[wdmaudiodev] Re: usbaudio 4ms in Vista / Win7 myth?
- From: First Last <molley45@xxxxxxxxx>
- To: wdmaudiodev@xxxxxxxxxxxxx
- Date: Wed, 28 Oct 2009 13:21:48 -0700 (PDT)
Great... that's good news, thx DJ. I've implemented this in our own ASIO
driver, and the performance is good so far in my tests. Certainly acceptable
for playing instruments, etc. I also mentioned this to the ASIO4ALL guy since
he also supports usbaudio devices and probably has a very large user base for
testing, etc.
However, the obvious problem is : What if MS decides to remove it again in some
upcoming SP for Win 7.
One might even hope (erroneously) that the opposite might be true, and they'll
release an SP for XP and Vista with this fixed, though I doubt that's even on
the table as they're trying to get people to upgrade to 7.
________________________________
From: DJ Sisolak <dsisolak@xxxxxxxxxxxxxxxxxxxxxx>
To: "wdmaudiodev@xxxxxxxxxxxxx" <wdmaudiodev@xxxxxxxxxxxxx>
Sent: Mon, October 26, 2009 9:31:42 PM
Subject: [wdmaudiodev] Re: usbaudio 4ms in Vista / Win7 myth?
This is not a Myth for Vista Gold. The driver did send 4ms requests
in this OS release. The problem was that the extra DPC time required to service
the extra requests was deemed as a performance hindrance and vendors requested
that the original 10ms request size from XP be restored. This was done in the
ensuing service pack (I believe it was SP1). The 10ms request was selected
because it matched the engine cadence and thus was the most efficient number for
the driver to use for the most common uses of audio capture, being usually
webcams and messenger type applications. Windows 7 however, has a means to set
the request size down to 3ms or any value between 3 and 10ms. There is a new
Property that can be sent to the driver pin: KSPROPERTY_AUDIO_BUFFER_DURATION
(see MSDN http://msdn.microsoft.com/en-us/library/ee220334.aspx). The caveat
being that you will sacrifice CPU time for latency. This value is also used in
the audio engine to match the exclusive mode period to the packet request count
as closely as possible in this same range.
Thx,
DJ
This posting is provided "AS IS" with no warranties,
and confers no rights.
From:wdmaudiodev-bounce@xxxxxxxxxxxxx
[mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx] On Behalf Of Matt Gonzalez
Sent: Sunday, October 25, 2009 7:45 PM
To: wdmaudiodev@xxxxxxxxxxxxx
Cc: wdmaudiodev@xxxxxxxxxxxxx
Subject: [wdmaudiodev] Re: usbaudio 4ms in Vista / Win7 myth?
I suspect that the 10 msec number is due to both the XP and
Vista audio mixers running at a 10 msec native buffer size.
Matt
On Oct 23, 2009, at 10:07 AM, First Last <molley45@xxxxxxxxx>
wrote:
>
>>
>>
>>
>Hi,
>
>>I have an USB Audio Class device, which I've been using on XP x64 with
>usbaudio.sys and the asio4all driver. I've been using usbtrace to look at
>the actual packets to try and determine the actual latencies.
>
>>It seems usbaudio.sys always asks for 10 packets at a time (i.e 10ms of data)
>from the IN endpoint, however, when using the asio4all driver I see that
>usbaudio can send as little as 1ms at a time, which is great. I
>understand that usbaudio is still usb 1.0, so anything less than 1ms is not
>possible given that it cannot support any interval other than 1 (meaning only
>frames, no microframes for packets).
>
>>So I have a couple questions regarding this...
>
>>1) Why 10ms only for the input stream? If you take into consideration the
>round trip latency 10ms input + ~2ms audio processing + 1ms output + (whatever
>time it actually takes to go up and down the stack), that's like at least 15ms
>in a best case scenario with usbaudio, no? more? Anyone who plays an
>instrument, even someone who doesn't know how long 15ms is or how many feet
>from the speakers that equates to will tell you "that doesn't feel
>right". Why so much latency on the input stream?
>
>>2) I've heard rumors that in Vista / Win7, that it uses packets in groups of 4
>instead of 10, or 4ms at a time. How do you get this elusive
>number? I have a Win7 test machine here, and darned if it doesn't still
>ask for 10 packets / ms at a time. Is 4ms a myth? How is the
>NumberOfPackets determined? Is it determined at the kernel level? in
>usbaudio? or perhaps even in the descriptors of the device? Or is it
>enforced by DirectSound, or I suppose DirectKS or WaveRT when using ASIO?
>How do I get that magic number?
>
>>Any suggestions welcome, thanks.
>
>
>
Other related posts:
