Hello-
i just discover that the windowsXP firewall does filter all message from pd (0.38) to it's gui. this can cause lot's of click or other trouble in the audio processing.
Where are the clicks happening or are likely to happen? I've been running PD .37 up to .38 on XP machines with scarcely a problem.
I do notice that Norton and the Windows firewall attempt to block communication but I have told the firewall to Unblock PD. It prompted me upon opening PD recently.
Again... all my patches have been click free from quite some time (even the hairy granular patches). I still haven't gotten into working with my own FFT patches but I suspect that maybe this is where the click could happen? Really I'm clueless as to what the trouble is all about.
I will note that while using the Asio Full Duplex Device that I will get an occassional click or pop here and there but that's most likely due to the driver (worst when using the multiedia device so don't do it). When I'm using my MOTU interface, I haven't had a problem since I downloaded the Windows firewire update.
Also, clicks and other distortion can occur when trying to bring the latency down too far (below the buffer size on your audio card). I'm not sure what the best advice here is though. Perhaps drop the buffer size on the audio card down and then reduce latency to just longer than the buffer size? A 1028 sample buffer is equivalent to just under 24 ms (44.1). I run at about 20 ms latency with 512 sample buffer (44.1) just to be safe. I suppose I could get the latency down smaller say 10 ms by using a 256 sample buffer but I haven't tried that yet.
PS : use winXP without firwall at your own risk.
yeah just make sure that you're not connected to the net (physically disconnect or power off network adapters) and that you've done a virus and spyware sweep
Phil
Do you Yahoo!? Yahoo! Small Business - Try our new Resources site http://smallbusiness.yahoo.com/resources/
hi all,
having a firewall will affect the pd/gui communication as long as a network socket is used. having a network socket will be a low latency issue unless the network communication is done in a separate thread ...
the best advice here is though. Perhaps drop the buffer size on the audio card down and then reduce latency to just longer than the buffer size? A 1028 sample buffer is equivalent to just under 24 ms (44.1). I run at about 20 ms latency with 512 sample buffer (44.1) just to be safe. I suppose I could get the latency down smaller say 10 ms by using a 256 sample buffer but I haven't tried that yet.
the current implementation of pablio (that's used for asio / coreaudio) is really bad ... i haven't been able to run safely below 25 ms of latency ... in devel, there is an implementation of a callback-driven scheduler, that directly talks to portaudio without using pablio, giving MUCH better latencies ...
if you need lowest latencies, i'd suggest, you try devel :-)
cheers ... tim