NHW Project - Is there interested people?
Raphael Canut <[email protected]> Fri, 7 May 2021 17:18:15 +0200
| Newsgroups | gmane.comp.multimedia.ogg.theora.general |
|---|---|
| Message-ID | <CAKE58qEAw8c2qApqMGf1+XswE+G1x=TG7JExq0CoLw72UMzNwg@mail.gmail.com> |
--===============6108923930536684397== Content-Type: multipart/alternative; boundary="00000000000087fe4c05c1bef048" --00000000000087fe4c05c1bef048 Content-Type: text/plain; charset="UTF-8" Hello, Just a quick note, some people told me in the past that I must adapt NHW to any image size else nobody will consider and use it.That's certainly right, but again I am working on NHW on my spare time and this task is too exhausting for me.Also, I feel now that it is useless, as I have read that in the very next months, the great AVIF and JPEG XL will be definitely adopted everywhere on the Internet. Some people also told me to forget about image compression and turn my codec into a video codec, but that's also too much work for my free time.So I have come to the conclusion that very soon when AVIF and JPEG XL will be adopted, then I will start to progressively abandon/stop the NHW Project, because it is also nearly 15 years that I am working on it and since the last years, I have seriously started to lose my energy for this project, and furthermore now there is no future for it. However it would be great to find someone that could take over the NHW Project.Turning it into a video codec for example, because there was interest in NHW for a MotionJPEG replacement, because a company was considering MJPEG for low-power camera devices, and MJPEG was too heavyweight for their processor (dual-core 240MHz Xtensa LX6, if I remember correctly).For now with rough calculation based on the latest Independent JPEG Group (IJG) C source code, NHW is 1.5x faster to encode than JPEG (and 2x faster to decode), but with switching off heavy processings with very few quality loss, NHW can be 2x faster to encode than JPEG and with a notably better quality/compression.This engineer told me that if NHW will be 2x faster to encode than MJPEG (and with a better quality furthermore) then this could be very interesting for such embedded products. So again, as I think it will be soon pointless to work on NHW image compression whereas 2 new standards will be adopted on the Internet, I wanted to let you know that it would be great to find someone/company/organization that are intereted in NHW and would like to take over the project and continue its research and development notably for some video codec use cases. So I would like to start to take contact as I am progressively stopping my work on NHW, if you think you could and want to work on and take over the NHW Project, or would know such a person, do not hesitate to contact me, would be great! Many thanks. Cheers, Raphael --00000000000087fe4c05c1bef048 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable <div dir=3D"ltr">Hello,<br><br>Just a quick note, some people told me in th= e past that I must adapt NHW to any image size else nobody will consider an= d use it.That's certainly right, but again I am working on NHW on my sp= are time and this task is too exhausting for me.Also, I feel now that it is= useless, as I have read that in the very next months, the great AVIF and J= PEG XL will be definitely adopted everywhere on the Internet.<br><br>Some p= eople also told me to forget about image compression and turn my codec into= a video codec, but that's also too much work for my free time.So I hav= e come to the conclusion that very soon when AVIF and JPEG XL will be adopt= ed, then I will start to progressively abandon/stop the NHW Project, becaus= e it is also nearly 15 years that I am working on it and since the last yea= rs, I have seriously started to lose my energy for this project, and furthe= rmore now there is no future for it.<br><br>However it would be great to fi= nd someone that could take over the NHW Project.Turning it into a video cod= ec for example, because there was interest in NHW for a MotionJPEG replacem= ent, because a company was considering MJPEG for low-power camera devices, = and MJPEG was too heavyweight for their processor (dual-core 240MHz Xtensa = LX6, if I remember correctly).For now with rough calculation based on the l= atest Independent JPEG Group (IJG) C source code, NHW is 1.5x faster to enc= ode than JPEG (and 2x faster to decode), but with switching off heavy proce= ssings with very few quality loss, NHW can be 2x faster to encode than JPEG= and with a notably better quality/compression.This engineer told me that i= f NHW will be 2x faster to encode than MJPEG (and with a better quality fur= thermore) then this could be very interesting for such embedded products.<b= r><br>So again, as I think it will be soon pointless to work on NHW image c= ompression whereas 2 new standards will be adopted on the Internet, I wante= d to let you know that it would be great to find someone/company/organizati= on that are intereted in NHW and would like to take over the project and co= ntinue its research and development notably for some video codec use cases.= <br><br>So I would like to start to take contact as I am progressively stop= ping my work on NHW, if you think you could and want to work on and take ov= er the NHW Project, or would know such a person, do not hesitate to contact= me, would be great!<br><br>Many thanks.<br>Cheers,<br>Raphael</div> --00000000000087fe4c05c1bef048-- --===============6108923930536684397== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: base64 Content-Disposition: inline X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18KdGhlb3JhIG1h aWxpbmcgbGlzdAp0aGVvcmFAeGlwaC5vcmcKaHR0cDovL2xpc3RzLnhpcGgub3JnL21haWxtYW4v bGlzdGluZm8vdGhlb3JhCg== --===============6108923930536684397==--