I am using libopus for my implementation. I wonder if anybody in the list have any experience on how to make libopus dynamically adjust its bitrate? On Mar 3, 2015 10:42 PM, "Benjamin Schwartz" <benjamin.m.schwartz at gmail.com> wrote:> It sounds like your software isn't adjusting the opus bitrate in response > to network conditions. For example, many WebRTC implementations do not > adjust the opus bitrate, because it is small in comparison to the video > bitrate. However, opus itself does support continuously varying the > bitrate over a wide range. > > On Tue, Mar 3, 2015, 12:58 AM Kelvin Chua <kelchy at gmail.com> wrote: > >> Hi guys, >> >> I have been reading a lot about the "adaptiveness" of opus and i quote: >> >> ... can still change, e.g. to adapt to changing network conditions. >> useinbandfec ... >> >> can somebody please enlighten me on this "adaptiveness"? >> whatever way I do our tests, it sticks to the same sampling rate and the >> same average bitrate, it would go up, down a bit but that's it. >> When we get some network issues, bandwidth utilization stays the same. >> Am I interpreting it incorrectly? >> >> Kelvin Chua >> _______________________________________________ >> opus mailing list >> opus at xiph.org >> http://lists.xiph.org/mailman/listinfo/opus >> >-------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/opus/attachments/20150304/e8a2c442/attachment.htm
Hi Kelvin,
You can use something like
:opus_encoder_ctl(enc,OPUS_SET_BITRATE(bitrate));opus_encoder_ctl(enc,OPUS_SET_MAX_BANDWIDTH(bandpass));
bandpass is the audio bandpass?, eg: OPUS_BANDWIDTH_WIDEBAND .
You will need to calculate the codec bitrate from the available network bitrate
(by taking into account the size of the IP+UDP+RTP headers which is 40 bytes ) .
By default the audio bandwidth (bandpass) setting is OPUS_BANDWIDTH_FULLBAND ,
which will utilize more network bandwidth .
Regards,Dragos
From: Kelvin Chua <kelchy at gmail.com>
To: Benjamin Schwartz <benjamin.m.schwartz at gmail.com>
Cc: opus at xiph.org
Sent: Wednesday, March 4, 2015 2:27 AM
Subject: Re: [opus] adaptive bandwidth
I am using libopus for my implementation. I wonder if anybody in the list have
any experience on how to make libopus dynamically adjust its bitrate?On Mar 3,
2015 10:42 PM, "Benjamin Schwartz" <benjamin.m.schwartz at
gmail.com> wrote:
It sounds like your software isn't adjusting the opus bitrate in response to
network conditions.? For example, many WebRTC implementations do not adjust the
opus bitrate, because it is small in comparison to the video bitrate.? However,
opus itself does support continuously varying the bitrate over a wide range.
On Tue, Mar 3, 2015, 12:58 AM?Kelvin Chua <kelchy at gmail.com> wrote:
Hi guys,
I have been reading a lot about the "adaptiveness" of opus and i
quote:
... can still change, e.g. to adapt to changing network conditions. useinbandfec
...
can somebody please enlighten me on this "adaptiveness"?whatever way I
do our tests, it sticks to the same sampling rate and the same average bitrate,
it would go up, down a bit but that's it.When we get some network issues,
bandwidth utilization stays the same.Am I interpreting it incorrectly?
Kelvin Chua_______________________________________________
opus mailing list
opus at xiph.org
http://lists.xiph.org/mailman/listinfo/opus
_______________________________________________
opus mailing list
opus at xiph.org
http://lists.xiph.org/mailman/listinfo/opus
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.xiph.org/pipermail/opus/attachments/20150304/7dd3dcd7/attachment.htm
Thanks Dragos, I assume I will be setting those parameters during initialization of encoder right? Question is, if connection gets too lossy, how will opus adapt to it? Can it automatically shift bitrate down to minimize impact? Mark from IRC suggests that the app has to be aware of the losses and change it on the fly. Has anybody on the list tried this? Kelvin Chua On Wed, Mar 4, 2015 at 5:53 PM, Dragos Oancea <droancea at yahoo.com> wrote:> Hi Kelvin, > > You can use something like : > opus_encoder_ctl(enc,OPUS_SET_BITRATE(bitrate)); > opus_encoder_ctl(enc,OPUS_SET_MAX_BANDWIDTH(bandpass)); > bandpass is the audio bandpass , eg: OPUS_BANDWIDTH_WIDEBAND . > > You will need to calculate the codec bitrate from the available network > bitrate (by taking into account the size of the IP+UDP+RTP headers which is > 40 bytes ) . > By default the audio bandwidth (bandpass) setting is > OPUS_BANDWIDTH_FULLBAND , which will utilize more network bandwidth . > > > Regards, > Dragos > > ------------------------------ > *From:* Kelvin Chua <kelchy at gmail.com> > *To:* Benjamin Schwartz <benjamin.m.schwartz at gmail.com> > *Cc:* opus at xiph.org > *Sent:* Wednesday, March 4, 2015 2:27 AM > *Subject:* Re: [opus] adaptive bandwidth > > I am using libopus for my implementation. I wonder if anybody in the list > have any experience on how to make libopus dynamically adjust its bitrate? > On Mar 3, 2015 10:42 PM, "Benjamin Schwartz" < > benjamin.m.schwartz at gmail.com> wrote: > > It sounds like your software isn't adjusting the opus bitrate in response > to network conditions. For example, many WebRTC implementations do not > adjust the opus bitrate, because it is small in comparison to the video > bitrate. However, opus itself does support continuously varying the > bitrate over a wide range. > > On Tue, Mar 3, 2015, 12:58 AM Kelvin Chua <kelchy at gmail.com> wrote: > > Hi guys, > > I have been reading a lot about the "adaptiveness" of opus and i quote: > > ... can still change, e.g. to adapt to changing network conditions. > useinbandfec ... > > can somebody please enlighten me on this "adaptiveness"? > whatever way I do our tests, it sticks to the same sampling rate and the > same average bitrate, it would go up, down a bit but that's it. > When we get some network issues, bandwidth utilization stays the same. > Am I interpreting it incorrectly? > > Kelvin Chua > _______________________________________________ > opus mailing list > opus at xiph.org > http://lists.xiph.org/mailman/listinfo/opus > > > _______________________________________________ > opus mailing list > opus at xiph.org > http://lists.xiph.org/mailman/listinfo/opus > > >-------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/opus/attachments/20150304/6cf4906d/attachment.htm