[pulseaudio-discuss] General Pulse audio async api questions

[Forwind info]

Thx Tanu. Will heed the advise.

Conor


On Wed, Jan 13, 2010 at 4:45 AM, Tanu Kaskinen <tanuk at iki.fi> wrote:

> ti, 2010-01-12 kello 11:48 +0000, Forwind info kirjoitti:
> > Hi all,
> >
> >
> > I have some notes below I made about using the Pulse Audio
> > asynchronous API. If you have a minute would you mind casting your
> > eyes over my proposed strategy. Any comments are very welcome.
> >
> >
> > thx,
> > Conor
> >
> >
> >
> >
> > 1.  To obtain the volume of primary Output Device.
> >     ans: Fetch the sink info via pa_context_get_sink_info_by_index().
> > The return value will  contain volume data type (pa_cvolume).
>
> I don't know what you mean by "primary output device". But if you
> somehow already know the index for the primary sink, whatever that
> means, then yes, use that function. In case you think pulseaudio
> recognizes a "primary sink" concept, please read
> http://pulseaudio.org/wiki/DefaultDevice
>
> > 2.Differentiate between headphone volume if plugged in and speaker
> > volume
> >     ans: Detect headphone connection - PulseAudio intend to expose
> > jack-sensing once it becomes available to them from below. Its in the
> > pipeline apparently. I need to do some more investigation around this.
> > Some devices expose individual headphone volumes as opposed to speaker
> > volumes others don't.
>
> It's not clear why you need to differentiate between headphone volume
> and the speaker volume - are you writing a volume control app that
> should be able to change the volume independently for headphones and
> speakers? You don't need jack sensing for this. Jack sensing is for
> switching the mode automatically when eg. the headphones are plugged in.
>
> The mode can already be switched manually with current pulseaudio - the
> concept is called "device ports". Sinks and sources can have multiple
> ports, and one port is always active (but some devices don't have ports
> at all!). There are functions in the client API for getting a list of
> available ports for a device, and for switching the port.
>
> When the port is switched, pulseaudio will control the correct alsa
> mixer element regardless whether the headphones and the speakers use the
> same or different mixer elements.
>
> > 3. Detect no output devices(sinks)
> >    ans: pa_context_get_sink_info_list - if the returned array is empty
> > then there are no output devices.
>
> The function doesn't return an array. It calls your callback multiple
> times, and the last call has the eol parameter set to a positive value
> and the info struct pointer set to NULL. But yes, in principle that's
> how you do it. Note, though, that the default configuration loads
> module-always-sink, which loads a null sink if no other sinks are
> present, so usually the sink list won't be empty. If you want to treat
> the lone null sink as equal to having no sinks at all (which may or may
> not be sensible), you should probably check whether there is only one
> sink present, and if that's the case, check whether the only sink's
> module is "module-null-sink".
>
> > 4. If the sound is muted and some exterior application attempts to
> > play sound then flag it
> >     ans: Subscribe to the PA_SUBSCRIPTION_EVENT_SINK_INPUT event and
> > on the callback check to see if that sink is muted (boolean). Note if
> > the client application is paused, the sink is then muted and the
> > client app is restart the event will not be triggered. This might be
> > not satisfy UI spec requirements. Check with MPT.
>
> This is correct, including the note about problems with unpausing. You
> don't get any kind of a notification when a stream is unpaused (uncorked
> in pulseaudio terminology). Maybe the cork status should be exposed in
> the sink input and source output info structs, and events should be sent
> whenever the status changes.
>
> This would still not work for applications that implement muting by
> sending silence. I don't know how usual common that is. Rhythmbox seems
> to use proper corking.
>
> > 5. Ability to mute all output sinks.
> >     ans: pa_context_set_sink_mute_by_index()- iterate through all
> > sinks returned via pa_context_get_sink_info_list().         Maybe
> > there is a mute all switch - could not see it in the docs right now -
> > will ask on #pa on freenode.
>
> Correct. There is no mute all function. Personally, I think there should
> be (perhaps not in the core API, but provided by an extension module).
>
> > 6. Unmute all outputs and return to previous volumes.
> >    ans: PA's module-stream-restore - more investigation. Apparently
> > this will store previous volumes of all sinks and restore them when
> > required.
>
> Sink volumes are restored by module-device-restore.
> module-stream-restore restores stream volumes. But your objective sounds
> misguided: muting a sink doesn't change the sink volume. Therefore
> unmuting the sink restores the previous volume even without any helper
> modules.
>
> > 7. Set individual output volumes to 0 (distinct from mute)
> >     pa_context_set_sink_volume_by_index() or by name.
>
> Yes.
>
> > 8. Set individual outputs volumes to max.
> >    ans: Set the sink (output device) in question to PA_VOLUME_NORM.
> >    pa_cvolume_reset() - should do this for the sink in question.
> > PA_VOLUME_NORM roughly corresponds to 0DB => MAX volume without
> > clipping.
>
> Use the base volume value that is stored in the sink info struct. That's
> the level that the sound card reports as 0dB. Higher values may or may
> not clip, depending whether the mixer element controls a digital or
> analog amplifier.
>
> HTH.
>
> --
> Tanu Kaskinen
>
> _______________________________________________
> pulseaudio-discuss mailing list
> pulseaudio-discuss at mail.0pointer.de
> https://tango.0pointer.de/mailman/listinfo/pulseaudio-discuss
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/pulseaudio-discuss/attachments/20100113/e979b18a/attachment.htm>