Simultaneous outgoing audio streams

Simultaneous outgoing audio streams

[Mark Stanovich]

Sorry if there are multiple messages, it seems that my other email was blocked?

Does anyone know if it is possible to send audio to multiple headsets from one adapter?

I have looked around the source, docs, and the web but have been unable to find an answer.

Currently I can send audio to one headset but once I try to send audio to another headset I get the following messages from bluetoothd:

bluetoothd[17481]: Accepted new client connection on unix socket (fd=22)
bluetoothd[17481]: Audio API: BT_REQUEST <- BT_GET_CAPABILITIES
bluetoothd[17481]: Audio API: BT_RESPONSE -> BT_GET_CAPABILITIES
bluetoothd[17481]: Audio API: BT_REQUEST <- BT_OPEN
bluetoothd[17481]: open sco - object=ANY source=ANY destination=00:00:00:00:00:00 lock=write
bluetoothd[17481]: Audio API: BT_RESPONSE -> BT_OPEN
bluetoothd[17481]: Audio API: BT_REQUEST <- BT_SET_CONFIGURATION
bluetoothd[17481]: State changed /org/bluez/17481/hci0/dev_00_00_00_00_00_00: HEADSET_STATE_DISCONNECTED -> HEADSET_STATE_CONNECTING
bluetoothd[17481]: adapter_get_device(00:16:8F:24:6E:A6)
bluetoothd[17481]: Unable to get service record: Host is down (112)
bluetoothd[17481]: config failed
bluetoothd[17481]: sending error Input/output error(5)
bluetoothd[17481]: Audio API: BT_ERROR -> BT_SET_CONFIGURATION
bluetoothd[17481]: State changed /org/bluez/17481/hci0/dev_00_00_00_00_00_00: HEADSET_STATE_CONNECTING -> HEADSET_STATE_DISCONNECTED
bluetoothd[17481]: Unix client disconnected (fd=22)
bluetoothd[17481]: client_free(0xb8dfa648)

The mac addresses have been changed to all zeros.

Maybe this is a limitation of the hardware or the device driver?

I have set the MaxConnected to greater than 1, actually 15.  Are there any other settings I should be aware of.

Thanks


      

Re: Simultaneous outgoing audio streams

[Luiz Augusto von Dentz]

Hi,

On Wed, Nov 11, 2009 at 7:54 PM, Mark Stanovich <stanmj@yahoo.com> wrote:
> Sorry if there are multiple messages, it seems that my other email was blocked?
>
> Does anyone know if it is possible to send audio to multiple headsets from one adapter?
>
> I have looked around the source, docs, and the web but have been unable to find an answer.
>
> Currently I can send audio to one headset but once I try to send audio to another headset I get the following messages from bluetoothd:
>
> bluetoothd[17481]: Accepted new client connection on unix socket (fd=22)
> bluetoothd[17481]: Audio API: BT_REQUEST <- BT_GET_CAPABILITIES
> bluetoothd[17481]: Audio API: BT_RESPONSE -> BT_GET_CAPABILITIES
> bluetoothd[17481]: Audio API: BT_REQUEST <- BT_OPEN
> bluetoothd[17481]: open sco - object=ANY source=ANY destination=00:00:00:00:00:00 lock=write
> bluetoothd[17481]: Audio API: BT_RESPONSE -> BT_OPEN
> bluetoothd[17481]: Audio API: BT_REQUEST <- BT_SET_CONFIGURATION
> bluetoothd[17481]: State changed /org/bluez/17481/hci0/dev_00_00_00_00_00_00: HEADSET_STATE_DISCONNECTED -> HEADSET_STATE_CONNECTING
> bluetoothd[17481]: adapter_get_device(00:16:8F:24:6E:A6)
> bluetoothd[17481]: Unable to get service record: Host is down (112)
> bluetoothd[17481]: config failed
> bluetoothd[17481]: sending error Input/output error(5)
> bluetoothd[17481]: Audio API: BT_ERROR -> BT_SET_CONFIGURATION
> bluetoothd[17481]: State changed /org/bluez/17481/hci0/dev_00_00_00_00_00_00: HEADSET_STATE_CONNECTING -> HEADSET_STATE_DISCONNECTED
> bluetoothd[17481]: Unix client disconnected (fd=22)
> bluetoothd[17481]: client_free(0xb8dfa648)
>
> The mac addresses have been changed to all zeros.
>
> Maybe this is a limitation of the hardware or the device driver?
>
> I have set the MaxConnected to greater than 1, actually 15.  Are there any other settings I should be aware of.

For hsp/hfp you need to set MaxConnected, but for a2dp it is required
to change the number of sources/sinks to do the same, both can be
configured via audio.conf. Now about the address, it really seems that
you are using 00:00:00:00:00:00 to connect which explain why you are
getting the host is down error, you can either choose a valid address
or a valid dbus object.


