• Open Menu Close Menu
  • Apple
  • Shopping Bag
  • Apple
  • Mac
  • iPad
  • iPhone
  • Watch
  • TV
  • Music
  • Support
  • Search apple.com
  • Shopping Bag

Lists

Open Menu Close Menu
  • Terms and Conditions
  • Lists hosted on this site
  • Email the Postmaster
  • Tips for posting to public mailing lists
RE: buffer size magical reduction
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: buffer size magical reduction


  • Subject: RE: buffer size magical reduction
  • From: Stéphane Beauchemin <email@hidden>
  • Date: Mon, 29 Nov 2010 08:20:13 -0500
  • Thread-topic: buffer size magical reduction

4096 is the default buffer size when you are in the default audio category for you Audio Session (default is kAudioSessionCategory_AmbientSound when you don't specifically initialize it).

Try doing the following:

AudioSessionInitialize ( NULL, NULL, NULL, NULL);
UInt32 sessionCategory = kAudioSessionCategory_SoloAmbientSound
AudioSessionSetProperty (kAudioSessionProperty_AudioCategory, sizeof(sessionCategory), &sessionCategory);

If you are in the mode kAudioSessionCategory_AmbientSound, you won't get requested buffer size because you not the only app talking to the hardware. With kAudioSessionCategory_SoloAmbientSound you should get the requested buffer size unless you are asking for something really small. 2048 should be ok, I was able to successfully request 1024.

Side effect of this -> other apps will be muted :-)

I hope this helps,

Stephane


-----Original Message-----
From: coreaudio-api-bounces+sbeauchemin=email@hidden [mailto:coreaudio-api-bounces+sbeauchemin=email@hidden] On Behalf Of David Plans
Sent: November-28-10 4:40 PM
To: Paul Davis
Cc: email@hidden
Subject: Re: buffer size magical reduction


On Nov 28, 2010, at 3:43 PM, Paul Davis wrote:

> you get something supported by the h/w and other layers of the
> CoreAudio API; you can find out its value as described by the quoted
> sentence you mentioned.

But the thing is, we consistently get _half_ of the buffersize we ask for...that is, no matter what we ask for, we get exactly that, I can't see why.

Surely, if it had to do something with hardware supported, there would be some immovable low/high limit that I'd hit when I set it very low or very high?

Which isn't the case, it just gets cut in half.

> btw, PureData doesn't "have" a blocksize per-se that matters in this
> sense.. Its quite happy to adapt to the blocksize imposed by outside
> factors, like the audio h/w, OS audio API etc. etc.

Well, by 'have' I meant that Miller Puckette wrote:

#define DEFDACBLKSIZE 64

into s_stuff.h and that traditionally, that's PD's blocksize

 _______________________________________________
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


 _______________________________________________
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

References: 
 >Re: buffer size magical reduction (From: David Plans <email@hidden>)

  • Prev by Date: Re: Generating pretty sounding Shepard Tones
  • Next by Date: audio priority and multitasking on iOS
  • Previous by thread: Re: buffer size magical reduction
  • Next by thread: gcd/asynchronous task and Core Audio
  • Index(es):
    • Date
    • Thread