Plans to support AV1 video codec?

Bit stream is supposed to be frozen this fall. 

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[Sebastian Dröge-3]

On Fri, 2017-06-23 at 22:04 -0400, Aaron Boxer wrote:
> Bit stream is supposed to be frozen this fall. 

Someone would need to write a elements around the decoder/encoder
library and add support to the relevant muxers/demuxer.

I'm not aware of any concrete plans, it's mostly a matter of someone
finding the time to do so. Go ahead if you were asking because you
wanted to spend some time on that :)

--
Sebastian Dröge, Centricular Ltd · http://www.centricular.com
_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

On Sat, Jun 24, 2017 at 3:04 AM, Sebastian Dröge <[hidden email]> wrote:

On Fri, 2017-06-23 at 22:04 -0400, Aaron Boxer wrote:
> Bit stream is supposed to be frozen this fall. 

Someone would need to write a elements around the decoder/encoder
library and add support to the relevant muxers/demuxer.

I'm not aware of any concrete plans, it's mostly a matter of someone
finding the time to do so. Go ahead if you were asking because you
wanted to spend some time on that :)

Don't really have the time :) but I would like to get this started.

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

The sad thing about this new codec is that after the inventor of part of the encoder helped

Google for 3 years to integrate his algorithm with AV1, they are now trying to patent his invention:

https://news.ycombinator.com/item?id=14583691

On Sat, Jun 24, 2017 at 8:27 AM, Aaron Boxer <[hidden email]> wrote:

On Sat, Jun 24, 2017 at 3:04 AM, Sebastian Dröge <[hidden email]> wrote:

On Fri, 2017-06-23 at 22:04 -0400, Aaron Boxer wrote:
> Bit stream is supposed to be frozen this fall. 

Someone would need to write a elements around the decoder/encoder
library and add support to the relevant muxers/demuxer.

I'm not aware of any concrete plans, it's mostly a matter of someone
finding the time to do so. Go ahead if you were asking because you
wanted to spend some time on that :)

Don't really have the time :) but I would like to get this started.

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[Sebastian Dröge-3]

On Sun, 2017-06-25 at 08:45 -0400, Aaron Boxer wrote:
> The sad thing about this new codec is that after the inventor of part
> of the encoder helped
> Google for 3 years to integrate his algorithm with AV1, they are now
> trying to patent his invention:
>
> https://news.ycombinator.com/item?id=14583691

AFAIU that says that Google applied for a patent. So this generally
needs more context to say anything, but keep in mind that Google also
supposedly owns patents related to VP8/9/10 (which it got together with
On2 back then), but it grants a free (etc) patent license for every use
(see the libvpx license for details/accurate wording). In theory for
AV1 this could also be for defensive measures. Who knows? They should
really clarify that somewhere though.

However I doubt that they plan to make AV1 useless (could as well use
HEVC then) considering how much time and effort they put into it.

--
Sebastian Dröge, Centricular Ltd · http://www.centricular.com
_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[Carlos Rafael Giani]

From reading their license, and from the Reddit comments here: https://www.reddit.com/r/programming/comments/6h08z5/google_is_currently_trying_to_patent_video/

it seems that this is intended to be a defensive patent. Which is much more in line with the whole goal of AOM.

On 2017-06-25 18:53, Sebastian Dröge wrote:

On Sun, 2017-06-25 at 08:45 -0400, Aaron Boxer wrote:

The sad thing about this new codec is that after the inventor of part
of the encoder helped
Google for 3 years to integrate his algorithm with AV1, they are now
trying to patent his invention:

https://news.ycombinator.com/item?id=14583691

AFAIU that says that Google applied for a patent. So this generally
needs more context to say anything, but keep in mind that Google also
supposedly owns patents related to VP8/9/10 (which it got together with
On2 back then), but it grants a free (etc) patent license for every use
(see the libvpx license for details/accurate wording). In theory for
AV1 this could also be for defensive measures. Who knows? They should
really clarify that somewhere though.

However I doubt that they plan to make AV1 useless (could as well use
HEVC then) considering how much time and effort they put into it.

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

On Sun, Jun 25, 2017 at 12:53 PM, Sebastian Dröge <[hidden email]> wrote:

On Sun, 2017-06-25 at 08:45 -0400, Aaron Boxer wrote:
> The sad thing about this new codec is that after the inventor of part
> of the encoder helped
> Google for 3 years to integrate his algorithm with AV1, they are now
> trying to patent his invention:
>
> https://news.ycombinator.com/item?id=14583691

AFAIU that says that Google applied for a patent. So this generally
needs more context to say anything, but keep in mind that Google also
supposedly owns patents related to VP8/9/10 (which it got together with
On2 back then), but it grants a free (etc) patent license for every use
(see the libvpx license for details/accurate wording). In theory for
AV1 this could also be for defensive measures. Who knows? They should
really clarify that somewhere though.