-- 
Luiz Augusto von Dentz
Engenheiro de Computação

Re: Simultaneous outgoing audio streams

[Mark Stanovich]

> > Does anyone know if it is possible to send audio to multiple headsets from one adapter?
> >
> > I have looked around the source, docs, and the web but have been unable to find an answer.
> >
> > Currently I can send audio to one headset but once I
> try to send audio to another headset I get the following
> messages from bluetoothd:
> >
> > bluetoothd[17481]: Accepted new client connection on
> unix socket (fd=22)
> > bluetoothd[17481]: Audio API: BT_REQUEST <-
> BT_GET_CAPABILITIES
> > bluetoothd[17481]: Audio API: BT_RESPONSE ->
> BT_GET_CAPABILITIES
> > bluetoothd[17481]: Audio API: BT_REQUEST <-
> BT_OPEN
> > bluetoothd[17481]: open sco - object=ANY source=ANY
> destination=00:00:00:00:00:00 lock=write
> > bluetoothd[17481]: Audio API: BT_RESPONSE ->
> BT_OPEN
> > bluetoothd[17481]: Audio API: BT_REQUEST <-
> BT_SET_CONFIGURATION
> > bluetoothd[17481]: State changed
> /org/bluez/17481/hci0/dev_00_00_00_00_00_00:
> HEADSET_STATE_DISCONNECTED -> HEADSET_STATE_CONNECTING
> > bluetoothd[17481]:
> adapter_get_device(00:16:8F:24:6E:A6)
> > bluetoothd[17481]: Unable to get service record: Host
> is down (112)
> > bluetoothd[17481]: config failed
> > bluetoothd[17481]: sending error Input/output
> error(5)
> > bluetoothd[17481]: Audio API: BT_ERROR ->
> BT_SET_CONFIGURATION
> > bluetoothd[17481]: State changed
> /org/bluez/17481/hci0/dev_00_00_00_00_00_00:
> HEADSET_STATE_CONNECTING -> HEADSET_STATE_DISCONNECTED
> > bluetoothd[17481]: Unix client disconnected (fd=22)
> > bluetoothd[17481]: client_free(0xb8dfa648)
> >
> > The mac addresses have been changed to all zeros.
> >
> > Maybe this is a limitation of the hardware or the
> device driver?
> >
> > I have set the MaxConnected to greater than 1,
> actually 15.  Are there any other settings I should be
> aware of.
> 
> For hsp/hfp you need to set MaxConnected, but for a2dp it

I have changed the MaxConnected under [Headset] to 15, is this the same as hsp/hfp, or should I create a new section?

> is required to change the number of sources/sinks to do the same, both
> can be configured via audio.conf. Now about the address, it really
> seems that you are using 00:00:00:00:00:00 to connect which explain

I just changed these to all zeros for privacy, and actually I missed one :)  Realistically, I guess it doesn't matter for other to see the mac, I just saw other people blocking their's out, so I thought maybe there was some privacy/security concern.

> why you are getting the host is down error, you can either choose a
> valid address or a valid dbus object.

Thanks,
Mark



      

Re: Simultaneous outgoing audio streams

[Luiz Augusto von Dentz]

Hi,

On Thu, Nov 12, 2009 at 10:20 AM, Mark Stanovich <stanmj@yahoo.com> wrote:
>> For hsp/hfp you need to set MaxConnected, but for a2dp it
>
> I have changed the MaxConnected under [Headset] to 15, is this the same as hsp/hfp, or should I create a new section?
>
>> is required to change the number of sources/sinks to do the same, both
>> can be configured via audio.conf. Now about the address, it really
>> seems that you are using 00:00:00:00:00:00 to connect which explain
>
> I just changed these to all zeros for privacy, and actually I missed one :)  Realistically, I guess it doesn't matter for other to see the mac, I just saw other people blocking their's out, so I thought maybe there was some privacy/security concern.
>
>> why you are getting the host is down error, you can either choose a
>> valid address or a valid dbus object.

Sorry I just interpret this all wrong, as the application has been
sending 00:00:00:00:00:00, but yeah you seem to be doing it right
although Im not sure you can have more than one SCO connection active,
but that doesn't explain the host is down error, so perhaps there is
something else blocking you.

-- 
Luiz Augusto von Dentz
Engenheiro de Computação

Re: Simultaneous outgoing audio streams

[Mark Stanovich]

Luiz Augusto von Dentz wrote:


> Sorry I just interpret this all wrong, as the application has been
> sending 00:00:00:00:00:00, but yeah you seem to be doing it right
> although Im not sure you can have more than one SCO connection active,
> but that doesn't explain the host is down error, so perhaps there is
> something else blocking you.

