[PD] portaudio compatibility with some ASIO drivers (Windows only)

[Stefano Papetti]

Thomas Grill ha scritto:
>> - the RME ASIO driver simply doesn't work, and nobody actually knows why...
>> However some people report to get ultra-low latencies (3ms) with the MMIO
>> drivers. Moreover, even with the MMIO driver they can use the multi-channel
>> capability of the RME only when set at 48000 Hz (but this is another
>> story...). I will ask RME directly about these issues.
>>     
>
> I can say that i've never had problems with the RME drivers,
> especially ASIO with various applications.
> If there are problems i suspect them to be related to some old
> portaudio code used in Pure Data.
> Up-to-date portaudio versions work flawlessly with RME drivers.
>   
I don't know many other portaudio-based applications, but I can say that 
the latest Audacity also can't properly see the Fireface ASIO drivers... 
Of course native ASIO applications (e.g., Ableton Live, Pianoteq, 
Reaper) work flawlessly.
As for pd-extended 0.40-3, 0.41-4, and pd vanilla 0.42-5 the Fireface 
driver report:

"Fireface Analog (1+2) MME" ... up to "(9+10) MME"
"Fireface Analog (1+2)" ... up to "(9+10)"

so I thought that the latest were reffering the ASIO driver, even if no 
explicitly mentioned. Indeed pd allows to select a pair (loosing the 
multichannel feature...) of ins/outs from the last line as ASIO drivers, 
however they don't work unless I set latency to 40ms... I tried on two 
different machines - a desktop and a laptop - using the latest RME 
drivers (v2.95) with identical results.
Also, Audacity reports the same list of devices above plus a prefix 
"MME: ", a nothing concerning ASIO.
Finally, I often use asio4all with a pc's internal audio card in order 
to get decently low latencies, however both pd and Audacity can't see 
the virtual driver ASIO4ALL while the Fireface is connected, i.e. as 
soon as it is disconnected the driver ASIO4ALL shows up again...
Well, I guess I'd better post this on portaudio's mailing list.
> Concerning MMIO i can hardly imagine how people can reach low
> latencies with it, given the architecture of this interface.
>   
I agree... I tried myself and I can't go below 40ms.

Best,
S
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20090702/1d47f73d/attachment.htm>