When setting up a vsti, like Garritan's Aria Player, or East West's Play, I have tended to use just the first putput. (Excuse me, that should be output - sometimes spell-check mistakes are just too funny to delete). Now, after all these years, it dawned on me that this might not be a good idea. Like maybe the sounds get too intermixed and create mud, which I admit my pieces have too much of. So, should I be using more of the outputs, in the hope of producing a better mix? I know this is really basic, but sometimes we develop habits that we never question, until one day an epiphany occurs (didn't need spell-check for that one - I used to be a poet). Like, for example, it suddenly occurred to me that the reason why East West S-O basses cause my Sonar to crash, is because they are on the same putout (there I go again) as a bunch of other instruments, and maybe this problem would go away if I  gave them their own outlet (I'm just going to stay away from that word, I can't type it for the life of me).  So, any thoughts? Please try to resist the urge to express shock that I don't know this by now.

I believe you should use multi outs as much as possible. It will give you more flexibility in the arranging and mixing phases of your projects. Great question!!!

The benefits of using multi outputs is that you can mix those "voices/channels/etc" inside Sonar instead of inside the plugin. That means you can use the faders inside Sonar on each part and add Sonar FX (or any other FX you have installed) on the individual parts.
 
So that gives you some extra control over the individual sounds BUT if you are merely using the raw sounds inside the plug and it has it's own internal mixer then that stereo out essentially just acts like a bus (so like if you took all those multi output channels then sent them to a bus it's essentially the same thing).
 
Essentially if you do not intend to add any Sonar or third party effects and your plugin has an internal mixer that you are comfortable working with (instead of doing the mix in Sonar instead) then your single stereo output merely acts like a bus. This could be much simpler if you don't need all that extra control.
 
As far as crashes? I seriously doubt using multi output to individual channels in Sonar would minimize that. If anything the extra tracks would increase the resource consumption of the project (especially if you start adding effects).
 
I personally always use multi outs on my drums and other such instruments because I prefer mixing in Sonar and having the option to have the Prochannel strip/FX rack available but it all gets dumped down to a bus anyway so if those weren't factors I could very happily just use a single stereo out.
 
I hope that makes sense. Definitely play around with it though to see whether the multi out option works better for you.
 
Cheers.

I like using multi out for the drum VSTi too.  I've done a couple of mixes with the console emulator on all tracks and I swear it sweetens things.  If it doesn't, please don't tell me.  :)

michael diemer
...until one day an epiphany...

 
Gesundheit.

michael diemer
 ...the reason why East West S-O basses cause my Sonar to crash, is because they are on the same putout (there I go again) as a bunch of other instruments, and maybe this problem would go away if I  gave them their own outlet

No. If your East West S-Ob's are crashing Sonar, there is no way this is the cause.

Yes so much more flexibility to be had from using multiple outputs, especially in terms of being able to clean up the mud you're referring to. And creating depth. 
 
Another thing to consider is that maybe your CPU's cores will be used more efficiently by processing all of the tracks separately with their own effects as opposed to processing them with the VSTi's own onboard effects unit, which may or may not spike one of your cores. This is something I have noticed with Native Instrument's Drummer libraries - it has its own internal mixer with effects processing, but it tends to spike one of my cores like crazy if I use it. However I also have all of the effects it uses as separate VST's and if I route all of the drums to multiple outputs and process them separately using those same VST's, I don't get the spike. 

Always better to put instruments within one vsti on separate outputs. Maybe your spell- checker was encouraging that with the technical term putput ;-)

I usually select the single output for VSTi's that are drumkits or have multiple instruments like orchestral plug-ins. Then come time to mix, I solo each drum/instrument in the plug-in and bounce it down to an audio track. I like seeing the waveforms while I mix. I know there's a waveform preview for the multi-out VSTi tracks, but I prefer the audio track. It's a little more work, though, for sure.

sharke
Another thing to consider is that maybe your CPU's cores will be used more efficiently by processing all of the tracks separately with their own effects as opposed to processing them with the VSTi's own onboard effects unit, which may or may not spike one of your cores. This is something I have noticed with Native Instrument's Drummer libraries - it has its own internal mixer with effects processing, but it tends to spike one of my cores like crazy if I use it. However I also have all of the effects it uses as separate VST's and if I route all of the drums to multiple outputs and process them separately using those same VST's, I don't get the spike. 

That's a very interesting tidbit, thanks. I sure wish there was a way for the collective knowledge that surfaces in the forum could be put into some kind of database, other than Google search of course...

Anderton

sharke
Another thing to consider is that maybe your CPU's cores will be used more efficiently by processing all of the tracks separately with their own effects as opposed to processing them with the VSTi's own onboard effects unit, which may or may not spike one of your cores. This is something I have noticed with Native Instrument's Drummer libraries - it has its own internal mixer with effects processing, but it tends to spike one of my cores like crazy if I use it. However I also have all of the effects it uses as separate VST's and if I route all of the drums to multiple outputs and process them separately using those same VST's, I don't get the spike. 

That's a very interesting tidbit, thanks. I sure wish there was a way for the collective knowledge that surfaces in the forum could be put into some kind of database, other than Google search of course...

The whole issue of core peaking and load spreading seems to be very vague and driven more by anecdotal evidence than anything else. I've never heard anyone assert a definitive statement on how it all works tbh. I'm sure the Bakers know a lot more about it than us.