Any plans for lossless VP8/WebM ?
599 views
Skip to first unread message
Dove
Sep 1, 2010, 10:40:18 AM9/1/10
to WebM Discussion
I'm interesting are any plans for lossless VP8/WebM ?
Paul Wilkins
Sep 1, 2010, 10:55:43 AM9/1/10
to WebM Discussion
Certainly in future iterations / generations of the codec this is
something that is on our radar.
something that is on our radar.
Dove
Sep 2, 2010, 2:36:56 PM9/2/10
to WebM Discussion
Good news! It's so needed! Could it be already VP9 ;) ?
And could you add lossless part to the roadmap list
http://www.webmproject.org/code/roadmap/ ?
On 1 сен, 18:55, Paul Wilkins <paulwilk...@google.com> wrote:
> Certainly in future iterations / generations of the codec this is
> something that is on our radar.
>
Danny Piccirillo
Oct 15, 2010, 5:00:30 AM10/15/10
to WebM Discussion
Seeing this on the roadmap would be great! I it's a concern that
people shouldn't have to dig up a google groups thread to find the
answer to =]
On Sep 2, 2:36 pm, Dove <dimac...@bk.ru> wrote:
> Good news! It's so needed! Could it be already VP9 ;) ?
> And could you add lossless part to the roadmap listhttp://www.webmproject.org/code/roadmap/ ?
>
people shouldn't have to dig up a google groups thread to find the
answer to =]
On Sep 2, 2:36 pm, Dove <dimac...@bk.ru> wrote:
> Good news! It's so needed! Could it be already VP9 ;) ?
> And could you add lossless part to the roadmap listhttp://www.webmproject.org/code/roadmap/ ?
>
> On 1 ÓÅÎ, 18:55, Paul Wilkins <paulwilk...@google.com> wrote:
>
>
>
>
>
>
>
> > Certainly in future iterations / generations of the codec this is
> > something that is on our radar.
>
>
>
>
>
>
>
>
> > Certainly in future iterations / generations of the codec this is
> > something that is on our radar.
>
Paul Wilkins
Oct 15, 2010, 11:45:42 AM10/15/10
to WebM Discussion
Just to be clear, the current VP8 bitstream does not support a
lossless mode, so this is something for consideration in the next
generation bitstream, but it IS something we have identified as
important.
The project is open and as such anyone can contribute ideas or
features that they think might be good for a “next generation
bitstream” and we have created an experimental repository branch, in
which people can build and test out their ideas, to facilitate this.
So it is not just a case of “our” roadmap; we perhaps need a dedicated
page where contributors can detail ideas that they are working on and/
or features they think need to be added.
On Oct 15, 10:00 am, Danny Piccirillo
lossless mode, so this is something for consideration in the next
generation bitstream, but it IS something we have identified as
important.
The project is open and as such anyone can contribute ideas or
features that they think might be good for a “next generation
bitstream” and we have created an experimental repository branch, in
which people can build and test out their ideas, to facilitate this.
So it is not just a case of “our” roadmap; we perhaps need a dedicated
page where contributors can detail ideas that they are working on and/
or features they think need to be added.
On Oct 15, 10:00 am, Danny Piccirillo
Carlos Jose Lenarts Ramis
Oct 15, 2010, 11:48:35 PM10/15/10
to webm-d...@webmproject.org
2010/10/15 Paul Wilkins <paulw...@google.com>
Just to be clear, the current VP8 bitstream does not support a
lossless mode, so this is something for consideration in the next
generation bitstream, but it IS something we have identified as
important.
The project is open and as such anyone can contribute ideas or
features that they think might be good for a “next generation
bitstream” and we have created an experimental repository branch, in
which people can build and test out their ideas, to facilitate this.
So it is not just a case of “our” roadmap; we perhaps need a dedicated
page where contributors can detail ideas that they are working on and/
or features they think need to be added.
I think that a dedicated page is the way to go and a dedicated mailing list can also be a good thing for the next generation development.
For example a feature that I find interesting is the capability of bilevel multiresolution, the application is a little different and is for content distribution over air or disc, is basically that we can decode the video at 1/4 of the original resolution with the basic data and we need the complementary data to recover the full
resolution but this complementary data can be encrypted in a pay per view scheme, as example this can be used by tv channels and a way for the free to air channels to make money from the HD in a hybrid free/pay manner with a minimum impact on the needed bandwidth for the HD signal.
Numien
Oct 22, 2010, 8:20:17 PM10/22/10
to WebM Discussion
On Oct 15, 11:48 pm, Carlos Jose Lenarts Ramis <goda...@gmail.com>
wrote:
needing multiple LODs as well.
For completely selfish reasons my first thought is in games; embedded
videos could be streamed at low res until a player gets close, upon
which the high res data is added. This saves having to either stream
and decode high res all the time, or do an expensive (in performance
terms) stream disconnect/reconnect to switch qualities.
I could see it being used in other places too, such as YouTube's
multiple resolutions. Instead of having to discard everything that's
been cached on a resolution change, the viewer could just start
downloading the higher res data to add to it.
wrote:
> For example a feature that I find interesting is the capability of bilevel
> multiresolution, the application is a little different and is for content
> distribution over air or disc, is basically that we can decode the video at
> 1/4 of the original resolution with the basic data and we need the
> complementary data to recover the full
> resolution but this complementary data can be encrypted in a pay per view
> scheme, as example this can be used by tv channels and a way for the free to
> air channels to make money from the HD in a hybrid free/pay manner with a
> minimum impact on the needed bandwidth for the HD signal.
That's a very interesting idea. It would be useful in other situations
> multiresolution, the application is a little different and is for content
> distribution over air or disc, is basically that we can decode the video at
> 1/4 of the original resolution with the basic data and we need the
> complementary data to recover the full
> resolution but this complementary data can be encrypted in a pay per view
> scheme, as example this can be used by tv channels and a way for the free to
> air channels to make money from the HD in a hybrid free/pay manner with a
> minimum impact on the needed bandwidth for the HD signal.
needing multiple LODs as well.
For completely selfish reasons my first thought is in games; embedded
videos could be streamed at low res until a player gets close, upon
which the high res data is added. This saves having to either stream
and decode high res all the time, or do an expensive (in performance
terms) stream disconnect/reconnect to switch qualities.
I could see it being used in other places too, such as YouTube's
multiple resolutions. Instead of having to discard everything that's
been cached on a resolution change, the viewer could just start
downloading the higher res data to add to it.
Reply all
Reply to author
Forward
0 new messages