Re: Problems in creating aggregate device
Re: Problems in creating aggregate device
- Subject: Re: Problems in creating aggregate device
- From: Jeff Moore <email@hidden>
- Date: Fri, 22 May 2009 13:13:29 -0700
I think my point here is that the fact that AMS is showing the
incorrect stream/channel count is most likely a bug in either AMS or
the HAL. There isn't anything you can do about it from your process,
which likely does not have any problems.
On May 22, 2009, at 1:05 PM, Philip Lukidis wrote:
As you say, I am creating an empty aggregate device, and then adding
the subdevice(s) to it after the fact. I certainly *don't* expect
to see 0 input and output channels. However, *sometimes* AMS sees
no 0 input and output channels, while hallab does see them. Hallab
sees the streams as well.
I had assumed that if AMS did not see the channels then there was a
problem in creating the aggregate device. It would be interesting
to know how AMS queries the channel state, so I could do the same,
if possible. Or am I looking in the wrong direction?
thank you,
Philip Lukidis
PS:
Note that I mentioned subdevice(s) instead of subdevices, because
10.5.7 seems to aggregate input and output subdevices of USB into 1
device, unlike earlier versions of OSX. So in order to maintain
backwards compatibility with some applications which look for our
aggregate device by name, I create an aggregate device with the old
device name which the software could use.
-----Original Message-----
From: coreaudio-api-bounces+plukidis=email@hidden
on behalf of Jeff Moore
Sent: Fri 5/22/2009 3:29 PM
To: CoreAudio API
Subject: Re: Problems in creating aggregate device
Wait. I guess I don't understand the question. You are creating an
aggregate device and adding some sub-devices to it, right? Why would
you expect that the resulting aggregate would not have any channels?
That's not what should happen. You should at least see all the
streams/
channels for the master device plus those of the sub-devices whose
sample rate is the same as the master.
It seems to me that it is AMS that is not seeing the proper number of
streams/channels.
On May 22, 2009, at 11:19 AM, Philip Lukidis wrote:
> Thank you for taking the time to answer. I am running on 10.5.7
> with the issue, and on 10.4.11 I don't (so far) seem to be running
> into any problems. My daemon programmatically creates the aggregate
> device when it detects that our USB device has arrived. I do give a
> healthy amount of pausing just to make sure that CoreAudio is
> sufficiently ready, but this is not the best solution, I think.
>
> However, since the MIDI audio properties tab seems to see that the
> device has no channels available, surely I should be able to do the
> same? Do you know if this is possible?
>
> thank you,
>
> Philip Lukidis
>
>
> -----Original Message-----
> From: coreaudio-api-bounces+plukidis=email@hidden
> on behalf of Jeff Moore
> Sent: Fri 5/22/2009 2:10 PM
> To: CoreAudio API
> Subject: Re: Problems in creating aggregate device
>
> First off, make sure you are running on 10.5.7. There were a
number of
> fixes to the aggregate device support in that release - several of
> which addressed issues with creation and editing.
>
> At any rate, there really isn't anything you can do from your own
> process to detect what is going on in the AMS process with respect
to
> the composition of an aggregate device.
>
> Finally, it is possible that you are running into some kind of bug
in
> the aggregate device support. So be sure to file a bug report for us
> to look at.
>
>
> On May 22, 2009, at 6:10 AM, Philip Lukidis wrote:
>
> > I am trying to programmatically create an aggregate device, and
for
> > the most part, it works. Sometimes, however, the coreaudio API
> > calls succeed, but in the audio-midi tab, the aggregate device
has 0
> > inputs and 0 outputs. I would like to be able to detect this,
> > delete the aggregate audio device, and try again. However, I
cannot
> > detect this condition myself. Hallab cannot detect it either, for
> > it displays the correct (non-zero) number of input and output
> > channels. When I try to use the
> > kAudioDevicePropertyStreamConfiguration property and count the
> > number of channels, I always get the correct number.
> >
> > So, can anyone tell me how else I can count the number of
available
> > inputs and outputs? The audio/midi properties tab gets it right,
> > and I would like to be able to do the same.
>
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Coreaudio-api mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden