Serious audio problems with Kontakt 5 & Strawberry Electric Guitar

[rangelot1]

Yeah, it looks like the D2/G2/B2 uses a tiny bit more CPU than D2/F#2/A2, but only 1% more. I feel like there might be something exasperating the problem.

I've been reading up on related issues--would you mind checking something for me? Go to Kontakt's options, and then to the "MIDI" tab. Click on "outputs". Are all the outputs set to "off"? There are a lot of note events going on in Strawberry, and I wonder if that's bogging down the system, if those note events are also getting transmitted back through the mLAN16e.

I also have a Yamaha Motif XS, but I don't have the mLAN16e card.

I will do this tonight and let you know- thanks.

Greg, all MIDI ourputs are set to off. A friend of mine has a delta 1010lt that he might be able to loan me. I'll see if I can get that hooked up next week sometime and test. this sucks!

[Gregjazz]

The reason I was wondering about the MIDI outputs was in case it's echoing MIDI data back to the Motif--that was something somebody suggested elsewhere when running into problems with the mLAN16e. Of course, we have to isolate what's causing the problem by taking factors out of the equation--which is why I'm not eager to blame it on the mLAN16e itself.

[rangelot1]

The reason I was wondering about the MIDI outputs was in case it's echoing MIDI data back to the Motif--that was something somebody suggested elsewhere when running into problems with the mLAN16e. Of course, we have to isolate what's causing the problem by taking factors out of the equation--which is why I'm not eager to blame it on the mLAN16e itself.

I've never had audio issues - like this anyway, so I wasn't too sure about the audio interface. The only reason why I am considering it is because in CUBASE the ASIO I/O meter gets pegged when this problem occurs. But I'm not sure if that's a symptom of a CPU problem. I did one other thing that may or may not help to isoloate. I started reading through NI's help document on optimizing Win7 audio. Most of what is recommended is already done on my system. But I did download a DCP Latency checker (www.thesycon.de) and I ran it while testing. No spikes occurred in the DCP tool. In fact the green bars never went above the first green line. But I also ran Windows process monitor and when the audio drop outs occur the CPU usage meter hovers between 6-14% while playing a key mash in the range above b2 until I release all keys. Then the CPU monitor jumps to 55%. But the last three individual core monitors (CPU Usage History) spike to 100%. The first one only gets as high as ~30%. Do think anything is odd with that? BTW what kind of system did you do your K5 test on. Thanks again Greg.

[Gregjazz]

I'm currently testing it on a W7 machine--it's an i7 2600K CPU, and my audio interface is a Fireface UFX (Firewire, but I use a TI-chip based PCIe card).

I also have used Strawberry in K5 and K4 on an old duo-core Intel XP machine, and it ran fine. Of course, the CPU usage was higher, but it didn't cause dropouts or other problems.

[cphoenix]

I experience the same problems with my large sample library. In particular, Mojo Horns exaggerates the issue the easiest.

I use an Echo Audiofire 4 which has exceptional drivers (best below an RME). HD is 7200RPM, 500GB. 16GB DDR3 1333 RAM. The only thing me and the OP share is the AMD processor. However my system is run on a 3.2GHz AMDPhenom II, quad. I have it overclocked to 3.8GHz running with no hiccups.


I should not be having these problems lol. I noticed that the issue arrises when the number of voices starts to get above 300-400. That's when the cracks take place. Any instruments that cannot get the voices that high do not cause the cracks in playback. However, when loading a Mojo Horns multi of 3 horns, and setting them at 10 players per horn, u can see how the voices multiplies.

Shouldn't my setup be able to handle that many voices? I'd assume so.. b/c if not... why have they made an instrument that can't be played?

I've tried almost every way I could think of to limit the voices. Nothing has worked. {I'm not one of those guys who is quick to blame a program, but accidentally has a setting wrong lol.. i've been troubleshooting this problem for 4 months now, almost nonstop.) I've lowered the individual max voice settings on each instrument. I've increased buffer sizes for preloading. I've played with the multicore settings. I've played with the CPU overload protection options. I've placed my library on an external harddrive. When I tell you nothing works..... nothing works. I've done all of thisin StandAlone mode as well. Same issue. Something is seriously wrong.


Try to find an instrument that will get your # of voices to go above 500, and I guarantee you will see the same issues with your i7. Please please test this theory for me if you can. I'm really trying to get to the bottom of this. I'm sure there's many people out there who have this problem.

[EvilDragon]

Well, that IS a LOT of voices, after all.

[cphoenix]

Any ideas on how to cut it down?

Took me forever to realize that has to be the problem.

The horn multi sounds so unbelievably rich, but its unplayable.