[pulseaudio-discuss] [PATCH v13 10/10] bluetooth: policy: Treat bi-directional A2DP profiles as suitable for VOIP

Hyperion h1p8r10n at yandex.com
Mon Nov 4 09:08:42 UTC 2019


Bluez multi-codec support is no longer a subject since today :
So I propose to test Pali's patch in master and see if it gets any regression
reports. If automatic fallback to legacy profile works, there shouldn't be any.

31.10.2019, 14:02, "Tanu Kaskinen" <tanuk at iki.fi>:
> On Mon, 2019-10-28 at 17:12 +0100, Hyperion wrote:
>>  - Bluetooth packet size is not constant : BT traffic is splited in radio slots
>>  that have fixed time lengh : so for a given bandwidth : slots size vary
>
> It's still not clear to me whether the bluetooth packets have constant
> size within one A2DP stream or not. Pali wrote earlier:
>
>     For established connection, bluetooth packet size is fixed. SBC
>     packet size depends on SBC bitpool (and other) parameters. And
>     because pulseaudio does not support generating fragmented SBC
>     packets, it can only fill those SBC frames which fits into
>     bluetooth packet size.
>
> That sounded like reducing the bitpool and therefore the SBC packet
> size mid-stream wouldn't help, because the lower level bluetooth packet
> size is constant for the duration of the stream. However, if the
> bluetooth packet size dynamically reacts to changes in the SBC packet
> sizes, then reducing the bitpool mid-stream is useful. Can someone
> confirm which it is?
>
> --
> Tanu
>
> https://www.patreon.com/tanuk
> https://liberapay.com/tanuk


More information about the pulseaudio-discuss mailing list