Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Opus in <audio> demos
61 views
Skip to first unread message
Ralph Giles
May 3, 2012, 7:04:54 PM5/3/12
to dev-...@lists.mozilla.org, rob...@ocallahan.org
All,
Our basic support for Opus playback in the <audio> tag has landed in
inbound and will be in Firefox 15 Nightly in a few days. Robert and I
were discussing a few ideas on how to popularize this.
Forwarding the discussion so far...
On 12-05-03 4:38 AM, Robert O'Callahan wrote:
> How can we construct some demos to show it off?
On 12-05-03 9:27 AM, Ralph Giles wrote:
> We can make files with opusenc from
> https://git.xiph.org/?p=users/greg/opus-tools.git. I suppose
> we should do some to demonstrate the capabilities, with some
> CC licensed music and audio books. Then blog about it?
> Put up a demo page at wiki.mozilla.org?
On 12-05-03 3:05 PM, Robert O'Callahan wrote:
> Those are all good things. What about the magical Opus features
> like low delay etc?
On 12-05-03 3:31 PM, Ralph Giles wrote:
> Those don't really apply with http streaming. Our hope for Opus
> in <audio> was that it would get some traction as an unencumbered
> baseline for audio file distribution. The lower bitrate for
> voice-only recordings is about the only dramatic technical difference
> over Ogg in this context.
>
> We could also do fancier things with text tracks, which are hard to
> add to Ogg Vorbis (or mp3) files because of compatibility issues with
> older players...once our text track support is further along.
> For example, webvtt supports chapter markers, which are a common
> feature request for audio books.
>
> Maybe we could do a demo with a special server sending the same data
> over vorbis and opus, and show the difference in lag? Or better yet,
> the live mode switching. Until we have webrtc it might be sufficient >
to generate interest. David Richards' might be willing to do the
> server-side part; he's experimented with Opus streaming and remote
> websocket encoder controls.
On 12-05-03 3:35 PM, Robert O'Callahan wrote:
> Yes, if we had an Icecast streaming server set up to iterate through
> various bandwidth levels in a fixed pattern, that could make a pretty
> good demo (with visual feedback in the Web page showing the transfer
> rate over time).
Our basic support for Opus playback in the <audio> tag has landed in
inbound and will be in Firefox 15 Nightly in a few days. Robert and I
were discussing a few ideas on how to popularize this.
Forwarding the discussion so far...
On 12-05-03 4:38 AM, Robert O'Callahan wrote:
> How can we construct some demos to show it off?
On 12-05-03 9:27 AM, Ralph Giles wrote:
> We can make files with opusenc from
> https://git.xiph.org/?p=users/greg/opus-tools.git. I suppose
> we should do some to demonstrate the capabilities, with some
> CC licensed music and audio books. Then blog about it?
> Put up a demo page at wiki.mozilla.org?
On 12-05-03 3:05 PM, Robert O'Callahan wrote:
> Those are all good things. What about the magical Opus features
> like low delay etc?
On 12-05-03 3:31 PM, Ralph Giles wrote:
> Those don't really apply with http streaming. Our hope for Opus
> in <audio> was that it would get some traction as an unencumbered
> baseline for audio file distribution. The lower bitrate for
> voice-only recordings is about the only dramatic technical difference
> over Ogg in this context.
>
> We could also do fancier things with text tracks, which are hard to
> add to Ogg Vorbis (or mp3) files because of compatibility issues with
> older players...once our text track support is further along.
> For example, webvtt supports chapter markers, which are a common
> feature request for audio books.
>
> Maybe we could do a demo with a special server sending the same data
> over vorbis and opus, and show the difference in lag? Or better yet,
> the live mode switching. Until we have webrtc it might be sufficient >
to generate interest. David Richards' might be willing to do the
> server-side part; he's experimented with Opus streaming and remote
> websocket encoder controls.
On 12-05-03 3:35 PM, Robert O'Callahan wrote:
> Yes, if we had an Icecast streaming server set up to iterate through
> various bandwidth levels in a fixed pattern, that could make a pretty
> good demo (with visual feedback in the Web page showing the transfer
> rate over time).
Ralph Giles
May 3, 2012, 7:06:21 PM5/3/12
to dev-...@lists.mozilla.org, rob...@ocallahan.org
On 12-05-03 3:35 PM, Robert O'Callahan wrote:
> Yes, if we had an Icecast streaming server set up to iterate through
> various bandwidth levels in a fixed pattern, that could make a pretty
> good demo (with visual feedback in the Web page showing the transfer
> rate over time).
An interactive version is always more impressive, but could do a pretty
> Yes, if we had an Icecast streaming server set up to iterate through
> various bandwidth levels in a fixed pattern, that could make a pretty
> good demo (with visual feedback in the Web page showing the transfer
> rate over time).
good static version of this with popcorn.js.
-r
Cullen Jennings
May 3, 2012, 9:55:43 PM5/3/12
to Ralph Giles, dev-...@lists.mozilla.org, rob...@ocallahan.org
Looking forward to seeing the demo whatever it is. Once you have the demo going, and firefox 15 is out, it would be great if someone could send a link to the IETF Last Call on opus - that will help show that people really are implementing opus.
Cullen
> dev-media mailing list
> dev-...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-media
0 new messages