Re: achieving very low latency
Re: achieving very low latency
- Subject: Re: achieving very low latency
- From: Peter April <email@hidden>
- Date: Thu, 05 Jul 2012 00:54:21 -0400
Or one could attain microsecond latencies using specialized hardware specifically designed with this purpose in mind; eg, DATAPixx:
On 2012-07-04, at 10:42 PM, Paul Davis wrote:
On Wed, Jul 4, 2012 at 2:28 PM, Daphne Ippolito <email@hidden> wrote:
I've tried buffers of 32 bytes, and I still get ~30 ms of delay. I
don't think I can go much lower than that? I am using an only slightly
revised version of the sample code from the Learning Core Audio book.
you need to use HALLab (IIRC) to adjust the size of the device driver "safety buffer". this may not even be possible on later versions of OS X. the safety buffer is not controlled by your application, it exists to deal with possible errors in the DLL/PLL that is used within coreaudio to estimate the current buffer position being used by the h/w.
as a general comment, the lowest measured latencies for audio come from a very well tuned linux system, which can go quite a bit lower than OS X or windows. but even there, sub-millisecond is probably impossible. 1-3 is doable but is very, very, very dependent on hardware and a carefully tuned software stack.
|
_______________________________________________
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