I attempted briefly to go through the code and see what exactly is causing the error.  It appears that bluetoothd starts a search for the headset and is unable to find it (or prevented from finding it).  I am unclear why it cannot find the headset when another audio stream is in progress.  It could be that there is some sort of locking mechanism in place.  The code is relatively new to me.  Maybe as you mentioned it is requesting another SCO connection and prevented from creating one with another device?  I will try to dig deeper and see where it is actually getting blocked.

   - Mark



      

Re: Simultaneous outgoing audio streams

[Mark Stanovich]

Mark Stanovich wrote:


> Maybe as you mentioned it is requesting another SCO connection and prevented 
> from creating one with another device?  I will try to dig deeper and see where 

Looking at the output again, I get something different, with a sco_bind: Address already in use error.  I guess this indicates that there can be only one SCO connection at a time?

bluetoothd[13455]: Accepted new client connection on unix socket (fd=22)
bluetoothd[13455]: Audio API: BT_REQUEST <- BT_GET_CAPABILITIES
bluetoothd[13455]: Audio API: BT_RESPONSE -> BT_GET_CAPABILITIES
bluetoothd[13455]: Audio API: BT_REQUEST <- BT_OPEN
bluetoothd[13455]: open sco - object=ANY source=ANY destination=00:18:13:87:3D:BF lock=write
bluetoothd[13455]: Audio API: BT_RESPONSE -> BT_OPEN
bluetoothd[13455]: Audio API: BT_REQUEST <- BT_SET_CONFIGURATION
bluetoothd[13455]: State changed /org/bluez/13455/hci0/dev_00_18_13_87_3D_BF: HEADSET_STATE_DISCONNECTED -> HEADSET_STATE_CONNECTING
bluetoothd[13455]: link_key_request (sba=00:11:67:00:00:00, dba=00:18:13:87:3D:BF)
bluetoothd[13455]: kernel auth requirements = 0x00
bluetoothd[13455]: stored link key type = 0x00
bluetoothd[13455]: adapter_get_device(00:18:13:87:3D:BF)
bluetoothd[13455]: Discovered Handsfree service on channel 2
bluetoothd[13455]: /org/bluez/13455/hci0/dev_00_18_13_87_3D_BF: Connecting to 00:18:13:87:3D:BF channel 2
bluetoothd[13455]: /org/bluez/13455/hci0/dev_00_18_13_87_3D_BF: Connected to 00:18:13:87:3D:BF
bluetoothd[13455]: Received AT+BRSF=24
bluetoothd[13455]: HFP HF features: "Voice recognition activation" "Remote volume control" 
bluetoothd[13455]: Received AT+CIND=?
bluetoothd[13455]: Received AT+CIND?
bluetoothd[13455]: Received AT+CMER=3, 0, 0, 1
bluetoothd[13455]: Event reporting (CMER): mode=3, ind=1
bluetoothd[13455]: HFP Service Level Connection established
bluetoothd[13455]: telephony-dummy: device 0xb916fb68 connected
bluetoothd[13455]: State changed /org/bluez/13455/hci0/dev_00_18_13_87_3D_BF: HEADSET_STATE_CONNECTING -> HEADSET_STATE_CONNECTED
bluetoothd[13455]: Audio API: BT_RESPONSE -> BT_SET_CONFIGURATION
bluetoothd[13455]: Audio API: BT_REQUEST <- BT_START_STREAM
bluetoothd[13455]: sco_bind: Address already in use (98)
bluetoothd[13455]: start_resume: resume failed
bluetoothd[13455]: sending error Input/output error(5)
bluetoothd[13455]: Audio API: BT_ERROR -> BT_START_STREAM
bluetoothd[13455]: Unix client disconnected (fd=22)
bluetoothd[13455]: client_free(0xb91691f8)
bluetoothd[13455]: Received AT+VGS=09
bluetoothd[13455]: Received AT*SEAUDIO=4,4022
bluetoothd[13455]: Badly formated or unrecognized command: AT*SEAUDIO=4,4022
bluetoothd[13455]: Received AT*SEAM="HBH-PV705",13
bluetoothd[13455]: Badly formated or unrecognized command: AT*SEAM="HBH-PV705",13
bluetoothd[13455]: Received AT*SEABS=0
bluetoothd[13455]: Badly formated or unrecognized command: AT*SEABS=0
bluetoothd[13455]: telephony-dummy: device 0xb916fb68 disconnected
bluetoothd[13455]: State changed /org/bluez/13455/hci0/dev_00_18_13_87_3D_BF: HEADSET_STATE_CONNECTED -> HEADSET_STATE_DISCONNECTED



      

Re: Simultaneous outgoing audio streams

[Brad Midgley]

Mark

> Looking at the output again, I get something different, with a sco_bind: Address already in use error.  I guess this indicates that there can be only one SCO connection at a time?

>From memory, it seems this was a usb resource allocation issue. I
don't remember all the specifics at the moment, only that the
alternate setting of the isochronous interface needed to be adjusted
dynamically but we didn't have the right hook to do it.

-- 
Brad Midgley