> >> You're right.I have thought that the new next-generation codecs like JVET >> (MPEG) VVC, AOM AV1, XVC will be too computational-intensive for mobile >> phones.So the de facto codecs for mobiles devices are H.264 and HEVC, I >> thought that my codec that is royalty-free and a lot faster could move the >> lines, but you're right Apple is "attached" to MPEG and Google and the >> others are "attached" to AOM. >> >> > Look to Asia >> >> I am always very attracted by new opportunities for my NHW Project! But >> don't you think there will be a problem of compatibility? For example a >> brand wants to use the NHW Project in its new mobile phone, but what will >> happen if the NHW Project is not supported by web browsers? >> >> Anyway, I am very interested to look to Asia, would you have some advice? >> >> Cheers, >> Raphael >> >> >> >> 2018-07-09 23:33 GMT+02:00 Dave Johnson Games <[email protected]>: >> >>> I think Apple is somehow "attached" to the MPEG group. They control the >>> "pools" too. Your codec sounds great. Google could -- if they wanted to -- >>> use it but they have Web^M and webP. Look to Asia. Asus has just launched a >>> new phone. ?? >>> >>> On Mon, Jul 9, 2018, 3:14 PM Raphael Canut <[email protected]> wrote: >>> >>>> Hello, >>>> >>>> I don't have advanced on the NHW Project this past month, but I have >>>> definitely validated its characteristics and this new version. >>>> >>>> The NHW Codec preserves or enhances image neatness and sharpness, which >>>> translates that on good quality images the NHW Codec is globally better >>>> than HEVC, and on degraded (with blur, artifacts) images it is worse than >>>> HEVC because it lacks of precision. I have made tests on 100 images from >>>> the Internet. >>>> >>>> The NHW Project still lacks very high compression, and I think now I >>>> will inspire from HEVC, that is to say, I would like to strongly remove >>>> details and preserve the contours, edges. It is not that easy to code, and >>>> again any help would be very very welcome! Again if you can provide any >>>> help to the NHW Project, do not hesitate to show up, would be so much >>>> appreciated!!! >>>> >>>> The NHW codec has a high speed, which makes it a good candidate for >>>> mobile, embedded devices.There are currently some exploration for a >>>> MotionJPEG replacement on low power processor (dual-core 240MHz Xtensa LX6) >>>> where MotionJPEG appears to be too heavyweight. >>>> >>>> With the recent updates, I think the NHW Project is still patent-free >>>> and royalty-free. And just the fact that it is royalty-free, should be >>>> worth the try. >>>> >>>> So with this new version, I hope the NHW Project could attract a >>>> company that would like to sponsorize the project (to adapt the codec to >>>> any size of image for example). >>>> >>>> Very quickly to finish, I have seen that on the Theora mailing list >>>> there are employees from big companies.This can be indiscreet and very >>>> impolite, but for example there are 2 employees from Apple who are on the >>>> Theora channel mailing list... Does Apple can be interested in such a codec >>>> as the NHW Project? If not, what does it lack to be of interest? Would you >>>> have some advice? >>>> >>>> Many thanks again! >>>> Cheers, >>>> Raphael >>>> >>>> >>>> >>>> >>>> _______________________________________________ >>>> theora mailing list >>>> [email protected] >>>> http://lists.xiph.org/mailman/listinfo/theora >>>> >>> >> _______________________________________________ theora mailing list [email protected] http://lists.xiph.org/mailman/listinfo/theora