20+ Ins and Outs - What hardware works for you?

I have found that the MOTU PCI424 with a 2408, 1224, and 24i, with most ins and outs enabled in the MOTU control panel, does not perform as needed. I cannot get glitch-free operation while trying to monitor at low latencies. CPU usage just will not stay down enough, and I get clicks on occation corresponding to MIDI activity. As best I can tell it is the MOTU drivers, as I have tried the same scenario in 2 computers with exactly the same result.

So, is there anyone having good experience where you are able to operate and monitor with 20 or more ins and outs active in the hardware driver? If so, what hardware is allowing you to accomplish this.

I just spent $800 on a new computer system thinking that my problem would be solved. Obviously, it's the MOTU drivers -- either the audio ones or midi ones, or both.

Thanks Jim.

What MIDI interface, if any? The problem shows up randomly when midi messages are received.

I find it very odd that a completely different system does exactly the same thing. I did find, however, that I can overclock my original 3.0GHz pentium to 3.75GHz, and that does ease the issue to a large extent, but it doesn't completely clear up. The occasional click still occurs during midi activity.

In addition, the same thing occurs with a PCI 324. I'm running on 32bit Windows XP on both systems, Home edition on the Pentium.

Jim,

The motherboard on the new system is an ASUS AZRock, http://www.newegg.com/Product/Product.aspx?Item=N82E16813157132. On the older Pentium, it is an ASUS P4C800E Deluxe.

I have tried MOTU driver settings at 32, 64 (my typical), and higher. I am going to do some more experimenting with these setting, but 64 or better is my aim.

As for the HD, I don't think that is at issue here. The activity shows nothing, and I'm not playing or recording tracks, just playing through with my midi keyboard and monitoring through Sonar.

Since I've just recently gotten this new system with the AZRock, I'm going to try some more tweeking with it and see how that goes. At this point, the older Pentium system is doing best of these two, being overclocked to 3.75GHz.

I have to wonder, if it's not the PCI424 driver, perhaps it is something with the MOTU MIDI drivers (MIDI Express 128, the parallel interface is nonfuntional on this system at this point). After all, the midi messages are causing most of the grief, it seems. MOTU MIDI Driver version 2.51.

I would love to talk to MOTU about this, but I can't imagine that happening.

Thanks.

Hi Jim,

Sorry, I think I am confusing the "latency" that you are referring to, and thinking about the MOTU driver buffers (?) setting. I was looking in the Bios settings on this new AZRock, and "PCI Latency" is set at 32. Thanks for that tip. I have made some other changes, so if those don't help, I'll try that next.

That is the PCI Latency to which you refer, right?

Thanks.

Hi Jim,

Well, after doing some adjusting in the BIOS I think I have something working. Thanks to that review of the motherboard on the Newegg site, I discovered that I needed to manually set the bus speed at 1066, and I think that could be the most important thing I have changed. I also turned off the "Cool and Quite" option in the BIOS, but I don't know if that had an effect. There were several things I wanted to change, and I did them all at once, but mostly the bus speed.

I did try changing the PCI Latency setting, but I didn't notice any difference. Of course, that was after I had upped the bus speed and things had seemed to work better. Any idea how that PCI Latency should affect the system in general?

At this point, it might actually be that I have a system I can use (except I need my parallel Midi Express XT to work now). The only possible thing is that I was noticing the occasional random pop in the audio (just playing my midi keyboard and monitoring the audio through Sonar). But, Windows was using my MOTU for the system sounds. I disabled the Windows system sounds, and installed the drivers for the onboard sound device so Windows could point to something other than my MOTU.

So far so good. I haven't been able to torture test yet, so we'll see.

Thank you very much for your response.

ORIGINAL: Junski

ORIGINAL: lahatte

...

I have tried MOTU driver settings at 32, 64 (my typical), and higher. I am going to do some more experimenting with these setting, but 64 or better is my aim.

...

Thanks.

If you mean the audio driver latency buffer setting for record/playback as samples ... 32 samples @ 44.1kHz equals ~0.73ms and @ 96kHz ~0.33ms latency ... etc.. Are you sure it's necessary to go this low (I can't notice difference between 2ms and 5ms)?

Junski

5.8 is ok for organ for me but piano 2.9. But I never got my P4 to go to 2.9. I just upgraded and Jim was a big help.

I've found that soft synths can have issues where CPU usage keeps spiraling up if they get "overworked". Could be a specific issue with a specific synth.

As for the mother board, I found the same with the P5-K Deluxe I have. Changing the BIOS setting from Auto to the appropriate FSB speed made a huge difference.