Re: How to make my AU graph the default output?
Re: How to make my AU graph the default output?
- Subject: Re: How to make my AU graph the default output?
- From: Ondřej Čada <email@hidden>
- Date: Sat, 28 Mar 2009 18:54:47 +0100
Jens,
On Mar 28, 2009, at 5:19 AM, Jens Alfke wrote:
But dammit all -- this inexplainable limitation makes the whole
Core Audio pretty much unusable! What the H did they think of
designing the stuff?!? :-O
Unusable for your specific application.
Quite; also, I am afraid, for any other application whose sound output
(a) was designed without Core Audio (be it since its capabilities were
not needed before or any other reason), (b) can't be easily completely
re-written now, (c) or for any reasons needs QuickTime.
Yours is not the only type of audio application. In fact I've never
heard anyone else make this specific complaint before.
Frankly that surprises me considerably -- I would suppose there'd be a
number of applications which fall into the aforementioned categories.
Perhaps there is not... or their writers did not even try to improve
the audio using CA :)
What does this app do? What API does it use to output the rest of
its audio?
It's a carputer software (http://www.qcar.nl). Does lots of things,
from playing music and movies (QT needed here) through diverse
warnings (often beeps, sometimes NSSound or QTMovie used for playing
simple sound files) to navigator (which speaks to the user through
NSSpeechSynth). It also offers an extendable plugin architecture for
modules some of which may use other sound APIs (including CA itself,
though, far as I can say, this never happened yet ;)).
Works for years in stereo.
Now, the need for Core Audio (or something similar) came when some of
our clients wanted the stereo output split to four channels normally
found in cars (front/rear pairs), along with a few special
considerations (like the parking assistant beeping from the direction
where the obstacle is, which should be easily achieved with the 3D
mixer, or so I think).
(What I actually need is simply to process the stereo sound to make
it multi-channel based on comparatively plain rules -- at the
moment I am pretty positive the Matrix Mixer AU's all I need, in
future I'd add some extras based on 3D mixer.)
It sounds like you could do this with AudioHijack (by Rogue Amoeba),
or WireTap Studio (by Ambrosia Software). These programs can take
the audio from any app, run it through AudioUnit effects, and send
it anywhere.
At the moment I am checking sunflower; preliminarily, it seems I
should be able
- to set up the system default output to Sunflower's (for far as I
understand, QT can't render sounds to anything else than the system
standard output);
- re-direct it to my auxiliary "sound processor" application which
will be run automatically and GUIless;
- it would use the CA graph to process the audio and send it out
through the real output device (depends on the client, but most often
this is an aggregate of the built-in output and something like an iMic
extra two channels).
It's actually rather ugly solution, but seems preliminarily to be
doable with a reasonable amount of work...
(Yes, Bill said this wasn't possible. These apps do some unorthodox
things like installing kernel extensions, which Apple isn't going to
condone or recommend that people do...)
It's not that big problem for us, since by the very nature of a
carputer we have more or less complete control over the environment;
still I don't like a KEXT-based solution and would much rather use a
cleaner one, if existed.
I still can't get the reasons why there is no "default output AU" to
be put in the graph: seems to me as clean solution as possible...
well, never mind :)
Thanks a lot,
---
Ondra Čada
OCSoftware: email@hidden http://www.ocs.cz
private email@hidden http://www.ocs.cz/oc
_______________________________________________
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