What I think about the Spark PC problem

[jimbobaggins]

Hi all , Kevin , Emilie

I have noticed in this forum there a lot of problems with Spark and PC users.

Most of these problems relate to high CPU usage and stuttering because of load.

These problems are usually expressed in terms of, I have this system and I can run X software without problems.

There seems to be very few people who can pinpoint what is happening in their system when their problems occur.

For example I was using a system in XP with 1Gb chips (4GB total) and had problems when the memory reached around 1.2GB being used
I changed these for 2GB chips (still 4GB total) and only had problems when memory reached around 1.7GB

Certain kits in Spark cause more problems than others.

When stuttering starts in a pattern I can press stop and start and the stuttering will go away until I have selected a number of patterns.
What number is not precise, but seems to indicate some buffer is being filled.

TSR (Terminate stay resident) programs have an effect. NIHardware, AntiVirus, Network all seem to have an effect. So I disconnect from the net and terminate all unnecessary programs using task manager.

I dont know whether OSX uses on-demand calls for support programs or if it uses TSR as Windows does.

I have recently upgraded to a PCIE soundcard and am using a unified driver which has less TSR stuff associated with it and the problems do not occur as much.

Anyway these are just some of the things I have noticed.

My suggestion to anyone having problems is to unplug any external hardware (controllers) terminate any drivers for them using task manager (dont worry they will come back next time you reboot) turn off the internet access and your antivirus, even disable your network card.

So basically you just have your bare machine with your soundcard/audio interface and try and run Spark, see if it still has a problem. (when I get to this point Spark runs fine for me.)

Then add in each of the above one at a time keeping a watch on your task manager  to see if it is a highend (running too many programs on the cpu, shows in amount of memory used) or a lowend (to many programs are trying to use too many resources, like direct sound to ASIO, shows in processes)

Personally I believe it is the latter but I am not the dev so I dont know how the port was done.

Anyway that's my 2c worth.

Baggins

[rodanmusic]

My suggestion is to get the power supply that is suggested in the manual. 

Not sure if that's what is happening to everyone else, but when I got the power supply most of my issues went away.

[jimbobaggins]

Hi all

I don't think the answer is in buying the power supply as the issues I have happen without the controller plugged in.

As suggested in the promo literature I am using a different midi controller to perform live.

A little more investigation shows me that converting a kits samples to 16-bit alleviates the problem a little.
(In other words it takes a bit longer before the stuttering and distortion occur)

So once again this leads me to believe it is a processing buffer problem in windows.

Although the problem can occur immediately upon starting the software it generally takes between 7 - 10 patterns before it starts
and as I said above if I am on the ball at the first click of distortion (a bit hard with D&B) I simply click the on and off on the beat and it goes along smoothly.

Still not an answer but hoping my observations may shed some light.

Baggins

One question I did have was as to whether the samples are loaded or re-loaded on each pattern change.

[jimbobaggins]

So the problem I was having in EDM was that on certain kits eg Basswave after selecting particular patterns (A7 and A8 in Basswave) by playing other patterns and then selecting the problem ones again. The system would start to get crackling distortion and playback would slow down.

So I copied the problem kit over to Spark CDM Library and tried to play it there.But I encountered the same problem?

Then I went to the bottom and looked at the sample window for each instrument and decided to set the length (moveable yellow line) to where the sample ended instead of all that blank space. I then saved the kit as a user kit and played it.

Surprise Surprise No Problem. 

So I copied the kit back to the user library of EDM and played it there and once again, NO Problem!

I am not sure what is going on here but I am now going looking at all the other kits, that might have an issue and see if giving them the same treatment will fix them.

Will keep you all advised.

Baggins
 

[Philippe]

Very interesting remark ! Thanks. I'm looking at it right now.