Re: Host callback
Re: Host callback
- Subject: Re: Host callback
- From: William Stewart <email@hidden>
- Date: Thu, 3 Mar 2005 11:49:17 -0800
Ah...
Well, I think I'll take this opportunity to remind people about the
auval list that we manage. This list is used to seed pre-release builds
of the "AU Dev" tool set. We of course seed the auval tool, but also
one of the things we've been seeding on this list is the AU Host
application (AU Lab) that we are shipping as part of the developer
tools for Tiger. This applications does implement all of these
callbacks correctly (and provides the ability to synch its "timeline"
to a MIDI source or an internal time base).
If you aren't on the validation list and would like to be, please send
me an email and we'll sort this out.
Thanks
Bill
On 03/03/2005, at 7:10 AM, Howard Moon wrote:
Hi all,
I thought I'd better chime in here about the specific example
given below:
On Mar 2, 2005, at 3:37 PM, Pavol Markovic wrote:
Hi Ralph,
here's example. The only relevant place/frequency to call callback
functions is in render/process thread, max once per render call (you
don't have to call it for every single frame, because the number of
frames to render are passed as parameter into render method/function)
... somewhere in render/process call stack...
bool haveSamplePos = false;
Float64 outCurrentSampleInTimeLine;
if ( IsTransportStateProcSafe() &&
mHostCallbackInfo.transportStateProc )
{
OSStatus result =
mHostCallbackInfo.transportStateProc(mHostCallbackInfo.hostUserData,
NULL, NULL, &outCurrentSampleInTimeLine, NULL, NULL, NULL);
Unfortunately, neither Logic nor Digital Performer currently give
valid information for the outCurrentSampleInTimeLine parameter.
You'll just get the value 0 for that. Other information should be
valid, but the time in samples may not be, depending on the host.
This has been a major headache for our software. It's strange that
the callback was implemented, but the proper setting of one of the
parameters left out. For hosts that give us 0 in that parameter,
we're relying on the (potentially much less reliable) time
information in the Render call.
if ( result == noErr )
{
haveSamplePos = true;
}
}
Please note, that currently very few hosts provide this info and
older versions of GarageBand and Logic may crash when tryin' to call
this. Find e-mail thread "TransportState Logic/GarageBand
workaround" in archives (8/18/2004).
Quite so. But you'd think that once a host (esp. one that Apple
owns!) added the callback support, they'd add *all* the information
it purports to support. (The only AU host I have that supports the
time info is Metro. At least *someone* got it right! :-))
-Howard
_______________________________________________
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
--
mailto:email@hidden
tel: +1 408 974 4056
________________________________________________________________________
__
"Much human ingenuity has gone into finding the ultimate Before.
The current state of knowledge can be summarized thus:
In the beginning, there was nothing, which exploded" - Terry Pratchett
________________________________________________________________________
__
_______________________________________________
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