Hi All -

I have a second machine I've performed this test on, a Pentium4 2.4 Ghz, WinXP Professional SP2, and I'm still getting 500 ms latency with a Tascam US-122 USB interface.  The setting in the ASIO control panel is 256 samples, and here Csound is not forcing the control panel to the highest option, as on my other machine.

I think with both these machines there is an issue with portaudio, asio, and maybe my understanding of how PA and ASIO work together, so I'll do a little research and hopefully come up with a fix, or at least more helpful info for anyone who can help me. :)  I assume someone else has seen this since I've seen it on both these machines.


----

One thing I wanted to ask the list: I'm have Cubase SL/SX installed on each of these machines - is anyone else out there running Cubase SL/SX and running csound from the cmd line (i.e. NOT CsoundVST)?  Your results may be consistent with mine.

In case I didn't mention it before, I get little or no latency within Cubase using the ASIO drivers for each device.  Does that maybe indicate a Portaudio issue?  Hmm.

Thanks for any help - b

 

On 1/5/06, Ben McAllister <benmca@gmail.com > wrote:
Update - what's happening is that csound is resetting the M-Audio's input latency to a very high 2048 samples after I set it myself to 64.  I didn't have much time to spend on this last night, but will explore more over the weekend.  Ring any bells for anyone?  Thanks in advance - Ben

PS - I am 100% positive I am passing csound the correct ASIO driver # for input and output on the commandline.


On 1/3/06, Ben McAllister < benmca@gmail.com> wrote:
Thanks for the feedback everyone - I'll need to stop by my studio tonight and check it out.  FWIW, the M-Audio ASIO driver performs well in other applications (Cubase, Logic), so I figured there was no need to check the control panel, _and_ that behavior would be consistent across apps.  One caveat - I'm debugging this in the context of a larger instrument (my vcs3 emulator, if anyone remembers me posting that).  Since it would be more helpful from a debugging perspective, I'll make a stripped-down test - say, a single instr containing only a simple in/out combo - and see if the results are consistent with what I'm seeing in my vcs3 instrument.
Thanks again - b


On 1/3/06, Michael Gogins < gogins@pipeline.com> wrote:
With Csound 5 and ASIO in the past, I have been able to get 10 to 15 ms delay without dropouts, 5 to 10 ms delay with some dropouts. I will be checking this again soon.

-----Original Message-----
>From: Ben McAllister <benmca@gmail.com>
>Sent: Jan 3, 2006 1:17 PM
>To: csound@lists.bath.ac.uk
>Subject: [Csnd] realtime input latency
>
>Hi All -
>
>What is the expected behavior of realtime input for the 12/13 build of
>csound5?  I'm running a 2.4gHz Win2k pc with 1.5 G Ram, M-Audio 1010LT
>soundcard.  Using the in, ins, and inch ugens, I'm getting ~300ms delay with
>the ASIO drivers included with my card at sr=44100 (this even with kr=441).
>Rt output is great, and RT MIDI input is good as well.  I will do more
>experimentation, but thought I'd also do a status check with the list, in
>case there is a known issue.  Thanks!
>
>Ben McAllister
> http://www.listenfaster.com



--
Send bugs reports to this list.
To unsubscribe, send email to csound-unsubscribe@lists.bath.ac.uk