7 Oct
2012
7 Oct
'12
10:03 p.m.
BODY { font-family:Arial, Helvetica, sans-serif;font-size:12px;
First, what I would like to do is pick which channel I send audio to
on the RME 9632 card.
My RME card is device 0.
When I call this function:
snd_pcm_open(&handle,device,SND_PCM_STREAM_PLAYBACK,0)
And set device = plughw:0,0 I get playback across all the RME
outputs.
(Just using hw:0,0 without the "plug" prefix does not work for some
reason which may or may not matter.)
So, how do I access the subdevices/channels?
When I run aplay -l I only see 1/1 subdevices.
Shouldn't I be able to set my device to hw:0,0,1 or plughw:0,0,1 to
access channel 2, for example?
There's no .asoundrc. Do I need it for this?
Finally, when attempting multichannel alsa, I assume I need to
simply make an array of pointers to the same datatype as "handle" and
run through
initializing each handle separately.
In the first time thought the loop, I can get handle[0] and all its
parameters to initialize without errors.
When I try to initialize handle[1] on a new device such as
plughw:0,0,1 I get an error that the device or resource is busy.
I feel like I've hit a brick wall here. Perhaps I am out of my
league. But, it feels tantilizingly close to working.
Thanks guys/gals!!
Grekim
8 Oct
8 Oct
12:40 a.m.
On Sun, Oct 7, 2012 at 4:03 PM, <grekimj(a)acousticrefuge.com> wrote:
First, what I would like to do is pick which channel I
send audio to on
the RME 9632 card.
My RME card is device 0.
When I call this function:
snd_pcm_open(&handle,device,SND_PCM_STREAM_PLAYBACK,0)
And set device = plughw:0,0 I get playback across all the RME outputs.
(Just using hw:0,0 without the "plug" prefix does not work for some
reason which may or may not matter.)
with ALSA, devices named "hw:..." can only be configured in ways that match
the hardware capabilities of the device. no configuration that requires the
driver or libasound to resample, reformat etc. etc. are allowed. note that
this includes the number of channels, so if the h/w has N channels you
cannot configure it for anything other than N channels.
"plughw:..." devices are much more flexible and will use the full
capabilities (such as they are) of the driver and libasound to provide a
much broader configuration space.
So, how do I access the subdevices/channels?
When I run aplay -l I only see 1/1 subdevices.
Shouldn't I be able to set my device to hw:0,0,1 or plughw:0,0,1 to access
channel 2, for example?
multichannel devices are multichannel devices. you don't open a different
device to get to different channels. you open it, and get a device with as
many channels as you specified. they will start at physical channel zero,
unless you use a psuedo-device (typically defined via an ~/.asoundrc file)
that maps physical channels to channels in the pseudo-device.
the way in which you access each channel depends on the configuration
parameters you use with the device. all the code i write uses
non-interleaved format, so that each channel's samples consist of a
contiguous block. some people prefer interleaved.
if you are just trying to write a functioning application rather than learn
to use ALSA, you will find it MUCH MUCH easier to just use JACK. if you
want to learn ALSA, be prepared to do a lot of reading of (a) the rather
non-explanatory reference documentation (b) finding examples of other
source code that uses ALSA in this way (e.g. JACK).
--p
10:52 a.m.
On Sun, Oct 07, 2012 at 06:40:17PM -0400, Paul Davis wrote:
if you are just trying to write a functioning
application rather than learn
to use ALSA, you will find it MUCH MUCH easier to just use JACK. if you
want to learn ALSA, be prepared to do a lot of reading of (a) the rather
non-explanatory reference documentation (b) finding examples of other
source code that uses ALSA in this way (e.g. JACK).
Or use zita-alsa-pcmi which is more or less the equivalent
of Jack's ALSA backend in an easy-to-use C++ class (original
versions were actually based on the Jack code). All you need
to add then is a RT-thread that takes the place of the one
normally created by Jack. Use by jaaa, japa, aeolus, zita-
ajbridge, and others.
Ciao,
--
FA
A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)
4688
days inactive
4689
days old
linux-audio-dev@lists.linuxaudio.org
2 comments
3 participants
participants (3)
-
Fons Adriaensen -
grekimj＠acousticrefuge.com -
Paul Davis