Beiträge von malfunct

This seems to be caused by USB, indeed.

Try routing to Analog and see if the problem recurs, if it does, it's caused by USB buffers running out.

The reason for this is many. On OS X, the primary cause using Live seems to be hyperthreading. Disable hyperthreading. You can also try disabling power saving options. I've got a core i7 processor, and this solved the same problem. You get in, it works for 10 minutes, and then distortion starts.

There is another strange thing with live. I've noticed that whenever CPU usage goes above 50% in Live, the Virus sync/crackling/FUBAR errors start. This MIGHT be due to a bad USB chip but I don't have high hopes for that. I think that has more to do with the kernel. When you're running Virus, try to run a program that accesses disk frequently. That will also cause Virus crackling/sync errors.

Somehow USB transfer doesn't have high priority. I don't know if this is the same in OS X 10.6.x. But I'm pretty sure there is a strange behavior in OS X 10.7.x.

I also completely disabled the dynamic pager! That works very well, and it doesn't break memory if you have enough RAM.

I am still seeing those annoying keThread::run - delay messages on OS X.

I am right now having to configure asl.conf so that debug level messages from the kernel aren't logged. Not a real solution to the problem.

Please remove them from the release version of the driver. They are annoying, take megabytes of space on the disk, and possibly slow down the system. As you know, syscalls don't come for cheap, so why do you make so many syscalls?

UPDATE: solution here was ineffective

Here is how my DAW experiment went. Problem recurs, but I've *found* that KeThread::delay messages happen only when there is some delay caused by the OS. This is mostly due to page faults. So if your system is swapping, you'll most likely lose Virus TI connectivity.

The problem is that on Lion, the system will swap even if there are GBs of available memory. It reserves what it calls "Inactive Memory". Good job apple!

I have no idea if that's fixable. It probably isn't without making the whole DAW run with real-time priority or something like that. Because obviously DAW will have to wait for kernel induced delays like page faults, and then buffer will be depleted.

It'd be nice if the DAW could somehow pin all the physical RAM it needs and be immune from page faults, but I doubt that DAW designers think that far. OS X 10.7.x has a rather ugly memory policy, that would probably make such attempts futile.

Good call. The konnekt 6 interfered with some silly daemon a plugin lib. installed and that's what mostly caused the stuttering. Now that the main reason for the stuttering is gone, I can try some recording and let's wake up the neighbors with a bit of practice =)

The bottom line is: be wary of any sound related daemons on OS X, they might have nasty consequences. Make sure you add that as an OS X FAQ? On OS X, it turns out it's also quite important to disable unnecessary daemons like the daemon that updates google chrome. I can now get OK playback with the konnekt, I'm testing just playback for a few hours, no DAW. I'll try again and report the results.

One thing for all to check: if there are several keThread::delay messages in kernel.log, you have a low-level problem.

I'm using the same USB port I used to use before. It's not shared with any other USB device, I checked it on the System Information utility. Which shows that the Virus TI is attached to a single USB Bus (with a unique PCI Device ID).

The number of notes played is small, I'm right now trying to record a single melody!!!

I'm using Desktop Konnekt 6, which also has some funny behavior from time to time. It's quite irritating how unstable an audio configuration can be.

IOW, changing the USB port won't help.

I'm using Live on OS X 10.7.2 and Desktop Konnekt 6. Buffer size 512.

Simple setup with 1 preset seems to work.

In a somewhat complex project I have 4 presets. I'm trying to record just one of the presets. I get constant popping and cracking. I'm trying to output to USB2L+R.

CPU never seems to max out. But I still get the sync errors.

What must I do now? Try to bypass USB and use analog out instead to see if it's a Virus problem or a USB problem?

Please help. It's driving me crazy, because this project used to work. I can't record any of the virus tracks now.

When I plug in my Virus TI Snow, I see the following messages on my kernel.log on OSX 10.6.7, both in 32-bit and 64-bit modes. I was using the 4.1.105 version. What can I do to correct these messages? They seem to be filling my kernel.log and slowing down my system!

Jun 23 15:03:38 orion kernel[0]: KeThread::run - delay:2803us
Jun 23 15:03:38 orion kernel[0]: KeThread::run - delay:2469us
Jun 23 15:03:39 orion kernel[0]: KeThread::run - delay:2189us
Jun 23 15:03:39 orion kernel[0]: KeThread::run - delay:2379us
Jun 23 15:03:40 orion kernel[0]: KeThread::run - delay:3916us
Jun 23 15:03:40 orion kernel[0]: KeThread::run - delay:1571us
Jun 23 15:03:41 orion kernel[0]: KeThread::run - delay:3281us
...