Re: AppleUSBAudio Driver Latency
Re: AppleUSBAudio Driver Latency
- Subject: Re: AppleUSBAudio Driver Latency
- From: Bill Stewart <email@hidden>
- Date: Tue, 13 Nov 2001 19:19:28 -0800
USB will add additional latency over PCI for two reasons...
Firstly - due to the innate architecture of USB in and out streams
are clocked differently so it is not possible to present a single
I/O cycle for a USB device - this can be worked around by having
say the output side of the device drive this and use the
AudioDeviceRead calls to read the appropriate data from the input
device... PCI and Firewire do not have this limitation
Secondly - USB itself has a transport latency of a msec per packet
and from memory I believe the safety offset for USB is an
additional msec - so that's 2msecs for input and 2msecs for output.
FW and PCI buses can operate with a very low latency in transport
and safety offset. (Safety offset is how close the HAL can get to
the read/write heads of the driver)
All of this information should be reported by drivers and can be
ascertained by an application allowing a good app to take account
of the latency:
I/O processing buffer size
+ Safety Offset
+ Hardware Latency
== how long your sample takes to get heard (or how long it took
your sample to arrive to you)
Bill
On Tuesday, November 13, 2001, at 04:32 PM, Sean Morrell wrote:
What is the round-trip latency of audio through OS X with the
AppleUSBAudio driver? How how does this compare to audio devices
on other
buses such as PCI? Is the latency known for the Audio HAL, the driver,
and other components separately?
Thanks,
Sean
_______________________________________________
coreaudio-api mailing list
email@hidden
http://www.lists.apple.com/mailman/listinfo/coreaudio-api
mailto:email@hidden
tel: +1 408 974 4056
__________________________________________________________________________
Cat: "Leave me alone, I'm trying to nap... Ah, What's that clicking
noise?"
__________________________________________________________________________