However I doubt that they plan to make AV1 useless (could as well use
HEVC then) considering how much time and effort they put into it.

Yes, most likely defensive. But, defensive or not, a patent shouldn't be granted

if there is prior art, of which there seems to be plenty. Hopefully USPTO

will see the folly of their ways :)

 

--
Sebastian Dröge, Centricular Ltd · http://www.centricular.com

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[Sean DuBois]

On Sun, Jun 25, 2017 at 11:12:18PM -0400, Aaron Boxer wrote:

> On Sun, Jun 25, 2017 at 12:53 PM, Sebastian Dröge <[hidden email]
> > wrote:
>
> > On Sun, 2017-06-25 at 08:45 -0400, Aaron Boxer wrote:
> > > The sad thing about this new codec is that after the inventor of part
> > > of the encoder helped
> > > Google for 3 years to integrate his algorithm with AV1, they are now
> > > trying to patent his invention:
> > >
> > > https://news.ycombinator.com/item?id=14583691
> >
> > AFAIU that says that Google applied for a patent. So this generally
> > needs more context to say anything, but keep in mind that Google also
> > supposedly owns patents related to VP8/9/10 (which it got together with
> > On2 back then), but it grants a free (etc) patent license for every use
> > (see the libvpx license for details/accurate wording). In theory for
> > AV1 this could also be for defensive measures. Who knows? They should
> > really clarify that somewhere though.
> >
> > However I doubt that they plan to make AV1 useless (could as well use
> > HEVC then) considering how much time and effort they put into it.
> >
>
> Yes, most likely defensive. But, defensive or not, a patent shouldn't be
> granted
> if there is prior art, of which there seems to be plenty. Hopefully USPTO
> will see the folly of their ways :)
>
>
>
>
>
>
> >
> > --
> > Sebastian Dröge, Centricular Ltd · http://www.centricular.com
> >
> > _______________________________________________
> > gstreamer-devel mailing list
> > [hidden email]
> > https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel
> >
> >

> _______________________________________________
> gstreamer-devel mailing list
> [hidden email]
> https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

I would love to take this on! I have only written very small patches
that have been committed, but have a bunch of elements that I wrote that
haven't been shared.

Does anyone have any recommendations for something I should use for
inspiration? vp8enc is in good, and maybe simpler then vp9enc?

thanks
_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

On Mon, Jun 26, 2017 at 1:14 AM, Sean DuBois <[hidden email]> wrote:

On Sun, Jun 25, 2017 at 11:12:18PM -0400, Aaron Boxer wrote:
> On Sun, Jun 25, 2017 at 12:53 PM, Sebastian Dröge <[hidden email]
> > wrote:
>
> > On Sun, 2017-06-25 at 08:45 -0400, Aaron Boxer wrote:
> > > The sad thing about this new codec is that after the inventor of part
> > > of the encoder helped
> > > Google for 3 years to integrate his algorithm with AV1, they are now
> > > trying to patent his invention:
> > >
> > > https://news.ycombinator.com/item?id=14583691
> >
> > AFAIU that says that Google applied for a patent. So this generally
> > needs more context to say anything, but keep in mind that Google also
> > supposedly owns patents related to VP8/9/10 (which it got together with
> > On2 back then), but it grants a free (etc) patent license for every use
> > (see the libvpx license for details/accurate wording). In theory for
> > AV1 this could also be for defensive measures. Who knows? They should
> > really clarify that somewhere though.
> >
> > However I doubt that they plan to make AV1 useless (could as well use
> > HEVC then) considering how much time and effort they put into it.
> >
>
> Yes, most likely defensive. But, defensive or not, a patent shouldn't be
> granted
> if there is prior art, of which there seems to be plenty. Hopefully USPTO
> will see the folly of their ways :)
>
>
>
>
>
>
> >
> > --
> > Sebastian Dröge, Centricular Ltd · http://www.centricular.com
> >
> > _______________________________________________
> > gstreamer-devel mailing list
> > [hidden email]
> > https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel
> >
> >

> _______________________________________________
> gstreamer-devel mailing list
> [hidden email]
> https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

Hi Sean,

 

I would love to take this on! I have only written very small patches
that have been committed, but have a bunch of elements that I wrote that
haven't been shared.

Does anyone have any recommendations for something I should use for
inspiration? vp8enc is in good, and maybe simpler then vp9enc?

Awesome! Here is the bugzilla issue:  https://bugzilla.gnome.org/show_bug.cgi?id=784160

And here is my very preliminary start to the patch, based on vp9:

https://github.com/GrokImageCompression/gst-plugins-base/tree/784160-av1
https://github.com/GrokImageCompression/gst-plugins-good/tree/784160-av1

You can fork these, or start from scratch, as you wish.

Cheers,

Aaron

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel